diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,69333 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 7.121021399028952, + "eval_steps": 500, + "global_step": 9900, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0007192950908110052, + "grad_norm": 64.34911621183012, + "learning_rate": 6.000000000000001e-08, + "loss": 4.4253, + "step": 1 + }, + { + "epoch": 0.0014385901816220104, + "grad_norm": 51.27350234777956, + "learning_rate": 1.2000000000000002e-07, + "loss": 4.157, + "step": 2 + }, + { + "epoch": 0.002157885272433016, + "grad_norm": 58.38570787634046, + "learning_rate": 1.8e-07, + "loss": 4.3753, + "step": 3 + }, + { + "epoch": 0.002877180363244021, + "grad_norm": 60.33603591467569, + "learning_rate": 2.4000000000000003e-07, + "loss": 4.7219, + "step": 4 + }, + { + "epoch": 0.0035964754540550262, + "grad_norm": 63.405779091424165, + "learning_rate": 3.0000000000000004e-07, + "loss": 4.4572, + "step": 5 + }, + { + "epoch": 0.004315770544866032, + "grad_norm": 57.43191624928036, + "learning_rate": 3.6e-07, + "loss": 4.3722, + "step": 6 + }, + { + "epoch": 0.005035065635677036, + "grad_norm": 45.80791469017583, + "learning_rate": 4.2000000000000006e-07, + "loss": 4.3694, + "step": 7 + }, + { + "epoch": 0.005754360726488042, + "grad_norm": 52.561254618344556, + "learning_rate": 4.800000000000001e-07, + "loss": 4.2124, + "step": 8 + }, + { + "epoch": 0.006473655817299047, + "grad_norm": 52.714545202649816, + "learning_rate": 5.4e-07, + "loss": 4.1141, + "step": 9 + }, + { + "epoch": 0.0071929509081100525, + "grad_norm": 63.569929213232236, + "learning_rate": 6.000000000000001e-07, + "loss": 4.4139, + "step": 10 + }, + { + "epoch": 0.007912245998921057, + "grad_norm": 52.68955747119893, + "learning_rate": 6.6e-07, + "loss": 4.37, + "step": 11 + }, + { + "epoch": 0.008631541089732063, + "grad_norm": 44.654498976303714, + "learning_rate": 7.2e-07, + "loss": 3.9797, + "step": 12 + }, + { + "epoch": 0.009350836180543068, + "grad_norm": 50.05528844599175, + "learning_rate": 7.8e-07, + "loss": 3.9772, + "step": 13 + }, + { + "epoch": 0.010070131271354072, + "grad_norm": 67.73011138752118, + "learning_rate": 8.400000000000001e-07, + "loss": 4.2362, + "step": 14 + }, + { + "epoch": 0.010789426362165079, + "grad_norm": 45.67621010960208, + "learning_rate": 9e-07, + "loss": 4.0604, + "step": 15 + }, + { + "epoch": 0.011508721452976083, + "grad_norm": 66.74676866766224, + "learning_rate": 9.600000000000001e-07, + "loss": 4.1516, + "step": 16 + }, + { + "epoch": 0.012228016543787088, + "grad_norm": 48.062555064438754, + "learning_rate": 1.0200000000000002e-06, + "loss": 4.1489, + "step": 17 + }, + { + "epoch": 0.012947311634598094, + "grad_norm": 42.64480446155265, + "learning_rate": 1.08e-06, + "loss": 3.5589, + "step": 18 + }, + { + "epoch": 0.013666606725409099, + "grad_norm": 42.07904498757481, + "learning_rate": 1.14e-06, + "loss": 3.9886, + "step": 19 + }, + { + "epoch": 0.014385901816220105, + "grad_norm": 48.72108700022635, + "learning_rate": 1.2000000000000002e-06, + "loss": 3.4519, + "step": 20 + }, + { + "epoch": 0.01510519690703111, + "grad_norm": 42.250033849621516, + "learning_rate": 1.26e-06, + "loss": 3.303, + "step": 21 + }, + { + "epoch": 0.015824491997842114, + "grad_norm": 33.70223333768381, + "learning_rate": 1.32e-06, + "loss": 3.6903, + "step": 22 + }, + { + "epoch": 0.01654378708865312, + "grad_norm": 40.15971992201178, + "learning_rate": 1.3800000000000001e-06, + "loss": 3.5432, + "step": 23 + }, + { + "epoch": 0.017263082179464127, + "grad_norm": 34.63347885795475, + "learning_rate": 1.44e-06, + "loss": 3.0621, + "step": 24 + }, + { + "epoch": 0.01798237727027513, + "grad_norm": 25.449455048743598, + "learning_rate": 1.5e-06, + "loss": 2.8903, + "step": 25 + }, + { + "epoch": 0.018701672361086136, + "grad_norm": 21.57750342260305, + "learning_rate": 1.56e-06, + "loss": 2.8485, + "step": 26 + }, + { + "epoch": 0.019420967451897142, + "grad_norm": 25.591838990527403, + "learning_rate": 1.6200000000000002e-06, + "loss": 2.8427, + "step": 27 + }, + { + "epoch": 0.020140262542708145, + "grad_norm": 21.888134103825376, + "learning_rate": 1.6800000000000002e-06, + "loss": 2.5554, + "step": 28 + }, + { + "epoch": 0.02085955763351915, + "grad_norm": 23.093437270571588, + "learning_rate": 1.7399999999999999e-06, + "loss": 2.5745, + "step": 29 + }, + { + "epoch": 0.021578852724330157, + "grad_norm": 22.408129250183837, + "learning_rate": 1.8e-06, + "loss": 2.8245, + "step": 30 + }, + { + "epoch": 0.02229814781514116, + "grad_norm": 21.62571464737288, + "learning_rate": 1.86e-06, + "loss": 2.7638, + "step": 31 + }, + { + "epoch": 0.023017442905952167, + "grad_norm": 22.1754794247378, + "learning_rate": 1.9200000000000003e-06, + "loss": 2.566, + "step": 32 + }, + { + "epoch": 0.023736737996763173, + "grad_norm": 17.36822178512086, + "learning_rate": 1.98e-06, + "loss": 2.3873, + "step": 33 + }, + { + "epoch": 0.024456033087574176, + "grad_norm": 20.82250738790257, + "learning_rate": 2.0400000000000004e-06, + "loss": 2.4609, + "step": 34 + }, + { + "epoch": 0.025175328178385182, + "grad_norm": 21.727374515734958, + "learning_rate": 2.1e-06, + "loss": 2.104, + "step": 35 + }, + { + "epoch": 0.025894623269196188, + "grad_norm": 25.192785218202438, + "learning_rate": 2.16e-06, + "loss": 2.2068, + "step": 36 + }, + { + "epoch": 0.026613918360007194, + "grad_norm": 22.949876065186068, + "learning_rate": 2.22e-06, + "loss": 2.1981, + "step": 37 + }, + { + "epoch": 0.027333213450818197, + "grad_norm": 20.452749514647596, + "learning_rate": 2.28e-06, + "loss": 2.1726, + "step": 38 + }, + { + "epoch": 0.028052508541629204, + "grad_norm": 19.783955544164957, + "learning_rate": 2.34e-06, + "loss": 2.0562, + "step": 39 + }, + { + "epoch": 0.02877180363244021, + "grad_norm": 21.913942051708446, + "learning_rate": 2.4000000000000003e-06, + "loss": 2.5344, + "step": 40 + }, + { + "epoch": 0.029491098723251213, + "grad_norm": 16.568586126115413, + "learning_rate": 2.4599999999999997e-06, + "loss": 2.1713, + "step": 41 + }, + { + "epoch": 0.03021039381406222, + "grad_norm": 15.232605106865144, + "learning_rate": 2.52e-06, + "loss": 2.2402, + "step": 42 + }, + { + "epoch": 0.030929688904873225, + "grad_norm": 15.106443009338856, + "learning_rate": 2.58e-06, + "loss": 2.304, + "step": 43 + }, + { + "epoch": 0.03164898399568423, + "grad_norm": 15.979582692190226, + "learning_rate": 2.64e-06, + "loss": 2.0808, + "step": 44 + }, + { + "epoch": 0.032368279086495234, + "grad_norm": 16.092017705403578, + "learning_rate": 2.7e-06, + "loss": 2.3539, + "step": 45 + }, + { + "epoch": 0.03308757417730624, + "grad_norm": 16.06526364336813, + "learning_rate": 2.7600000000000003e-06, + "loss": 2.2053, + "step": 46 + }, + { + "epoch": 0.03380686926811725, + "grad_norm": 14.316985257725126, + "learning_rate": 2.82e-06, + "loss": 1.9945, + "step": 47 + }, + { + "epoch": 0.03452616435892825, + "grad_norm": 14.958090177786831, + "learning_rate": 2.88e-06, + "loss": 2.1573, + "step": 48 + }, + { + "epoch": 0.03524545944973925, + "grad_norm": 14.026710889707836, + "learning_rate": 2.9400000000000002e-06, + "loss": 2.2067, + "step": 49 + }, + { + "epoch": 0.03596475454055026, + "grad_norm": 14.51423023014951, + "learning_rate": 3e-06, + "loss": 1.9742, + "step": 50 + }, + { + "epoch": 0.036684049631361265, + "grad_norm": 15.22192441616925, + "learning_rate": 2.999999961411183e-06, + "loss": 2.0712, + "step": 51 + }, + { + "epoch": 0.03740334472217227, + "grad_norm": 14.194763449787578, + "learning_rate": 2.9999998456447355e-06, + "loss": 1.9204, + "step": 52 + }, + { + "epoch": 0.03812263981298328, + "grad_norm": 13.667584041378175, + "learning_rate": 2.9999996527006628e-06, + "loss": 1.8722, + "step": 53 + }, + { + "epoch": 0.038841934903794284, + "grad_norm": 14.801818205020217, + "learning_rate": 2.9999993825789743e-06, + "loss": 1.9964, + "step": 54 + }, + { + "epoch": 0.03956122999460528, + "grad_norm": 14.031676772035228, + "learning_rate": 2.999999035279685e-06, + "loss": 1.7175, + "step": 55 + }, + { + "epoch": 0.04028052508541629, + "grad_norm": 15.806689286588222, + "learning_rate": 2.9999986108028114e-06, + "loss": 1.7233, + "step": 56 + }, + { + "epoch": 0.040999820176227296, + "grad_norm": 17.51440992749227, + "learning_rate": 2.9999981091483768e-06, + "loss": 1.9272, + "step": 57 + }, + { + "epoch": 0.0417191152670383, + "grad_norm": 13.741266330932344, + "learning_rate": 2.9999975303164056e-06, + "loss": 1.954, + "step": 58 + }, + { + "epoch": 0.04243841035784931, + "grad_norm": 18.39147673593491, + "learning_rate": 2.999996874306929e-06, + "loss": 1.8145, + "step": 59 + }, + { + "epoch": 0.043157705448660315, + "grad_norm": 18.716766390380016, + "learning_rate": 2.99999614111998e-06, + "loss": 1.8562, + "step": 60 + }, + { + "epoch": 0.04387700053947132, + "grad_norm": 15.030579384425387, + "learning_rate": 2.999995330755596e-06, + "loss": 1.7145, + "step": 61 + }, + { + "epoch": 0.04459629563028232, + "grad_norm": 16.27860968362536, + "learning_rate": 2.999994443213819e-06, + "loss": 1.4302, + "step": 62 + }, + { + "epoch": 0.04531559072109333, + "grad_norm": 18.137559091067445, + "learning_rate": 2.999993478494695e-06, + "loss": 2.0254, + "step": 63 + }, + { + "epoch": 0.04603488581190433, + "grad_norm": 15.51057143193802, + "learning_rate": 2.9999924365982735e-06, + "loss": 1.5322, + "step": 64 + }, + { + "epoch": 0.04675418090271534, + "grad_norm": 18.095173249301645, + "learning_rate": 2.9999913175246076e-06, + "loss": 1.7118, + "step": 65 + }, + { + "epoch": 0.047473475993526346, + "grad_norm": 17.045312399277137, + "learning_rate": 2.9999901212737557e-06, + "loss": 1.642, + "step": 66 + }, + { + "epoch": 0.04819277108433735, + "grad_norm": 19.706738120317652, + "learning_rate": 2.9999888478457784e-06, + "loss": 1.5376, + "step": 67 + }, + { + "epoch": 0.04891206617514835, + "grad_norm": 32.97972115668073, + "learning_rate": 2.9999874972407424e-06, + "loss": 1.7039, + "step": 68 + }, + { + "epoch": 0.04963136126595936, + "grad_norm": 21.563861795013054, + "learning_rate": 2.999986069458716e-06, + "loss": 1.4447, + "step": 69 + }, + { + "epoch": 0.050350656356770364, + "grad_norm": 17.419159113718333, + "learning_rate": 2.9999845644997736e-06, + "loss": 1.5097, + "step": 70 + }, + { + "epoch": 0.05106995144758137, + "grad_norm": 21.98894598263102, + "learning_rate": 2.9999829823639916e-06, + "loss": 1.4381, + "step": 71 + }, + { + "epoch": 0.051789246538392376, + "grad_norm": 17.957091655713384, + "learning_rate": 2.9999813230514524e-06, + "loss": 1.6295, + "step": 72 + }, + { + "epoch": 0.05250854162920338, + "grad_norm": 18.191843369307144, + "learning_rate": 2.9999795865622413e-06, + "loss": 1.394, + "step": 73 + }, + { + "epoch": 0.05322783672001439, + "grad_norm": 15.974616072378568, + "learning_rate": 2.9999777728964466e-06, + "loss": 1.153, + "step": 74 + }, + { + "epoch": 0.05394713181082539, + "grad_norm": 15.851738860415221, + "learning_rate": 2.999975882054163e-06, + "loss": 1.5358, + "step": 75 + }, + { + "epoch": 0.054666426901636395, + "grad_norm": 18.13076452589248, + "learning_rate": 2.999973914035487e-06, + "loss": 1.031, + "step": 76 + }, + { + "epoch": 0.0553857219924474, + "grad_norm": 20.323852306393164, + "learning_rate": 2.9999718688405194e-06, + "loss": 1.2072, + "step": 77 + }, + { + "epoch": 0.05610501708325841, + "grad_norm": 16.390106814037097, + "learning_rate": 2.9999697464693666e-06, + "loss": 1.1953, + "step": 78 + }, + { + "epoch": 0.056824312174069413, + "grad_norm": 13.93589642424244, + "learning_rate": 2.9999675469221375e-06, + "loss": 1.082, + "step": 79 + }, + { + "epoch": 0.05754360726488042, + "grad_norm": 12.913261699312807, + "learning_rate": 2.9999652701989444e-06, + "loss": 1.0928, + "step": 80 + }, + { + "epoch": 0.05826290235569142, + "grad_norm": 13.776128845237146, + "learning_rate": 2.9999629162999054e-06, + "loss": 1.3075, + "step": 81 + }, + { + "epoch": 0.058982197446502425, + "grad_norm": 15.81637358851, + "learning_rate": 2.999960485225141e-06, + "loss": 1.2325, + "step": 82 + }, + { + "epoch": 0.05970149253731343, + "grad_norm": 20.356411498268486, + "learning_rate": 2.999957976974777e-06, + "loss": 0.9464, + "step": 83 + }, + { + "epoch": 0.06042078762812444, + "grad_norm": 18.885207613766585, + "learning_rate": 2.999955391548942e-06, + "loss": 1.1445, + "step": 84 + }, + { + "epoch": 0.061140082718935444, + "grad_norm": 17.14246664289988, + "learning_rate": 2.999952728947768e-06, + "loss": 1.3846, + "step": 85 + }, + { + "epoch": 0.06185937780974645, + "grad_norm": 17.689753014018272, + "learning_rate": 2.999949989171394e-06, + "loss": 1.1818, + "step": 86 + }, + { + "epoch": 0.06257867290055745, + "grad_norm": 19.517032972882443, + "learning_rate": 2.9999471722199595e-06, + "loss": 1.0695, + "step": 87 + }, + { + "epoch": 0.06329796799136846, + "grad_norm": 17.38660383027941, + "learning_rate": 2.9999442780936104e-06, + "loss": 1.1583, + "step": 88 + }, + { + "epoch": 0.06401726308217946, + "grad_norm": 15.48079391320298, + "learning_rate": 2.999941306792495e-06, + "loss": 1.3267, + "step": 89 + }, + { + "epoch": 0.06473655817299047, + "grad_norm": 15.360783627227772, + "learning_rate": 2.999938258316766e-06, + "loss": 1.0709, + "step": 90 + }, + { + "epoch": 0.06545585326380148, + "grad_norm": 17.381248590193184, + "learning_rate": 2.999935132666581e-06, + "loss": 1.2396, + "step": 91 + }, + { + "epoch": 0.06617514835461248, + "grad_norm": 15.682577761416166, + "learning_rate": 2.9999319298420998e-06, + "loss": 1.1373, + "step": 92 + }, + { + "epoch": 0.06689444344542349, + "grad_norm": 15.080397284117504, + "learning_rate": 2.999928649843488e-06, + "loss": 0.9047, + "step": 93 + }, + { + "epoch": 0.0676137385362345, + "grad_norm": 14.62098720233514, + "learning_rate": 2.999925292670914e-06, + "loss": 1.1609, + "step": 94 + }, + { + "epoch": 0.0683330336270455, + "grad_norm": 13.450215274233114, + "learning_rate": 2.999921858324551e-06, + "loss": 0.9095, + "step": 95 + }, + { + "epoch": 0.0690523287178565, + "grad_norm": 15.936356086252578, + "learning_rate": 2.999918346804575e-06, + "loss": 1.0574, + "step": 96 + }, + { + "epoch": 0.0697716238086675, + "grad_norm": 11.361017434660084, + "learning_rate": 2.9999147581111674e-06, + "loss": 0.9188, + "step": 97 + }, + { + "epoch": 0.0704909188994785, + "grad_norm": 16.136922711242608, + "learning_rate": 2.9999110922445124e-06, + "loss": 0.8518, + "step": 98 + }, + { + "epoch": 0.07121021399028951, + "grad_norm": 14.479550155285834, + "learning_rate": 2.9999073492047983e-06, + "loss": 0.9733, + "step": 99 + }, + { + "epoch": 0.07192950908110052, + "grad_norm": 14.14720055173962, + "learning_rate": 2.9999035289922186e-06, + "loss": 0.9136, + "step": 100 + }, + { + "epoch": 0.07264880417191152, + "grad_norm": 14.460728505543798, + "learning_rate": 2.999899631606969e-06, + "loss": 0.7732, + "step": 101 + }, + { + "epoch": 0.07336809926272253, + "grad_norm": 14.915994889708552, + "learning_rate": 2.9998956570492504e-06, + "loss": 1.0222, + "step": 102 + }, + { + "epoch": 0.07408739435353354, + "grad_norm": 13.798774107026125, + "learning_rate": 2.9998916053192673e-06, + "loss": 0.9379, + "step": 103 + }, + { + "epoch": 0.07480668944434454, + "grad_norm": 16.45470711743046, + "learning_rate": 2.999887476417228e-06, + "loss": 1.1023, + "step": 104 + }, + { + "epoch": 0.07552598453515555, + "grad_norm": 16.82514571020224, + "learning_rate": 2.9998832703433456e-06, + "loss": 0.9055, + "step": 105 + }, + { + "epoch": 0.07624527962596656, + "grad_norm": 18.355401485360265, + "learning_rate": 2.9998789870978352e-06, + "loss": 1.0206, + "step": 106 + }, + { + "epoch": 0.07696457471677756, + "grad_norm": 17.01847252982604, + "learning_rate": 2.9998746266809183e-06, + "loss": 0.9563, + "step": 107 + }, + { + "epoch": 0.07768386980758857, + "grad_norm": 13.709892094376332, + "learning_rate": 2.9998701890928187e-06, + "loss": 0.6598, + "step": 108 + }, + { + "epoch": 0.07840316489839957, + "grad_norm": 17.46028171479583, + "learning_rate": 2.9998656743337654e-06, + "loss": 0.894, + "step": 109 + }, + { + "epoch": 0.07912245998921057, + "grad_norm": 13.80409609539761, + "learning_rate": 2.99986108240399e-06, + "loss": 0.7213, + "step": 110 + }, + { + "epoch": 0.07984175508002157, + "grad_norm": 17.175494336691322, + "learning_rate": 2.999856413303729e-06, + "loss": 0.9785, + "step": 111 + }, + { + "epoch": 0.08056105017083258, + "grad_norm": 18.60010165596918, + "learning_rate": 2.9998516670332227e-06, + "loss": 1.0314, + "step": 112 + }, + { + "epoch": 0.08128034526164359, + "grad_norm": 14.885433927059273, + "learning_rate": 2.9998468435927154e-06, + "loss": 0.8089, + "step": 113 + }, + { + "epoch": 0.08199964035245459, + "grad_norm": 16.42747769780196, + "learning_rate": 2.999841942982455e-06, + "loss": 0.9663, + "step": 114 + }, + { + "epoch": 0.0827189354432656, + "grad_norm": 15.279786015115542, + "learning_rate": 2.999836965202693e-06, + "loss": 0.7738, + "step": 115 + }, + { + "epoch": 0.0834382305340766, + "grad_norm": 14.116665198902021, + "learning_rate": 2.9998319102536868e-06, + "loss": 0.9416, + "step": 116 + }, + { + "epoch": 0.08415752562488761, + "grad_norm": 13.104497420126021, + "learning_rate": 2.9998267781356964e-06, + "loss": 0.6117, + "step": 117 + }, + { + "epoch": 0.08487682071569862, + "grad_norm": 14.089846442513345, + "learning_rate": 2.999821568848985e-06, + "loss": 0.9485, + "step": 118 + }, + { + "epoch": 0.08559611580650962, + "grad_norm": 17.090806029715722, + "learning_rate": 2.999816282393821e-06, + "loss": 0.9913, + "step": 119 + }, + { + "epoch": 0.08631541089732063, + "grad_norm": 17.008896450934316, + "learning_rate": 2.9998109187704765e-06, + "loss": 0.9245, + "step": 120 + }, + { + "epoch": 0.08703470598813164, + "grad_norm": 17.121880191450334, + "learning_rate": 2.999805477979227e-06, + "loss": 0.7626, + "step": 121 + }, + { + "epoch": 0.08775400107894264, + "grad_norm": 14.606147063625393, + "learning_rate": 2.9997999600203533e-06, + "loss": 0.8401, + "step": 122 + }, + { + "epoch": 0.08847329616975363, + "grad_norm": 17.397779605226102, + "learning_rate": 2.999794364894139e-06, + "loss": 0.7384, + "step": 123 + }, + { + "epoch": 0.08919259126056464, + "grad_norm": 16.957065813386407, + "learning_rate": 2.9997886926008713e-06, + "loss": 0.8409, + "step": 124 + }, + { + "epoch": 0.08991188635137565, + "grad_norm": 17.929248752097, + "learning_rate": 2.999782943140843e-06, + "loss": 0.8461, + "step": 125 + }, + { + "epoch": 0.09063118144218665, + "grad_norm": 12.05064245598457, + "learning_rate": 2.999777116514349e-06, + "loss": 0.7481, + "step": 126 + }, + { + "epoch": 0.09135047653299766, + "grad_norm": 15.70261153256519, + "learning_rate": 2.99977121272169e-06, + "loss": 0.8238, + "step": 127 + }, + { + "epoch": 0.09206977162380867, + "grad_norm": 17.78830125589999, + "learning_rate": 2.9997652317631693e-06, + "loss": 0.7911, + "step": 128 + }, + { + "epoch": 0.09278906671461967, + "grad_norm": 13.864250229628292, + "learning_rate": 2.9997591736390945e-06, + "loss": 0.7995, + "step": 129 + }, + { + "epoch": 0.09350836180543068, + "grad_norm": 15.502336013451172, + "learning_rate": 2.9997530383497773e-06, + "loss": 0.8459, + "step": 130 + }, + { + "epoch": 0.09422765689624168, + "grad_norm": 15.540040407710661, + "learning_rate": 2.999746825895534e-06, + "loss": 0.5619, + "step": 131 + }, + { + "epoch": 0.09494695198705269, + "grad_norm": 14.749204467275762, + "learning_rate": 2.9997405362766832e-06, + "loss": 0.5666, + "step": 132 + }, + { + "epoch": 0.0956662470778637, + "grad_norm": 12.622048470319367, + "learning_rate": 2.99973416949355e-06, + "loss": 0.8049, + "step": 133 + }, + { + "epoch": 0.0963855421686747, + "grad_norm": 14.279358326559988, + "learning_rate": 2.9997277255464603e-06, + "loss": 0.69, + "step": 134 + }, + { + "epoch": 0.09710483725948571, + "grad_norm": 14.37701622582175, + "learning_rate": 2.999721204435747e-06, + "loss": 0.7442, + "step": 135 + }, + { + "epoch": 0.0978241323502967, + "grad_norm": 14.970198887449568, + "learning_rate": 2.999714606161745e-06, + "loss": 0.7065, + "step": 136 + }, + { + "epoch": 0.09854342744110771, + "grad_norm": 11.087967774569819, + "learning_rate": 2.9997079307247934e-06, + "loss": 0.4393, + "step": 137 + }, + { + "epoch": 0.09926272253191872, + "grad_norm": 14.08146398456673, + "learning_rate": 2.999701178125236e-06, + "loss": 0.6798, + "step": 138 + }, + { + "epoch": 0.09998201762272972, + "grad_norm": 11.636590991525697, + "learning_rate": 2.999694348363421e-06, + "loss": 0.5427, + "step": 139 + }, + { + "epoch": 0.10070131271354073, + "grad_norm": 11.641045285267525, + "learning_rate": 2.9996874414396985e-06, + "loss": 0.4303, + "step": 140 + }, + { + "epoch": 0.10142060780435173, + "grad_norm": 15.17560296873663, + "learning_rate": 2.999680457354425e-06, + "loss": 0.648, + "step": 141 + }, + { + "epoch": 0.10213990289516274, + "grad_norm": 14.812459173916162, + "learning_rate": 2.999673396107959e-06, + "loss": 0.7599, + "step": 142 + }, + { + "epoch": 0.10285919798597375, + "grad_norm": 12.417508363153125, + "learning_rate": 2.9996662577006645e-06, + "loss": 0.5857, + "step": 143 + }, + { + "epoch": 0.10357849307678475, + "grad_norm": 13.864225195088665, + "learning_rate": 2.999659042132908e-06, + "loss": 0.9716, + "step": 144 + }, + { + "epoch": 0.10429778816759576, + "grad_norm": 12.987253142548148, + "learning_rate": 2.9996517494050618e-06, + "loss": 0.6616, + "step": 145 + }, + { + "epoch": 0.10501708325840677, + "grad_norm": 11.110045110963247, + "learning_rate": 2.9996443795175005e-06, + "loss": 0.6188, + "step": 146 + }, + { + "epoch": 0.10573637834921777, + "grad_norm": 16.827417071411812, + "learning_rate": 2.999636932470603e-06, + "loss": 0.5019, + "step": 147 + }, + { + "epoch": 0.10645567344002878, + "grad_norm": 14.899423395901154, + "learning_rate": 2.999629408264753e-06, + "loss": 0.7334, + "step": 148 + }, + { + "epoch": 0.10717496853083977, + "grad_norm": 15.329389828315854, + "learning_rate": 2.999621806900337e-06, + "loss": 0.7033, + "step": 149 + }, + { + "epoch": 0.10789426362165078, + "grad_norm": 15.779200047974253, + "learning_rate": 2.999614128377747e-06, + "loss": 0.8697, + "step": 150 + }, + { + "epoch": 0.10861355871246178, + "grad_norm": 18.170095046078853, + "learning_rate": 2.9996063726973773e-06, + "loss": 0.767, + "step": 151 + }, + { + "epoch": 0.10933285380327279, + "grad_norm": 12.716753914369793, + "learning_rate": 2.9995985398596274e-06, + "loss": 0.5188, + "step": 152 + }, + { + "epoch": 0.1100521488940838, + "grad_norm": 18.130394895813886, + "learning_rate": 2.9995906298649e-06, + "loss": 0.6062, + "step": 153 + }, + { + "epoch": 0.1107714439848948, + "grad_norm": 12.356114027945125, + "learning_rate": 2.9995826427136028e-06, + "loss": 0.5099, + "step": 154 + }, + { + "epoch": 0.11149073907570581, + "grad_norm": 16.007003122424482, + "learning_rate": 2.999574578406146e-06, + "loss": 0.5008, + "step": 155 + }, + { + "epoch": 0.11221003416651681, + "grad_norm": 15.156259796857944, + "learning_rate": 2.9995664369429444e-06, + "loss": 0.5808, + "step": 156 + }, + { + "epoch": 0.11292932925732782, + "grad_norm": 13.664453171228516, + "learning_rate": 2.9995582183244173e-06, + "loss": 0.7076, + "step": 157 + }, + { + "epoch": 0.11364862434813883, + "grad_norm": 17.933972247998355, + "learning_rate": 2.9995499225509876e-06, + "loss": 0.8485, + "step": 158 + }, + { + "epoch": 0.11436791943894983, + "grad_norm": 13.377851473531214, + "learning_rate": 2.9995415496230818e-06, + "loss": 0.5764, + "step": 159 + }, + { + "epoch": 0.11508721452976084, + "grad_norm": 16.776637613444294, + "learning_rate": 2.999533099541131e-06, + "loss": 0.5597, + "step": 160 + }, + { + "epoch": 0.11580650962057185, + "grad_norm": 10.738040442707506, + "learning_rate": 2.99952457230557e-06, + "loss": 0.4552, + "step": 161 + }, + { + "epoch": 0.11652580471138284, + "grad_norm": 15.306648055364423, + "learning_rate": 2.9995159679168376e-06, + "loss": 0.656, + "step": 162 + }, + { + "epoch": 0.11724509980219384, + "grad_norm": 12.17035497932848, + "learning_rate": 2.999507286375376e-06, + "loss": 0.5925, + "step": 163 + }, + { + "epoch": 0.11796439489300485, + "grad_norm": 13.543791536121919, + "learning_rate": 2.999498527681632e-06, + "loss": 0.4981, + "step": 164 + }, + { + "epoch": 0.11868368998381586, + "grad_norm": 13.118076682057563, + "learning_rate": 2.999489691836057e-06, + "loss": 0.4965, + "step": 165 + }, + { + "epoch": 0.11940298507462686, + "grad_norm": 12.71381775320525, + "learning_rate": 2.9994807788391045e-06, + "loss": 0.471, + "step": 166 + }, + { + "epoch": 0.12012228016543787, + "grad_norm": 12.399044755111824, + "learning_rate": 2.9994717886912343e-06, + "loss": 0.5591, + "step": 167 + }, + { + "epoch": 0.12084157525624888, + "grad_norm": 15.706549334877247, + "learning_rate": 2.999462721392908e-06, + "loss": 0.5651, + "step": 168 + }, + { + "epoch": 0.12156087034705988, + "grad_norm": 16.52354346421275, + "learning_rate": 2.9994535769445925e-06, + "loss": 0.8633, + "step": 169 + }, + { + "epoch": 0.12228016543787089, + "grad_norm": 18.27461329136108, + "learning_rate": 2.9994443553467584e-06, + "loss": 0.7281, + "step": 170 + }, + { + "epoch": 0.1229994605286819, + "grad_norm": 12.463490196038022, + "learning_rate": 2.99943505659988e-06, + "loss": 0.4946, + "step": 171 + }, + { + "epoch": 0.1237187556194929, + "grad_norm": 17.214653610695475, + "learning_rate": 2.9994256807044354e-06, + "loss": 0.7973, + "step": 172 + }, + { + "epoch": 0.12443805071030391, + "grad_norm": 14.550749335726373, + "learning_rate": 2.9994162276609078e-06, + "loss": 0.6072, + "step": 173 + }, + { + "epoch": 0.1251573458011149, + "grad_norm": 11.27619069006964, + "learning_rate": 2.9994066974697828e-06, + "loss": 0.4613, + "step": 174 + }, + { + "epoch": 0.1258766408919259, + "grad_norm": 16.14197403634099, + "learning_rate": 2.9993970901315513e-06, + "loss": 0.611, + "step": 175 + }, + { + "epoch": 0.1265959359827369, + "grad_norm": 11.947668987471774, + "learning_rate": 2.999387405646707e-06, + "loss": 0.5964, + "step": 176 + }, + { + "epoch": 0.12731523107354792, + "grad_norm": 11.917806216130602, + "learning_rate": 2.999377644015749e-06, + "loss": 0.5331, + "step": 177 + }, + { + "epoch": 0.12803452616435892, + "grad_norm": 10.823123910733955, + "learning_rate": 2.9993678052391797e-06, + "loss": 0.6022, + "step": 178 + }, + { + "epoch": 0.12875382125516993, + "grad_norm": 10.724683122415733, + "learning_rate": 2.9993578893175036e-06, + "loss": 0.3974, + "step": 179 + }, + { + "epoch": 0.12947311634598094, + "grad_norm": 11.844584382631318, + "learning_rate": 2.999347896251233e-06, + "loss": 0.3949, + "step": 180 + }, + { + "epoch": 0.13019241143679194, + "grad_norm": 11.831219301309252, + "learning_rate": 2.9993378260408804e-06, + "loss": 0.3115, + "step": 181 + }, + { + "epoch": 0.13091170652760295, + "grad_norm": 11.061144669012819, + "learning_rate": 2.9993276786869654e-06, + "loss": 0.3338, + "step": 182 + }, + { + "epoch": 0.13163100161841396, + "grad_norm": 14.56869049925627, + "learning_rate": 2.9993174541900088e-06, + "loss": 0.6595, + "step": 183 + }, + { + "epoch": 0.13235029670922496, + "grad_norm": 11.654208837607412, + "learning_rate": 2.9993071525505375e-06, + "loss": 0.5349, + "step": 184 + }, + { + "epoch": 0.13306959180003597, + "grad_norm": 15.254812193241476, + "learning_rate": 2.9992967737690808e-06, + "loss": 0.5788, + "step": 185 + }, + { + "epoch": 0.13378888689084698, + "grad_norm": 14.96019408719912, + "learning_rate": 2.9992863178461737e-06, + "loss": 0.6025, + "step": 186 + }, + { + "epoch": 0.13450818198165798, + "grad_norm": 15.053189519836618, + "learning_rate": 2.999275784782353e-06, + "loss": 0.5472, + "step": 187 + }, + { + "epoch": 0.135227477072469, + "grad_norm": 20.259725486038164, + "learning_rate": 2.9992651745781617e-06, + "loss": 0.747, + "step": 188 + }, + { + "epoch": 0.13594677216328, + "grad_norm": 11.486538136158382, + "learning_rate": 2.9992544872341453e-06, + "loss": 0.2998, + "step": 189 + }, + { + "epoch": 0.136666067254091, + "grad_norm": 12.60224429554145, + "learning_rate": 2.9992437227508536e-06, + "loss": 0.5738, + "step": 190 + }, + { + "epoch": 0.137385362344902, + "grad_norm": 12.630395204793585, + "learning_rate": 2.9992328811288407e-06, + "loss": 0.4701, + "step": 191 + }, + { + "epoch": 0.138104657435713, + "grad_norm": 16.38951406253093, + "learning_rate": 2.9992219623686637e-06, + "loss": 0.6143, + "step": 192 + }, + { + "epoch": 0.13882395252652402, + "grad_norm": 12.247087834394414, + "learning_rate": 2.9992109664708857e-06, + "loss": 0.5247, + "step": 193 + }, + { + "epoch": 0.139543247617335, + "grad_norm": 13.261523423865471, + "learning_rate": 2.999199893436071e-06, + "loss": 0.4652, + "step": 194 + }, + { + "epoch": 0.140262542708146, + "grad_norm": 16.34509283856271, + "learning_rate": 2.9991887432647905e-06, + "loss": 0.5315, + "step": 195 + }, + { + "epoch": 0.140981837798957, + "grad_norm": 12.950869035317044, + "learning_rate": 2.999177515957617e-06, + "loss": 0.3374, + "step": 196 + }, + { + "epoch": 0.14170113288976802, + "grad_norm": 12.263787315778378, + "learning_rate": 2.999166211515129e-06, + "loss": 0.4732, + "step": 197 + }, + { + "epoch": 0.14242042798057902, + "grad_norm": 12.788931392269063, + "learning_rate": 2.9991548299379073e-06, + "loss": 0.5551, + "step": 198 + }, + { + "epoch": 0.14313972307139003, + "grad_norm": 10.989185749610701, + "learning_rate": 2.9991433712265382e-06, + "loss": 0.423, + "step": 199 + }, + { + "epoch": 0.14385901816220104, + "grad_norm": 14.02524180139987, + "learning_rate": 2.999131835381611e-06, + "loss": 0.5528, + "step": 200 + }, + { + "epoch": 0.14457831325301204, + "grad_norm": 12.568161708376003, + "learning_rate": 2.9991202224037193e-06, + "loss": 0.5286, + "step": 201 + }, + { + "epoch": 0.14529760834382305, + "grad_norm": 15.398108360402558, + "learning_rate": 2.9991085322934607e-06, + "loss": 0.4578, + "step": 202 + }, + { + "epoch": 0.14601690343463405, + "grad_norm": 14.441341458041183, + "learning_rate": 2.9990967650514363e-06, + "loss": 0.5822, + "step": 203 + }, + { + "epoch": 0.14673619852544506, + "grad_norm": 14.364747704730616, + "learning_rate": 2.9990849206782516e-06, + "loss": 0.4663, + "step": 204 + }, + { + "epoch": 0.14745549361625607, + "grad_norm": 16.574475031532405, + "learning_rate": 2.999072999174516e-06, + "loss": 0.4073, + "step": 205 + }, + { + "epoch": 0.14817478870706707, + "grad_norm": 16.82965096041825, + "learning_rate": 2.9990610005408434e-06, + "loss": 0.5181, + "step": 206 + }, + { + "epoch": 0.14889408379787808, + "grad_norm": 8.874566029723, + "learning_rate": 2.9990489247778507e-06, + "loss": 0.2668, + "step": 207 + }, + { + "epoch": 0.14961337888868909, + "grad_norm": 10.609333465643331, + "learning_rate": 2.9990367718861596e-06, + "loss": 0.4078, + "step": 208 + }, + { + "epoch": 0.1503326739795001, + "grad_norm": 12.414073459052013, + "learning_rate": 2.999024541866395e-06, + "loss": 0.4306, + "step": 209 + }, + { + "epoch": 0.1510519690703111, + "grad_norm": 12.841986715041784, + "learning_rate": 2.9990122347191857e-06, + "loss": 0.5474, + "step": 210 + }, + { + "epoch": 0.1517712641611221, + "grad_norm": 15.865072441413076, + "learning_rate": 2.998999850445166e-06, + "loss": 0.6193, + "step": 211 + }, + { + "epoch": 0.1524905592519331, + "grad_norm": 11.45279052450105, + "learning_rate": 2.9989873890449723e-06, + "loss": 0.4464, + "step": 212 + }, + { + "epoch": 0.15320985434274412, + "grad_norm": 9.090892566687161, + "learning_rate": 2.998974850519246e-06, + "loss": 0.4324, + "step": 213 + }, + { + "epoch": 0.15392914943355512, + "grad_norm": 9.674584282140811, + "learning_rate": 2.9989622348686327e-06, + "loss": 0.4122, + "step": 214 + }, + { + "epoch": 0.15464844452436613, + "grad_norm": 6.684517331123412, + "learning_rate": 2.998949542093781e-06, + "loss": 0.2104, + "step": 215 + }, + { + "epoch": 0.15536773961517714, + "grad_norm": 14.33919684371673, + "learning_rate": 2.9989367721953438e-06, + "loss": 0.2846, + "step": 216 + }, + { + "epoch": 0.15608703470598814, + "grad_norm": 13.59688480883437, + "learning_rate": 2.9989239251739785e-06, + "loss": 0.5451, + "step": 217 + }, + { + "epoch": 0.15680632979679915, + "grad_norm": 15.17634264162047, + "learning_rate": 2.9989110010303464e-06, + "loss": 0.7904, + "step": 218 + }, + { + "epoch": 0.15752562488761016, + "grad_norm": 10.835490596827958, + "learning_rate": 2.998897999765111e-06, + "loss": 0.444, + "step": 219 + }, + { + "epoch": 0.15824491997842113, + "grad_norm": 11.121993133593538, + "learning_rate": 2.998884921378943e-06, + "loss": 0.3226, + "step": 220 + }, + { + "epoch": 0.15896421506923214, + "grad_norm": 12.4576180308362, + "learning_rate": 2.9988717658725144e-06, + "loss": 0.3672, + "step": 221 + }, + { + "epoch": 0.15968351016004315, + "grad_norm": 11.500297093293284, + "learning_rate": 2.9988585332465027e-06, + "loss": 0.3059, + "step": 222 + }, + { + "epoch": 0.16040280525085415, + "grad_norm": 11.692934284894786, + "learning_rate": 2.9988452235015877e-06, + "loss": 0.3584, + "step": 223 + }, + { + "epoch": 0.16112210034166516, + "grad_norm": 10.92625880939289, + "learning_rate": 2.998831836638455e-06, + "loss": 0.1749, + "step": 224 + }, + { + "epoch": 0.16184139543247616, + "grad_norm": 12.101401854465001, + "learning_rate": 2.9988183726577932e-06, + "loss": 0.2852, + "step": 225 + }, + { + "epoch": 0.16256069052328717, + "grad_norm": 10.696809101144598, + "learning_rate": 2.998804831560295e-06, + "loss": 0.3363, + "step": 226 + }, + { + "epoch": 0.16327998561409818, + "grad_norm": 13.259836928252481, + "learning_rate": 2.9987912133466573e-06, + "loss": 0.5374, + "step": 227 + }, + { + "epoch": 0.16399928070490918, + "grad_norm": 15.045405140587834, + "learning_rate": 2.9987775180175806e-06, + "loss": 0.4377, + "step": 228 + }, + { + "epoch": 0.1647185757957202, + "grad_norm": 5.533706659541429, + "learning_rate": 2.998763745573769e-06, + "loss": 0.096, + "step": 229 + }, + { + "epoch": 0.1654378708865312, + "grad_norm": 7.837804068147167, + "learning_rate": 2.9987498960159326e-06, + "loss": 0.16, + "step": 230 + }, + { + "epoch": 0.1661571659773422, + "grad_norm": 12.514268459194259, + "learning_rate": 2.9987359693447828e-06, + "loss": 0.4397, + "step": 231 + }, + { + "epoch": 0.1668764610681532, + "grad_norm": 8.16973774034906, + "learning_rate": 2.9987219655610362e-06, + "loss": 0.2041, + "step": 232 + }, + { + "epoch": 0.16759575615896422, + "grad_norm": 11.796620785675312, + "learning_rate": 2.9987078846654134e-06, + "loss": 0.5258, + "step": 233 + }, + { + "epoch": 0.16831505124977522, + "grad_norm": 9.1366011301431, + "learning_rate": 2.9986937266586394e-06, + "loss": 0.4041, + "step": 234 + }, + { + "epoch": 0.16903434634058623, + "grad_norm": 7.319660588016992, + "learning_rate": 2.998679491541442e-06, + "loss": 0.2205, + "step": 235 + }, + { + "epoch": 0.16975364143139723, + "grad_norm": 9.663758952207552, + "learning_rate": 2.9986651793145543e-06, + "loss": 0.2728, + "step": 236 + }, + { + "epoch": 0.17047293652220824, + "grad_norm": 9.432959955727869, + "learning_rate": 2.998650789978712e-06, + "loss": 0.2088, + "step": 237 + }, + { + "epoch": 0.17119223161301925, + "grad_norm": 14.911613342507426, + "learning_rate": 2.998636323534656e-06, + "loss": 0.3816, + "step": 238 + }, + { + "epoch": 0.17191152670383025, + "grad_norm": 9.91889765561175, + "learning_rate": 2.99862177998313e-06, + "loss": 0.485, + "step": 239 + }, + { + "epoch": 0.17263082179464126, + "grad_norm": 7.970415448239636, + "learning_rate": 2.998607159324883e-06, + "loss": 0.1747, + "step": 240 + }, + { + "epoch": 0.17335011688545227, + "grad_norm": 10.490791513990272, + "learning_rate": 2.998592461560667e-06, + "loss": 0.3607, + "step": 241 + }, + { + "epoch": 0.17406941197626327, + "grad_norm": 12.008227473172994, + "learning_rate": 2.998577686691238e-06, + "loss": 0.2864, + "step": 242 + }, + { + "epoch": 0.17478870706707428, + "grad_norm": 11.469251123315106, + "learning_rate": 2.998562834717356e-06, + "loss": 0.5207, + "step": 243 + }, + { + "epoch": 0.17550800215788528, + "grad_norm": 8.869318644710406, + "learning_rate": 2.9985479056397858e-06, + "loss": 0.267, + "step": 244 + }, + { + "epoch": 0.1762272972486963, + "grad_norm": 10.837545663504354, + "learning_rate": 2.9985328994592958e-06, + "loss": 0.3718, + "step": 245 + }, + { + "epoch": 0.17694659233950727, + "grad_norm": 9.438194242279943, + "learning_rate": 2.9985178161766568e-06, + "loss": 0.2437, + "step": 246 + }, + { + "epoch": 0.17766588743031828, + "grad_norm": 14.817412116589152, + "learning_rate": 2.9985026557926456e-06, + "loss": 0.5897, + "step": 247 + }, + { + "epoch": 0.17838518252112928, + "grad_norm": 9.465656960834561, + "learning_rate": 2.9984874183080427e-06, + "loss": 0.2346, + "step": 248 + }, + { + "epoch": 0.1791044776119403, + "grad_norm": 13.316057832497831, + "learning_rate": 2.998472103723631e-06, + "loss": 0.4892, + "step": 249 + }, + { + "epoch": 0.1798237727027513, + "grad_norm": 13.335893338717915, + "learning_rate": 2.9984567120401996e-06, + "loss": 0.5034, + "step": 250 + }, + { + "epoch": 0.1805430677935623, + "grad_norm": 12.60598126805529, + "learning_rate": 2.99844124325854e-06, + "loss": 0.4103, + "step": 251 + }, + { + "epoch": 0.1812623628843733, + "grad_norm": 11.39934372716221, + "learning_rate": 2.9984256973794474e-06, + "loss": 0.4216, + "step": 252 + }, + { + "epoch": 0.1819816579751843, + "grad_norm": 12.896340882610827, + "learning_rate": 2.998410074403723e-06, + "loss": 0.571, + "step": 253 + }, + { + "epoch": 0.18270095306599532, + "grad_norm": 7.954801231901637, + "learning_rate": 2.998394374332169e-06, + "loss": 0.2671, + "step": 254 + }, + { + "epoch": 0.18342024815680633, + "grad_norm": 13.453184658810134, + "learning_rate": 2.9983785971655947e-06, + "loss": 0.4554, + "step": 255 + }, + { + "epoch": 0.18413954324761733, + "grad_norm": 8.15725989141266, + "learning_rate": 2.9983627429048113e-06, + "loss": 0.2668, + "step": 256 + }, + { + "epoch": 0.18485883833842834, + "grad_norm": 12.551962320728792, + "learning_rate": 2.998346811550634e-06, + "loss": 0.2095, + "step": 257 + }, + { + "epoch": 0.18557813342923934, + "grad_norm": 17.30908229697925, + "learning_rate": 2.9983308031038837e-06, + "loss": 0.5998, + "step": 258 + }, + { + "epoch": 0.18629742852005035, + "grad_norm": 11.557334489143301, + "learning_rate": 2.9983147175653827e-06, + "loss": 0.3596, + "step": 259 + }, + { + "epoch": 0.18701672361086136, + "grad_norm": 8.856220867626062, + "learning_rate": 2.9982985549359598e-06, + "loss": 0.2382, + "step": 260 + }, + { + "epoch": 0.18773601870167236, + "grad_norm": 12.347054297084657, + "learning_rate": 2.9982823152164458e-06, + "loss": 0.3965, + "step": 261 + }, + { + "epoch": 0.18845531379248337, + "grad_norm": 7.181096905871709, + "learning_rate": 2.9982659984076764e-06, + "loss": 0.1093, + "step": 262 + }, + { + "epoch": 0.18917460888329438, + "grad_norm": 13.246743112939686, + "learning_rate": 2.9982496045104914e-06, + "loss": 0.5906, + "step": 263 + }, + { + "epoch": 0.18989390397410538, + "grad_norm": 9.777738264141377, + "learning_rate": 2.998233133525734e-06, + "loss": 0.3573, + "step": 264 + }, + { + "epoch": 0.1906131990649164, + "grad_norm": 8.247457435694521, + "learning_rate": 2.998216585454252e-06, + "loss": 0.2562, + "step": 265 + }, + { + "epoch": 0.1913324941557274, + "grad_norm": 13.976344251379821, + "learning_rate": 2.9981999602968967e-06, + "loss": 0.3708, + "step": 266 + }, + { + "epoch": 0.1920517892465384, + "grad_norm": 13.163142692336633, + "learning_rate": 2.9981832580545235e-06, + "loss": 0.4492, + "step": 267 + }, + { + "epoch": 0.1927710843373494, + "grad_norm": 13.001953400210564, + "learning_rate": 2.9981664787279915e-06, + "loss": 0.4655, + "step": 268 + }, + { + "epoch": 0.19349037942816041, + "grad_norm": 13.842506368966323, + "learning_rate": 2.998149622318164e-06, + "loss": 0.3648, + "step": 269 + }, + { + "epoch": 0.19420967451897142, + "grad_norm": 10.906398363882099, + "learning_rate": 2.998132688825909e-06, + "loss": 0.3095, + "step": 270 + }, + { + "epoch": 0.19492896960978243, + "grad_norm": 13.87216090717685, + "learning_rate": 2.998115678252097e-06, + "loss": 0.375, + "step": 271 + }, + { + "epoch": 0.1956482647005934, + "grad_norm": 7.815467917214922, + "learning_rate": 2.9980985905976037e-06, + "loss": 0.2633, + "step": 272 + }, + { + "epoch": 0.1963675597914044, + "grad_norm": 11.856422063191868, + "learning_rate": 2.9980814258633074e-06, + "loss": 0.2885, + "step": 273 + }, + { + "epoch": 0.19708685488221542, + "grad_norm": 10.880015187107556, + "learning_rate": 2.9980641840500925e-06, + "loss": 0.4076, + "step": 274 + }, + { + "epoch": 0.19780614997302642, + "grad_norm": 14.201705034672239, + "learning_rate": 2.9980468651588458e-06, + "loss": 0.5385, + "step": 275 + }, + { + "epoch": 0.19852544506383743, + "grad_norm": 15.601881547790963, + "learning_rate": 2.9980294691904575e-06, + "loss": 0.388, + "step": 276 + }, + { + "epoch": 0.19924474015464844, + "grad_norm": 15.390198877632194, + "learning_rate": 2.9980119961458237e-06, + "loss": 0.2607, + "step": 277 + }, + { + "epoch": 0.19996403524545944, + "grad_norm": 14.371856391914728, + "learning_rate": 2.997994446025843e-06, + "loss": 0.4136, + "step": 278 + }, + { + "epoch": 0.20068333033627045, + "grad_norm": 10.420784489540349, + "learning_rate": 2.9979768188314182e-06, + "loss": 0.2827, + "step": 279 + }, + { + "epoch": 0.20140262542708146, + "grad_norm": 13.104961484102828, + "learning_rate": 2.997959114563457e-06, + "loss": 0.6195, + "step": 280 + }, + { + "epoch": 0.20212192051789246, + "grad_norm": 10.38392710315252, + "learning_rate": 2.9979413332228692e-06, + "loss": 0.3119, + "step": 281 + }, + { + "epoch": 0.20284121560870347, + "grad_norm": 6.14429188712158, + "learning_rate": 2.9979234748105707e-06, + "loss": 0.2586, + "step": 282 + }, + { + "epoch": 0.20356051069951447, + "grad_norm": 6.487760743841203, + "learning_rate": 2.99790553932748e-06, + "loss": 0.182, + "step": 283 + }, + { + "epoch": 0.20427980579032548, + "grad_norm": 11.743271681054466, + "learning_rate": 2.9978875267745193e-06, + "loss": 0.4323, + "step": 284 + }, + { + "epoch": 0.2049991008811365, + "grad_norm": 9.956722592479277, + "learning_rate": 2.997869437152616e-06, + "loss": 0.161, + "step": 285 + }, + { + "epoch": 0.2057183959719475, + "grad_norm": 10.740249919417174, + "learning_rate": 2.997851270462701e-06, + "loss": 0.2286, + "step": 286 + }, + { + "epoch": 0.2064376910627585, + "grad_norm": 10.504140656380809, + "learning_rate": 2.997833026705708e-06, + "loss": 0.2957, + "step": 287 + }, + { + "epoch": 0.2071569861535695, + "grad_norm": 8.5000954682574, + "learning_rate": 2.997814705882577e-06, + "loss": 0.1918, + "step": 288 + }, + { + "epoch": 0.2078762812443805, + "grad_norm": 10.098724047024069, + "learning_rate": 2.99779630799425e-06, + "loss": 0.1764, + "step": 289 + }, + { + "epoch": 0.20859557633519152, + "grad_norm": 10.857302085391463, + "learning_rate": 2.997777833041674e-06, + "loss": 0.2413, + "step": 290 + }, + { + "epoch": 0.20931487142600252, + "grad_norm": 8.268792145880168, + "learning_rate": 2.9977592810257984e-06, + "loss": 0.1949, + "step": 291 + }, + { + "epoch": 0.21003416651681353, + "grad_norm": 13.155447855701645, + "learning_rate": 2.997740651947579e-06, + "loss": 0.3198, + "step": 292 + }, + { + "epoch": 0.21075346160762454, + "grad_norm": 10.049044718287215, + "learning_rate": 2.997721945807974e-06, + "loss": 0.3267, + "step": 293 + }, + { + "epoch": 0.21147275669843554, + "grad_norm": 10.560958792742182, + "learning_rate": 2.997703162607946e-06, + "loss": 0.2936, + "step": 294 + }, + { + "epoch": 0.21219205178924655, + "grad_norm": 9.690052481092012, + "learning_rate": 2.9976843023484606e-06, + "loss": 0.2943, + "step": 295 + }, + { + "epoch": 0.21291134688005756, + "grad_norm": 11.5968414234468, + "learning_rate": 2.997665365030489e-06, + "loss": 0.4533, + "step": 296 + }, + { + "epoch": 0.21363064197086856, + "grad_norm": 8.65941356334403, + "learning_rate": 2.997646350655005e-06, + "loss": 0.34, + "step": 297 + }, + { + "epoch": 0.21434993706167954, + "grad_norm": 14.093680848748951, + "learning_rate": 2.9976272592229876e-06, + "loss": 0.3068, + "step": 298 + }, + { + "epoch": 0.21506923215249055, + "grad_norm": 10.73945937673034, + "learning_rate": 2.997608090735419e-06, + "loss": 0.2158, + "step": 299 + }, + { + "epoch": 0.21578852724330155, + "grad_norm": 10.319541571696844, + "learning_rate": 2.997588845193284e-06, + "loss": 0.529, + "step": 300 + }, + { + "epoch": 0.21650782233411256, + "grad_norm": 8.049753202701346, + "learning_rate": 2.997569522597575e-06, + "loss": 0.2708, + "step": 301 + }, + { + "epoch": 0.21722711742492357, + "grad_norm": 10.15704158478656, + "learning_rate": 2.9975501229492847e-06, + "loss": 0.3963, + "step": 302 + }, + { + "epoch": 0.21794641251573457, + "grad_norm": 10.72051688835257, + "learning_rate": 2.997530646249412e-06, + "loss": 0.3836, + "step": 303 + }, + { + "epoch": 0.21866570760654558, + "grad_norm": 10.25860994558, + "learning_rate": 2.9975110924989584e-06, + "loss": 0.2646, + "step": 304 + }, + { + "epoch": 0.21938500269735658, + "grad_norm": 6.991686682325399, + "learning_rate": 2.997491461698931e-06, + "loss": 0.1934, + "step": 305 + }, + { + "epoch": 0.2201042977881676, + "grad_norm": 11.001826145243419, + "learning_rate": 2.9974717538503385e-06, + "loss": 0.1715, + "step": 306 + }, + { + "epoch": 0.2208235928789786, + "grad_norm": 12.894266580814731, + "learning_rate": 2.9974519689541953e-06, + "loss": 0.3524, + "step": 307 + }, + { + "epoch": 0.2215428879697896, + "grad_norm": 10.243557471185888, + "learning_rate": 2.9974321070115196e-06, + "loss": 0.3797, + "step": 308 + }, + { + "epoch": 0.2222621830606006, + "grad_norm": 11.622193991467736, + "learning_rate": 2.997412168023334e-06, + "loss": 0.2475, + "step": 309 + }, + { + "epoch": 0.22298147815141162, + "grad_norm": 13.620762783141206, + "learning_rate": 2.9973921519906634e-06, + "loss": 0.4691, + "step": 310 + }, + { + "epoch": 0.22370077324222262, + "grad_norm": 12.052922046994613, + "learning_rate": 2.997372058914538e-06, + "loss": 0.2727, + "step": 311 + }, + { + "epoch": 0.22442006833303363, + "grad_norm": 16.174380113680808, + "learning_rate": 2.997351888795992e-06, + "loss": 0.6491, + "step": 312 + }, + { + "epoch": 0.22513936342384464, + "grad_norm": 9.63241089464826, + "learning_rate": 2.9973316416360624e-06, + "loss": 0.2147, + "step": 313 + }, + { + "epoch": 0.22585865851465564, + "grad_norm": 17.522132794774755, + "learning_rate": 2.9973113174357908e-06, + "loss": 0.3858, + "step": 314 + }, + { + "epoch": 0.22657795360546665, + "grad_norm": 8.973491779012276, + "learning_rate": 2.9972909161962247e-06, + "loss": 0.1409, + "step": 315 + }, + { + "epoch": 0.22729724869627765, + "grad_norm": 8.243014165895838, + "learning_rate": 2.997270437918412e-06, + "loss": 0.2116, + "step": 316 + }, + { + "epoch": 0.22801654378708866, + "grad_norm": 7.407220332696786, + "learning_rate": 2.997249882603406e-06, + "loss": 0.1531, + "step": 317 + }, + { + "epoch": 0.22873583887789967, + "grad_norm": 9.18431680608267, + "learning_rate": 2.997229250252266e-06, + "loss": 0.3942, + "step": 318 + }, + { + "epoch": 0.22945513396871067, + "grad_norm": 10.355015508976939, + "learning_rate": 2.997208540866053e-06, + "loss": 0.2808, + "step": 319 + }, + { + "epoch": 0.23017442905952168, + "grad_norm": 15.603005344421826, + "learning_rate": 2.9971877544458325e-06, + "loss": 0.4613, + "step": 320 + }, + { + "epoch": 0.23089372415033269, + "grad_norm": 8.679251032096321, + "learning_rate": 2.9971668909926733e-06, + "loss": 0.322, + "step": 321 + }, + { + "epoch": 0.2316130192411437, + "grad_norm": 8.052456370936467, + "learning_rate": 2.9971459505076494e-06, + "loss": 0.2745, + "step": 322 + }, + { + "epoch": 0.2323323143319547, + "grad_norm": 8.42723801671621, + "learning_rate": 2.997124932991838e-06, + "loss": 0.1704, + "step": 323 + }, + { + "epoch": 0.23305160942276568, + "grad_norm": 9.695100590141454, + "learning_rate": 2.997103838446321e-06, + "loss": 0.167, + "step": 324 + }, + { + "epoch": 0.23377090451357668, + "grad_norm": 9.102937950822987, + "learning_rate": 2.9970826668721836e-06, + "loss": 0.28, + "step": 325 + }, + { + "epoch": 0.2344901996043877, + "grad_norm": 8.471200978402717, + "learning_rate": 2.9970614182705145e-06, + "loss": 0.2106, + "step": 326 + }, + { + "epoch": 0.2352094946951987, + "grad_norm": 11.863885987586409, + "learning_rate": 2.9970400926424076e-06, + "loss": 0.4534, + "step": 327 + }, + { + "epoch": 0.2359287897860097, + "grad_norm": 10.935872905596742, + "learning_rate": 2.9970186899889594e-06, + "loss": 0.3792, + "step": 328 + }, + { + "epoch": 0.2366480848768207, + "grad_norm": 11.52976338678996, + "learning_rate": 2.9969972103112724e-06, + "loss": 0.1784, + "step": 329 + }, + { + "epoch": 0.23736737996763171, + "grad_norm": 9.062366806315843, + "learning_rate": 2.996975653610451e-06, + "loss": 0.2023, + "step": 330 + }, + { + "epoch": 0.23808667505844272, + "grad_norm": 9.407895213960451, + "learning_rate": 2.996954019887604e-06, + "loss": 0.3943, + "step": 331 + }, + { + "epoch": 0.23880597014925373, + "grad_norm": 15.337835908445518, + "learning_rate": 2.9969323091438455e-06, + "loss": 0.5976, + "step": 332 + }, + { + "epoch": 0.23952526524006473, + "grad_norm": 12.85595473580567, + "learning_rate": 2.9969105213802915e-06, + "loss": 0.4178, + "step": 333 + }, + { + "epoch": 0.24024456033087574, + "grad_norm": 9.00368307970452, + "learning_rate": 2.9968886565980633e-06, + "loss": 0.1701, + "step": 334 + }, + { + "epoch": 0.24096385542168675, + "grad_norm": 12.983566330146337, + "learning_rate": 2.9968667147982864e-06, + "loss": 0.4775, + "step": 335 + }, + { + "epoch": 0.24168315051249775, + "grad_norm": 12.490135154595928, + "learning_rate": 2.996844695982089e-06, + "loss": 0.391, + "step": 336 + }, + { + "epoch": 0.24240244560330876, + "grad_norm": 10.662630389549339, + "learning_rate": 2.9968226001506042e-06, + "loss": 0.3546, + "step": 337 + }, + { + "epoch": 0.24312174069411976, + "grad_norm": 8.438847678165793, + "learning_rate": 2.9968004273049694e-06, + "loss": 0.2592, + "step": 338 + }, + { + "epoch": 0.24384103578493077, + "grad_norm": 9.506620260928214, + "learning_rate": 2.996778177446325e-06, + "loss": 0.2267, + "step": 339 + }, + { + "epoch": 0.24456033087574178, + "grad_norm": 10.231121716210279, + "learning_rate": 2.996755850575816e-06, + "loss": 0.3227, + "step": 340 + }, + { + "epoch": 0.24527962596655278, + "grad_norm": 7.227690713001611, + "learning_rate": 2.9967334466945903e-06, + "loss": 0.2186, + "step": 341 + }, + { + "epoch": 0.2459989210573638, + "grad_norm": 9.57947279862929, + "learning_rate": 2.996710965803802e-06, + "loss": 0.2507, + "step": 342 + }, + { + "epoch": 0.2467182161481748, + "grad_norm": 9.371660195527829, + "learning_rate": 2.996688407904607e-06, + "loss": 0.3166, + "step": 343 + }, + { + "epoch": 0.2474375112389858, + "grad_norm": 13.460697955358041, + "learning_rate": 2.9966657729981665e-06, + "loss": 0.233, + "step": 344 + }, + { + "epoch": 0.2481568063297968, + "grad_norm": 9.83124222393436, + "learning_rate": 2.9966430610856436e-06, + "loss": 0.2656, + "step": 345 + }, + { + "epoch": 0.24887610142060781, + "grad_norm": 10.010909354391735, + "learning_rate": 2.996620272168209e-06, + "loss": 0.4054, + "step": 346 + }, + { + "epoch": 0.24959539651141882, + "grad_norm": 4.824021065095197, + "learning_rate": 2.9965974062470333e-06, + "loss": 0.1997, + "step": 347 + }, + { + "epoch": 0.2503146916022298, + "grad_norm": 8.005322435112516, + "learning_rate": 2.9965744633232947e-06, + "loss": 0.2646, + "step": 348 + }, + { + "epoch": 0.25103398669304083, + "grad_norm": 7.343884721132704, + "learning_rate": 2.996551443398172e-06, + "loss": 0.1738, + "step": 349 + }, + { + "epoch": 0.2517532817838518, + "grad_norm": 8.360177818344797, + "learning_rate": 2.9965283464728508e-06, + "loss": 0.2571, + "step": 350 + }, + { + "epoch": 0.25247257687466285, + "grad_norm": 6.792580794942874, + "learning_rate": 2.996505172548519e-06, + "loss": 0.0952, + "step": 351 + }, + { + "epoch": 0.2531918719654738, + "grad_norm": 10.696204710229608, + "learning_rate": 2.9964819216263694e-06, + "loss": 0.3471, + "step": 352 + }, + { + "epoch": 0.25391116705628486, + "grad_norm": 9.32400869528146, + "learning_rate": 2.996458593707598e-06, + "loss": 0.1938, + "step": 353 + }, + { + "epoch": 0.25463046214709584, + "grad_norm": 6.526513107235073, + "learning_rate": 2.9964351887934043e-06, + "loss": 0.1676, + "step": 354 + }, + { + "epoch": 0.25534975723790687, + "grad_norm": 8.338472510715075, + "learning_rate": 2.9964117068849937e-06, + "loss": 0.1876, + "step": 355 + }, + { + "epoch": 0.25606905232871785, + "grad_norm": 9.621677975595418, + "learning_rate": 2.996388147983574e-06, + "loss": 0.2056, + "step": 356 + }, + { + "epoch": 0.2567883474195289, + "grad_norm": 12.484588912806705, + "learning_rate": 2.996364512090357e-06, + "loss": 0.2654, + "step": 357 + }, + { + "epoch": 0.25750764251033986, + "grad_norm": 8.106593765931988, + "learning_rate": 2.996340799206559e-06, + "loss": 0.2597, + "step": 358 + }, + { + "epoch": 0.2582269376011509, + "grad_norm": 14.282226452989242, + "learning_rate": 2.9963170093334004e-06, + "loss": 0.3816, + "step": 359 + }, + { + "epoch": 0.2589462326919619, + "grad_norm": 8.368840076338213, + "learning_rate": 2.996293142472105e-06, + "loss": 0.1286, + "step": 360 + }, + { + "epoch": 0.2596655277827729, + "grad_norm": 6.696843858149411, + "learning_rate": 2.9962691986239e-06, + "loss": 0.1924, + "step": 361 + }, + { + "epoch": 0.2603848228735839, + "grad_norm": 7.962899969224421, + "learning_rate": 2.9962451777900192e-06, + "loss": 0.1765, + "step": 362 + }, + { + "epoch": 0.26110411796439487, + "grad_norm": 17.73875889449904, + "learning_rate": 2.996221079971697e-06, + "loss": 0.6514, + "step": 363 + }, + { + "epoch": 0.2618234130552059, + "grad_norm": 10.088957533953707, + "learning_rate": 2.9961969051701735e-06, + "loss": 0.1896, + "step": 364 + }, + { + "epoch": 0.2625427081460169, + "grad_norm": 12.453133378487488, + "learning_rate": 2.9961726533866926e-06, + "loss": 0.3453, + "step": 365 + }, + { + "epoch": 0.2632620032368279, + "grad_norm": 11.899756096310341, + "learning_rate": 2.9961483246225027e-06, + "loss": 0.1886, + "step": 366 + }, + { + "epoch": 0.2639812983276389, + "grad_norm": 11.495582861425552, + "learning_rate": 2.996123918878855e-06, + "loss": 0.2784, + "step": 367 + }, + { + "epoch": 0.2647005934184499, + "grad_norm": 12.54192687953523, + "learning_rate": 2.9960994361570052e-06, + "loss": 0.2389, + "step": 368 + }, + { + "epoch": 0.2654198885092609, + "grad_norm": 15.193888755066343, + "learning_rate": 2.9960748764582134e-06, + "loss": 0.4148, + "step": 369 + }, + { + "epoch": 0.26613918360007194, + "grad_norm": 9.503164320388725, + "learning_rate": 2.9960502397837422e-06, + "loss": 0.1989, + "step": 370 + }, + { + "epoch": 0.2668584786908829, + "grad_norm": 8.269561546615632, + "learning_rate": 2.9960255261348606e-06, + "loss": 0.2405, + "step": 371 + }, + { + "epoch": 0.26757777378169395, + "grad_norm": 12.020635644499855, + "learning_rate": 2.996000735512839e-06, + "loss": 0.2988, + "step": 372 + }, + { + "epoch": 0.26829706887250493, + "grad_norm": 8.985638658238557, + "learning_rate": 2.995975867918954e-06, + "loss": 0.1619, + "step": 373 + }, + { + "epoch": 0.26901636396331596, + "grad_norm": 7.935651727610059, + "learning_rate": 2.9959509233544843e-06, + "loss": 0.2772, + "step": 374 + }, + { + "epoch": 0.26973565905412694, + "grad_norm": 13.423066123686887, + "learning_rate": 2.9959259018207133e-06, + "loss": 0.3276, + "step": 375 + }, + { + "epoch": 0.270454954144938, + "grad_norm": 8.090818837580843, + "learning_rate": 2.9959008033189287e-06, + "loss": 0.0981, + "step": 376 + }, + { + "epoch": 0.27117424923574895, + "grad_norm": 10.917968964794552, + "learning_rate": 2.995875627850422e-06, + "loss": 0.3252, + "step": 377 + }, + { + "epoch": 0.27189354432656, + "grad_norm": 8.49562529169398, + "learning_rate": 2.9958503754164883e-06, + "loss": 0.2594, + "step": 378 + }, + { + "epoch": 0.27261283941737097, + "grad_norm": 8.652543862175486, + "learning_rate": 2.9958250460184265e-06, + "loss": 0.271, + "step": 379 + }, + { + "epoch": 0.273332134508182, + "grad_norm": 7.735114801134429, + "learning_rate": 2.9957996396575405e-06, + "loss": 0.2967, + "step": 380 + }, + { + "epoch": 0.274051429598993, + "grad_norm": 7.461929682508416, + "learning_rate": 2.995774156335137e-06, + "loss": 0.1971, + "step": 381 + }, + { + "epoch": 0.274770724689804, + "grad_norm": 8.373943213507687, + "learning_rate": 2.995748596052528e-06, + "loss": 0.2388, + "step": 382 + }, + { + "epoch": 0.275490019780615, + "grad_norm": 7.692796650185384, + "learning_rate": 2.9957229588110277e-06, + "loss": 0.2689, + "step": 383 + }, + { + "epoch": 0.276209314871426, + "grad_norm": 11.5353347069809, + "learning_rate": 2.9956972446119556e-06, + "loss": 0.1924, + "step": 384 + }, + { + "epoch": 0.276928609962237, + "grad_norm": 9.236543467720582, + "learning_rate": 2.9956714534566343e-06, + "loss": 0.1729, + "step": 385 + }, + { + "epoch": 0.27764790505304804, + "grad_norm": 9.711051511926513, + "learning_rate": 2.9956455853463914e-06, + "loss": 0.2198, + "step": 386 + }, + { + "epoch": 0.278367200143859, + "grad_norm": 9.059058712001333, + "learning_rate": 2.9956196402825573e-06, + "loss": 0.3891, + "step": 387 + }, + { + "epoch": 0.27908649523467, + "grad_norm": 6.277351548744565, + "learning_rate": 2.9955936182664678e-06, + "loss": 0.1852, + "step": 388 + }, + { + "epoch": 0.27980579032548103, + "grad_norm": 7.242455927831628, + "learning_rate": 2.9955675192994615e-06, + "loss": 0.2527, + "step": 389 + }, + { + "epoch": 0.280525085416292, + "grad_norm": 10.367919504041973, + "learning_rate": 2.99554134338288e-06, + "loss": 0.2319, + "step": 390 + }, + { + "epoch": 0.28124438050710304, + "grad_norm": 8.790781759048409, + "learning_rate": 2.995515090518072e-06, + "loss": 0.2425, + "step": 391 + }, + { + "epoch": 0.281963675597914, + "grad_norm": 7.180223846706552, + "learning_rate": 2.995488760706386e-06, + "loss": 0.228, + "step": 392 + }, + { + "epoch": 0.28268297068872505, + "grad_norm": 6.964087388445663, + "learning_rate": 2.9954623539491793e-06, + "loss": 0.165, + "step": 393 + }, + { + "epoch": 0.28340226577953603, + "grad_norm": 10.702327770754659, + "learning_rate": 2.9954358702478084e-06, + "loss": 0.2385, + "step": 394 + }, + { + "epoch": 0.28412156087034707, + "grad_norm": 9.082516119263104, + "learning_rate": 2.995409309603637e-06, + "loss": 0.3171, + "step": 395 + }, + { + "epoch": 0.28484085596115805, + "grad_norm": 7.707621376635846, + "learning_rate": 2.995382672018032e-06, + "loss": 0.1612, + "step": 396 + }, + { + "epoch": 0.2855601510519691, + "grad_norm": 6.014506748872444, + "learning_rate": 2.995355957492363e-06, + "loss": 0.2554, + "step": 397 + }, + { + "epoch": 0.28627944614278006, + "grad_norm": 7.564226651367394, + "learning_rate": 2.9953291660280055e-06, + "loss": 0.2188, + "step": 398 + }, + { + "epoch": 0.2869987412335911, + "grad_norm": 9.981279147419581, + "learning_rate": 2.995302297626337e-06, + "loss": 0.2318, + "step": 399 + }, + { + "epoch": 0.28771803632440207, + "grad_norm": 8.532486186989155, + "learning_rate": 2.99527535228874e-06, + "loss": 0.3529, + "step": 400 + }, + { + "epoch": 0.2884373314152131, + "grad_norm": 7.82831385381344, + "learning_rate": 2.995248330016602e-06, + "loss": 0.2805, + "step": 401 + }, + { + "epoch": 0.2891566265060241, + "grad_norm": 8.805743130071244, + "learning_rate": 2.995221230811312e-06, + "loss": 0.2915, + "step": 402 + }, + { + "epoch": 0.2898759215968351, + "grad_norm": 9.094733511915015, + "learning_rate": 2.9951940546742653e-06, + "loss": 0.1482, + "step": 403 + }, + { + "epoch": 0.2905952166876461, + "grad_norm": 8.207026219832166, + "learning_rate": 2.9951668016068596e-06, + "loss": 0.1724, + "step": 404 + }, + { + "epoch": 0.29131451177845713, + "grad_norm": 6.867471145862766, + "learning_rate": 2.9951394716104974e-06, + "loss": 0.2305, + "step": 405 + }, + { + "epoch": 0.2920338068692681, + "grad_norm": 9.204708353903655, + "learning_rate": 2.9951120646865842e-06, + "loss": 0.286, + "step": 406 + }, + { + "epoch": 0.29275310196007914, + "grad_norm": 11.684321050750068, + "learning_rate": 2.9950845808365317e-06, + "loss": 0.2599, + "step": 407 + }, + { + "epoch": 0.2934723970508901, + "grad_norm": 12.972548620267979, + "learning_rate": 2.9950570200617524e-06, + "loss": 0.2231, + "step": 408 + }, + { + "epoch": 0.29419169214170116, + "grad_norm": 8.150918145053511, + "learning_rate": 2.995029382363665e-06, + "loss": 0.1222, + "step": 409 + }, + { + "epoch": 0.29491098723251213, + "grad_norm": 13.31114508219458, + "learning_rate": 2.995001667743691e-06, + "loss": 0.2331, + "step": 410 + }, + { + "epoch": 0.29563028232332317, + "grad_norm": 10.915018744772286, + "learning_rate": 2.9949738762032576e-06, + "loss": 0.2106, + "step": 411 + }, + { + "epoch": 0.29634957741413415, + "grad_norm": 11.212739819920543, + "learning_rate": 2.9949460077437932e-06, + "loss": 0.1485, + "step": 412 + }, + { + "epoch": 0.2970688725049452, + "grad_norm": 15.257880903781947, + "learning_rate": 2.994918062366733e-06, + "loss": 0.5924, + "step": 413 + }, + { + "epoch": 0.29778816759575616, + "grad_norm": 8.772076814667747, + "learning_rate": 2.9948900400735138e-06, + "loss": 0.2593, + "step": 414 + }, + { + "epoch": 0.29850746268656714, + "grad_norm": 9.238925942706313, + "learning_rate": 2.994861940865578e-06, + "loss": 0.2807, + "step": 415 + }, + { + "epoch": 0.29922675777737817, + "grad_norm": 12.327731313364321, + "learning_rate": 2.994833764744371e-06, + "loss": 0.1232, + "step": 416 + }, + { + "epoch": 0.29994605286818915, + "grad_norm": 9.820439632000484, + "learning_rate": 2.994805511711343e-06, + "loss": 0.3351, + "step": 417 + }, + { + "epoch": 0.3006653479590002, + "grad_norm": 8.000892509445384, + "learning_rate": 2.9947771817679474e-06, + "loss": 0.266, + "step": 418 + }, + { + "epoch": 0.30138464304981116, + "grad_norm": 10.352406847146282, + "learning_rate": 2.9947487749156418e-06, + "loss": 0.1777, + "step": 419 + }, + { + "epoch": 0.3021039381406222, + "grad_norm": 7.480424892887567, + "learning_rate": 2.9947202911558874e-06, + "loss": 0.2517, + "step": 420 + }, + { + "epoch": 0.3028232332314332, + "grad_norm": 6.623389444191224, + "learning_rate": 2.9946917304901506e-06, + "loss": 0.2247, + "step": 421 + }, + { + "epoch": 0.3035425283222442, + "grad_norm": 7.417147665562637, + "learning_rate": 2.9946630929199e-06, + "loss": 0.2303, + "step": 422 + }, + { + "epoch": 0.3042618234130552, + "grad_norm": 8.923072775228164, + "learning_rate": 2.9946343784466096e-06, + "loss": 0.226, + "step": 423 + }, + { + "epoch": 0.3049811185038662, + "grad_norm": 10.476843191096348, + "learning_rate": 2.9946055870717565e-06, + "loss": 0.4008, + "step": 424 + }, + { + "epoch": 0.3057004135946772, + "grad_norm": 8.703951018313962, + "learning_rate": 2.9945767187968227e-06, + "loss": 0.2258, + "step": 425 + }, + { + "epoch": 0.30641970868548823, + "grad_norm": 10.476161352254241, + "learning_rate": 2.994547773623293e-06, + "loss": 0.4341, + "step": 426 + }, + { + "epoch": 0.3071390037762992, + "grad_norm": 8.696807780643637, + "learning_rate": 2.9945187515526566e-06, + "loss": 0.287, + "step": 427 + }, + { + "epoch": 0.30785829886711025, + "grad_norm": 10.68503471179048, + "learning_rate": 2.994489652586407e-06, + "loss": 0.3893, + "step": 428 + }, + { + "epoch": 0.3085775939579212, + "grad_norm": 11.173799729660958, + "learning_rate": 2.994460476726041e-06, + "loss": 0.2456, + "step": 429 + }, + { + "epoch": 0.30929688904873226, + "grad_norm": 7.304502575495571, + "learning_rate": 2.9944312239730604e-06, + "loss": 0.258, + "step": 430 + }, + { + "epoch": 0.31001618413954324, + "grad_norm": 9.2050161333308, + "learning_rate": 2.9944018943289698e-06, + "loss": 0.1753, + "step": 431 + }, + { + "epoch": 0.3107354792303543, + "grad_norm": 7.25244315285998, + "learning_rate": 2.9943724877952783e-06, + "loss": 0.1689, + "step": 432 + }, + { + "epoch": 0.31145477432116525, + "grad_norm": 8.57014411753549, + "learning_rate": 2.9943430043734988e-06, + "loss": 0.1171, + "step": 433 + }, + { + "epoch": 0.3121740694119763, + "grad_norm": 5.12479042738401, + "learning_rate": 2.994313444065149e-06, + "loss": 0.0794, + "step": 434 + }, + { + "epoch": 0.31289336450278726, + "grad_norm": 8.167122669177642, + "learning_rate": 2.994283806871749e-06, + "loss": 0.1969, + "step": 435 + }, + { + "epoch": 0.3136126595935983, + "grad_norm": 5.889526847070679, + "learning_rate": 2.994254092794824e-06, + "loss": 0.1907, + "step": 436 + }, + { + "epoch": 0.3143319546844093, + "grad_norm": 7.803611222302982, + "learning_rate": 2.994224301835903e-06, + "loss": 0.2208, + "step": 437 + }, + { + "epoch": 0.3150512497752203, + "grad_norm": 17.112049783414655, + "learning_rate": 2.9941944339965187e-06, + "loss": 0.1361, + "step": 438 + }, + { + "epoch": 0.3157705448660313, + "grad_norm": 12.968078824689863, + "learning_rate": 2.9941644892782074e-06, + "loss": 0.3567, + "step": 439 + }, + { + "epoch": 0.31648983995684227, + "grad_norm": 6.543873111288228, + "learning_rate": 2.9941344676825104e-06, + "loss": 0.0938, + "step": 440 + }, + { + "epoch": 0.3172091350476533, + "grad_norm": 11.561968819083491, + "learning_rate": 2.9941043692109725e-06, + "loss": 0.5349, + "step": 441 + }, + { + "epoch": 0.3179284301384643, + "grad_norm": 7.899621134761218, + "learning_rate": 2.994074193865142e-06, + "loss": 0.121, + "step": 442 + }, + { + "epoch": 0.3186477252292753, + "grad_norm": 12.740350221282768, + "learning_rate": 2.994043941646571e-06, + "loss": 0.2477, + "step": 443 + }, + { + "epoch": 0.3193670203200863, + "grad_norm": 5.348027947374008, + "learning_rate": 2.9940136125568165e-06, + "loss": 0.0354, + "step": 444 + }, + { + "epoch": 0.3200863154108973, + "grad_norm": 11.986539406760485, + "learning_rate": 2.993983206597439e-06, + "loss": 0.5919, + "step": 445 + }, + { + "epoch": 0.3208056105017083, + "grad_norm": 8.767824151738221, + "learning_rate": 2.993952723770003e-06, + "loss": 0.1385, + "step": 446 + }, + { + "epoch": 0.32152490559251934, + "grad_norm": 11.490044874098949, + "learning_rate": 2.9939221640760773e-06, + "loss": 0.2664, + "step": 447 + }, + { + "epoch": 0.3222442006833303, + "grad_norm": 6.473208496720477, + "learning_rate": 2.993891527517233e-06, + "loss": 0.1232, + "step": 448 + }, + { + "epoch": 0.32296349577414135, + "grad_norm": 8.402492007996052, + "learning_rate": 2.9938608140950475e-06, + "loss": 0.3161, + "step": 449 + }, + { + "epoch": 0.32368279086495233, + "grad_norm": 10.887662933472471, + "learning_rate": 2.9938300238111007e-06, + "loss": 0.374, + "step": 450 + }, + { + "epoch": 0.32440208595576336, + "grad_norm": 7.531560746533734, + "learning_rate": 2.9937991566669765e-06, + "loss": 0.1566, + "step": 451 + }, + { + "epoch": 0.32512138104657434, + "grad_norm": 8.045718514802468, + "learning_rate": 2.9937682126642643e-06, + "loss": 0.1335, + "step": 452 + }, + { + "epoch": 0.3258406761373854, + "grad_norm": 7.20557052846816, + "learning_rate": 2.993737191804555e-06, + "loss": 0.2782, + "step": 453 + }, + { + "epoch": 0.32655997122819636, + "grad_norm": 7.3314078440514105, + "learning_rate": 2.9937060940894447e-06, + "loss": 0.0879, + "step": 454 + }, + { + "epoch": 0.3272792663190074, + "grad_norm": 12.314161812184437, + "learning_rate": 2.9936749195205335e-06, + "loss": 0.2937, + "step": 455 + }, + { + "epoch": 0.32799856140981837, + "grad_norm": 10.013293912777787, + "learning_rate": 2.993643668099426e-06, + "loss": 0.2753, + "step": 456 + }, + { + "epoch": 0.3287178565006294, + "grad_norm": 9.215564999455935, + "learning_rate": 2.99361233982773e-06, + "loss": 0.1477, + "step": 457 + }, + { + "epoch": 0.3294371515914404, + "grad_norm": 11.464109463075435, + "learning_rate": 2.9935809347070566e-06, + "loss": 0.373, + "step": 458 + }, + { + "epoch": 0.3301564466822514, + "grad_norm": 7.98721130374862, + "learning_rate": 2.9935494527390227e-06, + "loss": 0.0831, + "step": 459 + }, + { + "epoch": 0.3308757417730624, + "grad_norm": 8.461868928839882, + "learning_rate": 2.9935178939252478e-06, + "loss": 0.0513, + "step": 460 + }, + { + "epoch": 0.3315950368638734, + "grad_norm": 11.237735546551782, + "learning_rate": 2.9934862582673547e-06, + "loss": 0.3597, + "step": 461 + }, + { + "epoch": 0.3323143319546844, + "grad_norm": 6.777642651412544, + "learning_rate": 2.9934545457669722e-06, + "loss": 0.091, + "step": 462 + }, + { + "epoch": 0.33303362704549544, + "grad_norm": 6.099996550715816, + "learning_rate": 2.993422756425732e-06, + "loss": 0.1597, + "step": 463 + }, + { + "epoch": 0.3337529221363064, + "grad_norm": 10.10252724634063, + "learning_rate": 2.9933908902452688e-06, + "loss": 0.2642, + "step": 464 + }, + { + "epoch": 0.33447221722711745, + "grad_norm": 8.73958452919609, + "learning_rate": 2.9933589472272228e-06, + "loss": 0.1605, + "step": 465 + }, + { + "epoch": 0.33519151231792843, + "grad_norm": 10.637020856187041, + "learning_rate": 2.993326927373238e-06, + "loss": 0.3334, + "step": 466 + }, + { + "epoch": 0.3359108074087394, + "grad_norm": 7.948066977617699, + "learning_rate": 2.9932948306849607e-06, + "loss": 0.3347, + "step": 467 + }, + { + "epoch": 0.33663010249955044, + "grad_norm": 11.7979073268195, + "learning_rate": 2.9932626571640433e-06, + "loss": 0.5206, + "step": 468 + }, + { + "epoch": 0.3373493975903614, + "grad_norm": 7.492798403225308, + "learning_rate": 2.993230406812141e-06, + "loss": 0.1728, + "step": 469 + }, + { + "epoch": 0.33806869268117246, + "grad_norm": 7.385286357099298, + "learning_rate": 2.9931980796309132e-06, + "loss": 0.1249, + "step": 470 + }, + { + "epoch": 0.33878798777198343, + "grad_norm": 7.8909238454313995, + "learning_rate": 2.993165675622022e-06, + "loss": 0.2917, + "step": 471 + }, + { + "epoch": 0.33950728286279447, + "grad_norm": 7.794771394916748, + "learning_rate": 2.9931331947871364e-06, + "loss": 0.2058, + "step": 472 + }, + { + "epoch": 0.34022657795360545, + "grad_norm": 7.267841128071171, + "learning_rate": 2.9931006371279264e-06, + "loss": 0.1779, + "step": 473 + }, + { + "epoch": 0.3409458730444165, + "grad_norm": 6.995480773011945, + "learning_rate": 2.993068002646068e-06, + "loss": 0.1657, + "step": 474 + }, + { + "epoch": 0.34166516813522746, + "grad_norm": 6.687070292531621, + "learning_rate": 2.99303529134324e-06, + "loss": 0.1984, + "step": 475 + }, + { + "epoch": 0.3423844632260385, + "grad_norm": 7.04320434425698, + "learning_rate": 2.9930025032211246e-06, + "loss": 0.1614, + "step": 476 + }, + { + "epoch": 0.34310375831684947, + "grad_norm": 7.962174451472046, + "learning_rate": 2.99296963828141e-06, + "loss": 0.1542, + "step": 477 + }, + { + "epoch": 0.3438230534076605, + "grad_norm": 7.506214315558402, + "learning_rate": 2.9929366965257867e-06, + "loss": 0.0898, + "step": 478 + }, + { + "epoch": 0.3445423484984715, + "grad_norm": 7.815488630923755, + "learning_rate": 2.9929036779559494e-06, + "loss": 0.2657, + "step": 479 + }, + { + "epoch": 0.3452616435892825, + "grad_norm": 12.177242440046614, + "learning_rate": 2.9928705825735976e-06, + "loss": 0.2595, + "step": 480 + }, + { + "epoch": 0.3459809386800935, + "grad_norm": 9.561239344410755, + "learning_rate": 2.992837410380433e-06, + "loss": 0.2468, + "step": 481 + }, + { + "epoch": 0.34670023377090453, + "grad_norm": 9.73125276390459, + "learning_rate": 2.992804161378163e-06, + "loss": 0.248, + "step": 482 + }, + { + "epoch": 0.3474195288617155, + "grad_norm": 10.663792090176129, + "learning_rate": 2.992770835568499e-06, + "loss": 0.4104, + "step": 483 + }, + { + "epoch": 0.34813882395252654, + "grad_norm": 6.349392040007562, + "learning_rate": 2.9927374329531547e-06, + "loss": 0.0644, + "step": 484 + }, + { + "epoch": 0.3488581190433375, + "grad_norm": 8.458321034729055, + "learning_rate": 2.9927039535338493e-06, + "loss": 0.2923, + "step": 485 + }, + { + "epoch": 0.34957741413414856, + "grad_norm": 5.340889505512948, + "learning_rate": 2.9926703973123047e-06, + "loss": 0.0805, + "step": 486 + }, + { + "epoch": 0.35029670922495953, + "grad_norm": 9.371768873596796, + "learning_rate": 2.9926367642902484e-06, + "loss": 0.242, + "step": 487 + }, + { + "epoch": 0.35101600431577057, + "grad_norm": 6.487285839408496, + "learning_rate": 2.9926030544694098e-06, + "loss": 0.0783, + "step": 488 + }, + { + "epoch": 0.35173529940658155, + "grad_norm": 7.616234299134105, + "learning_rate": 2.9925692678515243e-06, + "loss": 0.0961, + "step": 489 + }, + { + "epoch": 0.3524545944973926, + "grad_norm": 6.992847248226069, + "learning_rate": 2.9925354044383292e-06, + "loss": 0.1936, + "step": 490 + }, + { + "epoch": 0.35317388958820356, + "grad_norm": 9.653437422048437, + "learning_rate": 2.992501464231568e-06, + "loss": 0.1853, + "step": 491 + }, + { + "epoch": 0.35389318467901454, + "grad_norm": 6.1322948238121775, + "learning_rate": 2.992467447232986e-06, + "loss": 0.202, + "step": 492 + }, + { + "epoch": 0.3546124797698256, + "grad_norm": 4.817342730839772, + "learning_rate": 2.9924333534443345e-06, + "loss": 0.0877, + "step": 493 + }, + { + "epoch": 0.35533177486063655, + "grad_norm": 9.542291528465045, + "learning_rate": 2.9923991828673674e-06, + "loss": 0.345, + "step": 494 + }, + { + "epoch": 0.3560510699514476, + "grad_norm": 13.503074532301907, + "learning_rate": 2.9923649355038418e-06, + "loss": 0.0948, + "step": 495 + }, + { + "epoch": 0.35677036504225856, + "grad_norm": 8.40388429041765, + "learning_rate": 2.992330611355521e-06, + "loss": 0.3541, + "step": 496 + }, + { + "epoch": 0.3574896601330696, + "grad_norm": 10.966619584674941, + "learning_rate": 2.9922962104241697e-06, + "loss": 0.3177, + "step": 497 + }, + { + "epoch": 0.3582089552238806, + "grad_norm": 7.13962635679365, + "learning_rate": 2.9922617327115596e-06, + "loss": 0.3161, + "step": 498 + }, + { + "epoch": 0.3589282503146916, + "grad_norm": 6.188551855241201, + "learning_rate": 2.9922271782194633e-06, + "loss": 0.3086, + "step": 499 + }, + { + "epoch": 0.3596475454055026, + "grad_norm": 8.454800054539927, + "learning_rate": 2.9921925469496592e-06, + "loss": 0.2869, + "step": 500 + }, + { + "epoch": 0.3603668404963136, + "grad_norm": 7.162322257370967, + "learning_rate": 2.992157838903929e-06, + "loss": 0.2941, + "step": 501 + }, + { + "epoch": 0.3610861355871246, + "grad_norm": 11.809733697609971, + "learning_rate": 2.992123054084059e-06, + "loss": 0.2529, + "step": 502 + }, + { + "epoch": 0.36180543067793564, + "grad_norm": 11.192862994065177, + "learning_rate": 2.992088192491838e-06, + "loss": 0.292, + "step": 503 + }, + { + "epoch": 0.3625247257687466, + "grad_norm": 6.543959350421433, + "learning_rate": 2.992053254129061e-06, + "loss": 0.1429, + "step": 504 + }, + { + "epoch": 0.36324402085955765, + "grad_norm": 6.468307908696831, + "learning_rate": 2.992018238997524e-06, + "loss": 0.2569, + "step": 505 + }, + { + "epoch": 0.3639633159503686, + "grad_norm": 6.7613249894524, + "learning_rate": 2.99198314709903e-06, + "loss": 0.0466, + "step": 506 + }, + { + "epoch": 0.36468261104117966, + "grad_norm": 8.295124851483795, + "learning_rate": 2.9919479784353837e-06, + "loss": 0.1143, + "step": 507 + }, + { + "epoch": 0.36540190613199064, + "grad_norm": 7.524959168455001, + "learning_rate": 2.991912733008395e-06, + "loss": 0.2229, + "step": 508 + }, + { + "epoch": 0.3661212012228017, + "grad_norm": 6.718934638289708, + "learning_rate": 2.991877410819877e-06, + "loss": 0.2365, + "step": 509 + }, + { + "epoch": 0.36684049631361265, + "grad_norm": 13.189631741873894, + "learning_rate": 2.9918420118716476e-06, + "loss": 0.4204, + "step": 510 + }, + { + "epoch": 0.3675597914044237, + "grad_norm": 10.64610812505103, + "learning_rate": 2.991806536165528e-06, + "loss": 0.3337, + "step": 511 + }, + { + "epoch": 0.36827908649523466, + "grad_norm": 12.436063489280153, + "learning_rate": 2.991770983703343e-06, + "loss": 0.279, + "step": 512 + }, + { + "epoch": 0.3689983815860457, + "grad_norm": 8.553716095731936, + "learning_rate": 2.991735354486922e-06, + "loss": 0.2034, + "step": 513 + }, + { + "epoch": 0.3697176766768567, + "grad_norm": 10.794690904916838, + "learning_rate": 2.991699648518099e-06, + "loss": 0.3749, + "step": 514 + }, + { + "epoch": 0.3704369717676677, + "grad_norm": 23.608326345542828, + "learning_rate": 2.99166386579871e-06, + "loss": 0.388, + "step": 515 + }, + { + "epoch": 0.3711562668584787, + "grad_norm": 8.114143071314698, + "learning_rate": 2.991628006330596e-06, + "loss": 0.2778, + "step": 516 + }, + { + "epoch": 0.3718755619492897, + "grad_norm": 7.394027502213448, + "learning_rate": 2.991592070115604e-06, + "loss": 0.1384, + "step": 517 + }, + { + "epoch": 0.3725948570401007, + "grad_norm": 10.12722355127731, + "learning_rate": 2.9915560571555805e-06, + "loss": 0.2268, + "step": 518 + }, + { + "epoch": 0.3733141521309117, + "grad_norm": 8.959721954921136, + "learning_rate": 2.9915199674523797e-06, + "loss": 0.154, + "step": 519 + }, + { + "epoch": 0.3740334472217227, + "grad_norm": 11.097370858106627, + "learning_rate": 2.991483801007859e-06, + "loss": 0.2331, + "step": 520 + }, + { + "epoch": 0.3747527423125337, + "grad_norm": 10.23729402319527, + "learning_rate": 2.9914475578238775e-06, + "loss": 0.2367, + "step": 521 + }, + { + "epoch": 0.3754720374033447, + "grad_norm": 8.759875368003424, + "learning_rate": 2.9914112379023017e-06, + "loss": 0.2951, + "step": 522 + }, + { + "epoch": 0.3761913324941557, + "grad_norm": 8.902771910725425, + "learning_rate": 2.9913748412449996e-06, + "loss": 0.3176, + "step": 523 + }, + { + "epoch": 0.37691062758496674, + "grad_norm": 7.050805312115364, + "learning_rate": 2.991338367853844e-06, + "loss": 0.2211, + "step": 524 + }, + { + "epoch": 0.3776299226757777, + "grad_norm": 7.734429974304813, + "learning_rate": 2.9913018177307105e-06, + "loss": 0.231, + "step": 525 + }, + { + "epoch": 0.37834921776658875, + "grad_norm": 5.43757036795608, + "learning_rate": 2.9912651908774814e-06, + "loss": 0.1995, + "step": 526 + }, + { + "epoch": 0.37906851285739973, + "grad_norm": 6.483873167419547, + "learning_rate": 2.9912284872960397e-06, + "loss": 0.2105, + "step": 527 + }, + { + "epoch": 0.37978780794821076, + "grad_norm": 7.3010543475949925, + "learning_rate": 2.9911917069882753e-06, + "loss": 0.0761, + "step": 528 + }, + { + "epoch": 0.38050710303902174, + "grad_norm": 8.32884538218666, + "learning_rate": 2.9911548499560794e-06, + "loss": 0.1122, + "step": 529 + }, + { + "epoch": 0.3812263981298328, + "grad_norm": 8.480023160282716, + "learning_rate": 2.9911179162013497e-06, + "loss": 0.2223, + "step": 530 + }, + { + "epoch": 0.38194569322064376, + "grad_norm": 12.045079861664414, + "learning_rate": 2.9910809057259844e-06, + "loss": 0.3155, + "step": 531 + }, + { + "epoch": 0.3826649883114548, + "grad_norm": 7.271205841558662, + "learning_rate": 2.99104381853189e-06, + "loss": 0.1235, + "step": 532 + }, + { + "epoch": 0.38338428340226577, + "grad_norm": 15.42344551275697, + "learning_rate": 2.991006654620973e-06, + "loss": 0.3456, + "step": 533 + }, + { + "epoch": 0.3841035784930768, + "grad_norm": 8.111247993915727, + "learning_rate": 2.9909694139951465e-06, + "loss": 0.0458, + "step": 534 + }, + { + "epoch": 0.3848228735838878, + "grad_norm": 6.181929860819091, + "learning_rate": 2.9909320966563264e-06, + "loss": 0.1166, + "step": 535 + }, + { + "epoch": 0.3855421686746988, + "grad_norm": 4.268457502173128, + "learning_rate": 2.9908947026064325e-06, + "loss": 0.1057, + "step": 536 + }, + { + "epoch": 0.3862614637655098, + "grad_norm": 9.901830929736837, + "learning_rate": 2.990857231847389e-06, + "loss": 0.2146, + "step": 537 + }, + { + "epoch": 0.38698075885632083, + "grad_norm": 6.815730598847511, + "learning_rate": 2.9908196843811242e-06, + "loss": 0.1423, + "step": 538 + }, + { + "epoch": 0.3877000539471318, + "grad_norm": 8.483742294843939, + "learning_rate": 2.990782060209569e-06, + "loss": 0.2209, + "step": 539 + }, + { + "epoch": 0.38841934903794284, + "grad_norm": 10.974153112435175, + "learning_rate": 2.99074435933466e-06, + "loss": 0.5092, + "step": 540 + }, + { + "epoch": 0.3891386441287538, + "grad_norm": 11.761229447582627, + "learning_rate": 2.9907065817583372e-06, + "loss": 0.4249, + "step": 541 + }, + { + "epoch": 0.38985793921956485, + "grad_norm": 5.364895028470896, + "learning_rate": 2.9906687274825434e-06, + "loss": 0.2539, + "step": 542 + }, + { + "epoch": 0.39057723431037583, + "grad_norm": 15.050771989918358, + "learning_rate": 2.990630796509227e-06, + "loss": 0.2649, + "step": 543 + }, + { + "epoch": 0.3912965294011868, + "grad_norm": 9.976972697423335, + "learning_rate": 2.9905927888403396e-06, + "loss": 0.302, + "step": 544 + }, + { + "epoch": 0.39201582449199784, + "grad_norm": 7.391920640469044, + "learning_rate": 2.990554704477836e-06, + "loss": 0.0516, + "step": 545 + }, + { + "epoch": 0.3927351195828088, + "grad_norm": 7.041586378117951, + "learning_rate": 2.990516543423677e-06, + "loss": 0.1401, + "step": 546 + }, + { + "epoch": 0.39345441467361986, + "grad_norm": 8.427415270807284, + "learning_rate": 2.990478305679825e-06, + "loss": 0.3231, + "step": 547 + }, + { + "epoch": 0.39417370976443084, + "grad_norm": 9.918189223212861, + "learning_rate": 2.9904399912482473e-06, + "loss": 0.2854, + "step": 548 + }, + { + "epoch": 0.39489300485524187, + "grad_norm": 5.0042513771867085, + "learning_rate": 2.9904016001309162e-06, + "loss": 0.2168, + "step": 549 + }, + { + "epoch": 0.39561229994605285, + "grad_norm": 7.54059565286538, + "learning_rate": 2.990363132329806e-06, + "loss": 0.1821, + "step": 550 + }, + { + "epoch": 0.3963315950368639, + "grad_norm": 7.863760165364239, + "learning_rate": 2.990324587846897e-06, + "loss": 0.2189, + "step": 551 + }, + { + "epoch": 0.39705089012767486, + "grad_norm": 9.16986964019923, + "learning_rate": 2.990285966684171e-06, + "loss": 0.3196, + "step": 552 + }, + { + "epoch": 0.3977701852184859, + "grad_norm": 9.123454530945438, + "learning_rate": 2.9902472688436166e-06, + "loss": 0.1538, + "step": 553 + }, + { + "epoch": 0.3984894803092969, + "grad_norm": 8.8030982967322, + "learning_rate": 2.990208494327224e-06, + "loss": 0.1839, + "step": 554 + }, + { + "epoch": 0.3992087754001079, + "grad_norm": 6.085133818971661, + "learning_rate": 2.990169643136988e-06, + "loss": 0.1835, + "step": 555 + }, + { + "epoch": 0.3999280704909189, + "grad_norm": 5.4845724521404176, + "learning_rate": 2.9901307152749085e-06, + "loss": 0.1069, + "step": 556 + }, + { + "epoch": 0.4006473655817299, + "grad_norm": 10.55518273666302, + "learning_rate": 2.990091710742988e-06, + "loss": 0.1236, + "step": 557 + }, + { + "epoch": 0.4013666606725409, + "grad_norm": 6.282292647659121, + "learning_rate": 2.9900526295432323e-06, + "loss": 0.1235, + "step": 558 + }, + { + "epoch": 0.40208595576335193, + "grad_norm": 6.778760028112376, + "learning_rate": 2.9900134716776535e-06, + "loss": 0.1123, + "step": 559 + }, + { + "epoch": 0.4028052508541629, + "grad_norm": 10.965481679600794, + "learning_rate": 2.989974237148266e-06, + "loss": 0.2948, + "step": 560 + }, + { + "epoch": 0.40352454594497394, + "grad_norm": 9.358817453662242, + "learning_rate": 2.9899349259570887e-06, + "loss": 0.3162, + "step": 561 + }, + { + "epoch": 0.4042438410357849, + "grad_norm": 9.072919156506938, + "learning_rate": 2.989895538106144e-06, + "loss": 0.1678, + "step": 562 + }, + { + "epoch": 0.40496313612659596, + "grad_norm": 7.334111730541219, + "learning_rate": 2.989856073597458e-06, + "loss": 0.1197, + "step": 563 + }, + { + "epoch": 0.40568243121740694, + "grad_norm": 4.1093610492547, + "learning_rate": 2.9898165324330617e-06, + "loss": 0.0688, + "step": 564 + }, + { + "epoch": 0.40640172630821797, + "grad_norm": 5.408633667840331, + "learning_rate": 2.9897769146149896e-06, + "loss": 0.1209, + "step": 565 + }, + { + "epoch": 0.40712102139902895, + "grad_norm": 9.990975918910715, + "learning_rate": 2.9897372201452804e-06, + "loss": 0.2834, + "step": 566 + }, + { + "epoch": 0.40784031648984, + "grad_norm": 7.2589940043717895, + "learning_rate": 2.9896974490259754e-06, + "loss": 0.041, + "step": 567 + }, + { + "epoch": 0.40855961158065096, + "grad_norm": 7.773056940759599, + "learning_rate": 2.989657601259122e-06, + "loss": 0.1928, + "step": 568 + }, + { + "epoch": 0.409278906671462, + "grad_norm": 6.567740277178054, + "learning_rate": 2.9896176768467703e-06, + "loss": 0.0657, + "step": 569 + }, + { + "epoch": 0.409998201762273, + "grad_norm": 10.81902515901499, + "learning_rate": 2.9895776757909735e-06, + "loss": 0.1264, + "step": 570 + }, + { + "epoch": 0.41071749685308395, + "grad_norm": 8.865799846970328, + "learning_rate": 2.989537598093791e-06, + "loss": 0.3872, + "step": 571 + }, + { + "epoch": 0.411436791943895, + "grad_norm": 15.254691911822684, + "learning_rate": 2.9894974437572835e-06, + "loss": 0.2536, + "step": 572 + }, + { + "epoch": 0.41215608703470596, + "grad_norm": 7.159124207249005, + "learning_rate": 2.9894572127835187e-06, + "loss": 0.1144, + "step": 573 + }, + { + "epoch": 0.412875382125517, + "grad_norm": 8.79482496857362, + "learning_rate": 2.989416905174565e-06, + "loss": 0.1438, + "step": 574 + }, + { + "epoch": 0.413594677216328, + "grad_norm": 7.027221158617834, + "learning_rate": 2.989376520932497e-06, + "loss": 0.2318, + "step": 575 + }, + { + "epoch": 0.414313972307139, + "grad_norm": 7.251576110185878, + "learning_rate": 2.989336060059393e-06, + "loss": 0.1345, + "step": 576 + }, + { + "epoch": 0.41503326739795, + "grad_norm": 5.7282401777393845, + "learning_rate": 2.989295522557334e-06, + "loss": 0.1738, + "step": 577 + }, + { + "epoch": 0.415752562488761, + "grad_norm": 7.512920906389525, + "learning_rate": 2.9892549084284058e-06, + "loss": 0.1235, + "step": 578 + }, + { + "epoch": 0.416471857579572, + "grad_norm": 5.087999073143586, + "learning_rate": 2.989214217674698e-06, + "loss": 0.048, + "step": 579 + }, + { + "epoch": 0.41719115267038304, + "grad_norm": 11.190401308971477, + "learning_rate": 2.9891734502983053e-06, + "loss": 0.4374, + "step": 580 + }, + { + "epoch": 0.417910447761194, + "grad_norm": 8.535846595307731, + "learning_rate": 2.9891326063013237e-06, + "loss": 0.3329, + "step": 581 + }, + { + "epoch": 0.41862974285200505, + "grad_norm": 8.045877789828847, + "learning_rate": 2.9890916856858554e-06, + "loss": 0.2156, + "step": 582 + }, + { + "epoch": 0.419349037942816, + "grad_norm": 6.369909292910831, + "learning_rate": 2.9890506884540064e-06, + "loss": 0.5132, + "step": 583 + }, + { + "epoch": 0.42006833303362706, + "grad_norm": 7.818690288087174, + "learning_rate": 2.989009614607885e-06, + "loss": 0.1819, + "step": 584 + }, + { + "epoch": 0.42078762812443804, + "grad_norm": 6.884612407988808, + "learning_rate": 2.9889684641496055e-06, + "loss": 0.0953, + "step": 585 + }, + { + "epoch": 0.4215069232152491, + "grad_norm": 8.551756168499713, + "learning_rate": 2.988927237081284e-06, + "loss": 0.3049, + "step": 586 + }, + { + "epoch": 0.42222621830606005, + "grad_norm": 6.796960057299801, + "learning_rate": 2.9888859334050435e-06, + "loss": 0.0677, + "step": 587 + }, + { + "epoch": 0.4229455133968711, + "grad_norm": 8.561352029905176, + "learning_rate": 2.9888445531230076e-06, + "loss": 0.0882, + "step": 588 + }, + { + "epoch": 0.42366480848768207, + "grad_norm": 7.6161044367457835, + "learning_rate": 2.988803096237306e-06, + "loss": 0.3083, + "step": 589 + }, + { + "epoch": 0.4243841035784931, + "grad_norm": 13.454930616180322, + "learning_rate": 2.988761562750071e-06, + "loss": 0.59, + "step": 590 + }, + { + "epoch": 0.4251033986693041, + "grad_norm": 8.0279068423302, + "learning_rate": 2.988719952663441e-06, + "loss": 0.1738, + "step": 591 + }, + { + "epoch": 0.4258226937601151, + "grad_norm": 6.064140395677239, + "learning_rate": 2.988678265979555e-06, + "loss": 0.1167, + "step": 592 + }, + { + "epoch": 0.4265419888509261, + "grad_norm": 7.566254479498275, + "learning_rate": 2.9886365027005595e-06, + "loss": 0.167, + "step": 593 + }, + { + "epoch": 0.4272612839417371, + "grad_norm": 7.160711044708271, + "learning_rate": 2.9885946628286028e-06, + "loss": 0.1023, + "step": 594 + }, + { + "epoch": 0.4279805790325481, + "grad_norm": 10.181651394413505, + "learning_rate": 2.9885527463658377e-06, + "loss": 0.257, + "step": 595 + }, + { + "epoch": 0.4286998741233591, + "grad_norm": 5.118949215894016, + "learning_rate": 2.98851075331442e-06, + "loss": 0.0944, + "step": 596 + }, + { + "epoch": 0.4294191692141701, + "grad_norm": 7.512608938043663, + "learning_rate": 2.988468683676512e-06, + "loss": 0.2957, + "step": 597 + }, + { + "epoch": 0.4301384643049811, + "grad_norm": 6.641450299767169, + "learning_rate": 2.9884265374542765e-06, + "loss": 0.1542, + "step": 598 + }, + { + "epoch": 0.43085775939579213, + "grad_norm": 6.652629687802819, + "learning_rate": 2.988384314649883e-06, + "loss": 0.252, + "step": 599 + }, + { + "epoch": 0.4315770544866031, + "grad_norm": 9.917527283754659, + "learning_rate": 2.988342015265504e-06, + "loss": 0.322, + "step": 600 + }, + { + "epoch": 0.43229634957741414, + "grad_norm": 7.346807434337759, + "learning_rate": 2.988299639303316e-06, + "loss": 0.3069, + "step": 601 + }, + { + "epoch": 0.4330156446682251, + "grad_norm": 12.027092956868286, + "learning_rate": 2.9882571867654986e-06, + "loss": 0.3036, + "step": 602 + }, + { + "epoch": 0.43373493975903615, + "grad_norm": 7.871003547860107, + "learning_rate": 2.9882146576542357e-06, + "loss": 0.1724, + "step": 603 + }, + { + "epoch": 0.43445423484984713, + "grad_norm": 11.4574400805975, + "learning_rate": 2.988172051971717e-06, + "loss": 0.5131, + "step": 604 + }, + { + "epoch": 0.43517352994065817, + "grad_norm": 6.963701130057769, + "learning_rate": 2.9881293697201334e-06, + "loss": 0.2412, + "step": 605 + }, + { + "epoch": 0.43589282503146914, + "grad_norm": 7.798993128468614, + "learning_rate": 2.988086610901681e-06, + "loss": 0.1144, + "step": 606 + }, + { + "epoch": 0.4366121201222802, + "grad_norm": 9.079845766536438, + "learning_rate": 2.9880437755185613e-06, + "loss": 0.312, + "step": 607 + }, + { + "epoch": 0.43733141521309116, + "grad_norm": 8.26055961405721, + "learning_rate": 2.9880008635729766e-06, + "loss": 0.2598, + "step": 608 + }, + { + "epoch": 0.4380507103039022, + "grad_norm": 6.518289068819356, + "learning_rate": 2.9879578750671356e-06, + "loss": 0.1522, + "step": 609 + }, + { + "epoch": 0.43877000539471317, + "grad_norm": 10.980823267639687, + "learning_rate": 2.987914810003249e-06, + "loss": 0.331, + "step": 610 + }, + { + "epoch": 0.4394893004855242, + "grad_norm": 8.462160580724552, + "learning_rate": 2.9878716683835343e-06, + "loss": 0.2036, + "step": 611 + }, + { + "epoch": 0.4402085955763352, + "grad_norm": 6.504916199265367, + "learning_rate": 2.9878284502102104e-06, + "loss": 0.2739, + "step": 612 + }, + { + "epoch": 0.4409278906671462, + "grad_norm": 9.6391017965284, + "learning_rate": 2.9877851554855007e-06, + "loss": 0.26, + "step": 613 + }, + { + "epoch": 0.4416471857579572, + "grad_norm": 6.822877387341788, + "learning_rate": 2.987741784211633e-06, + "loss": 0.1603, + "step": 614 + }, + { + "epoch": 0.44236648084876823, + "grad_norm": 8.354057412567972, + "learning_rate": 2.987698336390839e-06, + "loss": 0.2595, + "step": 615 + }, + { + "epoch": 0.4430857759395792, + "grad_norm": 7.712805431473086, + "learning_rate": 2.987654812025354e-06, + "loss": 0.3013, + "step": 616 + }, + { + "epoch": 0.44380507103039024, + "grad_norm": 6.820823276133487, + "learning_rate": 2.9876112111174175e-06, + "loss": 0.0679, + "step": 617 + }, + { + "epoch": 0.4445243661212012, + "grad_norm": 7.3218377848106275, + "learning_rate": 2.9875675336692728e-06, + "loss": 0.1217, + "step": 618 + }, + { + "epoch": 0.44524366121201225, + "grad_norm": 7.803188661672437, + "learning_rate": 2.9875237796831665e-06, + "loss": 0.2676, + "step": 619 + }, + { + "epoch": 0.44596295630282323, + "grad_norm": 8.325760523497854, + "learning_rate": 2.9874799491613515e-06, + "loss": 0.316, + "step": 620 + }, + { + "epoch": 0.4466822513936342, + "grad_norm": 6.090735806757629, + "learning_rate": 2.987436042106081e-06, + "loss": 0.291, + "step": 621 + }, + { + "epoch": 0.44740154648444524, + "grad_norm": 3.802907758694769, + "learning_rate": 2.9873920585196155e-06, + "loss": 0.0147, + "step": 622 + }, + { + "epoch": 0.4481208415752562, + "grad_norm": 4.945923938755192, + "learning_rate": 2.9873479984042177e-06, + "loss": 0.0283, + "step": 623 + }, + { + "epoch": 0.44884013666606726, + "grad_norm": 6.643269050647207, + "learning_rate": 2.987303861762154e-06, + "loss": 0.1419, + "step": 624 + }, + { + "epoch": 0.44955943175687824, + "grad_norm": 11.47672021422071, + "learning_rate": 2.987259648595696e-06, + "loss": 0.5138, + "step": 625 + }, + { + "epoch": 0.45027872684768927, + "grad_norm": 12.200240121757396, + "learning_rate": 2.9872153589071185e-06, + "loss": 0.2653, + "step": 626 + }, + { + "epoch": 0.45099802193850025, + "grad_norm": 5.487828060112799, + "learning_rate": 2.9871709926987e-06, + "loss": 0.0673, + "step": 627 + }, + { + "epoch": 0.4517173170293113, + "grad_norm": 7.96134008578685, + "learning_rate": 2.9871265499727227e-06, + "loss": 0.0956, + "step": 628 + }, + { + "epoch": 0.45243661212012226, + "grad_norm": 9.014733530641363, + "learning_rate": 2.9870820307314743e-06, + "loss": 0.099, + "step": 629 + }, + { + "epoch": 0.4531559072109333, + "grad_norm": 5.620593454289586, + "learning_rate": 2.9870374349772448e-06, + "loss": 0.0739, + "step": 630 + }, + { + "epoch": 0.4538752023017443, + "grad_norm": 12.261486479299116, + "learning_rate": 2.986992762712329e-06, + "loss": 0.4411, + "step": 631 + }, + { + "epoch": 0.4545944973925553, + "grad_norm": 9.810601747113425, + "learning_rate": 2.9869480139390247e-06, + "loss": 0.2169, + "step": 632 + }, + { + "epoch": 0.4553137924833663, + "grad_norm": 12.102061342834876, + "learning_rate": 2.9869031886596357e-06, + "loss": 0.1651, + "step": 633 + }, + { + "epoch": 0.4560330875741773, + "grad_norm": 5.488367023817783, + "learning_rate": 2.9868582868764667e-06, + "loss": 0.1201, + "step": 634 + }, + { + "epoch": 0.4567523826649883, + "grad_norm": 7.617074599084058, + "learning_rate": 2.986813308591829e-06, + "loss": 0.1553, + "step": 635 + }, + { + "epoch": 0.45747167775579933, + "grad_norm": 5.5786678434242605, + "learning_rate": 2.986768253808037e-06, + "loss": 0.2147, + "step": 636 + }, + { + "epoch": 0.4581909728466103, + "grad_norm": 8.989821213248181, + "learning_rate": 2.9867231225274075e-06, + "loss": 0.3028, + "step": 637 + }, + { + "epoch": 0.45891026793742135, + "grad_norm": 5.091181429753673, + "learning_rate": 2.986677914752264e-06, + "loss": 0.0591, + "step": 638 + }, + { + "epoch": 0.4596295630282323, + "grad_norm": 10.911399896249282, + "learning_rate": 2.986632630484932e-06, + "loss": 0.5262, + "step": 639 + }, + { + "epoch": 0.46034885811904336, + "grad_norm": 6.656110126300123, + "learning_rate": 2.9865872697277417e-06, + "loss": 0.1292, + "step": 640 + }, + { + "epoch": 0.46106815320985434, + "grad_norm": 10.0668832546612, + "learning_rate": 2.9865418324830263e-06, + "loss": 0.3637, + "step": 641 + }, + { + "epoch": 0.46178744830066537, + "grad_norm": 8.427653507735005, + "learning_rate": 2.986496318753124e-06, + "loss": 0.107, + "step": 642 + }, + { + "epoch": 0.46250674339147635, + "grad_norm": 4.764136536876361, + "learning_rate": 2.986450728540377e-06, + "loss": 0.1999, + "step": 643 + }, + { + "epoch": 0.4632260384822874, + "grad_norm": 3.9011328290520995, + "learning_rate": 2.986405061847131e-06, + "loss": 0.0203, + "step": 644 + }, + { + "epoch": 0.46394533357309836, + "grad_norm": 6.459055537304067, + "learning_rate": 2.986359318675735e-06, + "loss": 0.0821, + "step": 645 + }, + { + "epoch": 0.4646646286639094, + "grad_norm": 7.965736761871791, + "learning_rate": 2.9863134990285425e-06, + "loss": 0.4077, + "step": 646 + }, + { + "epoch": 0.4653839237547204, + "grad_norm": 6.2188393746309965, + "learning_rate": 2.9862676029079114e-06, + "loss": 0.3448, + "step": 647 + }, + { + "epoch": 0.46610321884553135, + "grad_norm": 7.443160591071639, + "learning_rate": 2.986221630316203e-06, + "loss": 0.1343, + "step": 648 + }, + { + "epoch": 0.4668225139363424, + "grad_norm": 7.108644129952765, + "learning_rate": 2.986175581255783e-06, + "loss": 0.2194, + "step": 649 + }, + { + "epoch": 0.46754180902715337, + "grad_norm": 6.813646701011755, + "learning_rate": 2.9861294557290204e-06, + "loss": 0.2985, + "step": 650 + }, + { + "epoch": 0.4682611041179644, + "grad_norm": 6.585814058876483, + "learning_rate": 2.986083253738288e-06, + "loss": 0.1636, + "step": 651 + }, + { + "epoch": 0.4689803992087754, + "grad_norm": 8.545202459019087, + "learning_rate": 2.986036975285964e-06, + "loss": 0.3853, + "step": 652 + }, + { + "epoch": 0.4696996942995864, + "grad_norm": 8.29350019433357, + "learning_rate": 2.9859906203744294e-06, + "loss": 0.2736, + "step": 653 + }, + { + "epoch": 0.4704189893903974, + "grad_norm": 7.26657270385466, + "learning_rate": 2.985944189006068e-06, + "loss": 0.3296, + "step": 654 + }, + { + "epoch": 0.4711382844812084, + "grad_norm": 4.716989904329773, + "learning_rate": 2.9858976811832696e-06, + "loss": 0.1396, + "step": 655 + }, + { + "epoch": 0.4718575795720194, + "grad_norm": 5.303771176013311, + "learning_rate": 2.9858510969084277e-06, + "loss": 0.2398, + "step": 656 + }, + { + "epoch": 0.47257687466283044, + "grad_norm": 5.264001879842815, + "learning_rate": 2.985804436183938e-06, + "loss": 0.2143, + "step": 657 + }, + { + "epoch": 0.4732961697536414, + "grad_norm": 8.491284647304514, + "learning_rate": 2.985757699012202e-06, + "loss": 0.4539, + "step": 658 + }, + { + "epoch": 0.47401546484445245, + "grad_norm": 9.197947922453682, + "learning_rate": 2.9857108853956246e-06, + "loss": 0.3073, + "step": 659 + }, + { + "epoch": 0.47473475993526343, + "grad_norm": 8.184881279794077, + "learning_rate": 2.9856639953366134e-06, + "loss": 0.2799, + "step": 660 + }, + { + "epoch": 0.47545405502607446, + "grad_norm": 6.276697037518068, + "learning_rate": 2.9856170288375815e-06, + "loss": 0.2185, + "step": 661 + }, + { + "epoch": 0.47617335011688544, + "grad_norm": 3.5239432360839817, + "learning_rate": 2.9855699859009463e-06, + "loss": 0.036, + "step": 662 + }, + { + "epoch": 0.4768926452076965, + "grad_norm": 7.090429537477977, + "learning_rate": 2.9855228665291273e-06, + "loss": 0.3242, + "step": 663 + }, + { + "epoch": 0.47761194029850745, + "grad_norm": 9.354126086365557, + "learning_rate": 2.9854756707245487e-06, + "loss": 0.1249, + "step": 664 + }, + { + "epoch": 0.4783312353893185, + "grad_norm": 8.30907469094051, + "learning_rate": 2.9854283984896394e-06, + "loss": 0.3307, + "step": 665 + }, + { + "epoch": 0.47905053048012947, + "grad_norm": 9.96926455120718, + "learning_rate": 2.9853810498268316e-06, + "loss": 0.2644, + "step": 666 + }, + { + "epoch": 0.4797698255709405, + "grad_norm": 10.064518275162772, + "learning_rate": 2.9853336247385607e-06, + "loss": 0.1975, + "step": 667 + }, + { + "epoch": 0.4804891206617515, + "grad_norm": 9.843722243189204, + "learning_rate": 2.985286123227268e-06, + "loss": 0.1901, + "step": 668 + }, + { + "epoch": 0.4812084157525625, + "grad_norm": 13.833537456986445, + "learning_rate": 2.9852385452953966e-06, + "loss": 0.5601, + "step": 669 + }, + { + "epoch": 0.4819277108433735, + "grad_norm": 10.319994056237286, + "learning_rate": 2.985190890945395e-06, + "loss": 0.2334, + "step": 670 + }, + { + "epoch": 0.4826470059341845, + "grad_norm": 5.913399332588977, + "learning_rate": 2.9851431601797144e-06, + "loss": 0.0998, + "step": 671 + }, + { + "epoch": 0.4833663010249955, + "grad_norm": 9.09397873095818, + "learning_rate": 2.9850953530008115e-06, + "loss": 0.1238, + "step": 672 + }, + { + "epoch": 0.4840855961158065, + "grad_norm": 8.559550481125624, + "learning_rate": 2.985047469411146e-06, + "loss": 0.22, + "step": 673 + }, + { + "epoch": 0.4848048912066175, + "grad_norm": 5.346905375652378, + "learning_rate": 2.984999509413181e-06, + "loss": 0.2827, + "step": 674 + }, + { + "epoch": 0.4855241862974285, + "grad_norm": 5.272507163621891, + "learning_rate": 2.984951473009384e-06, + "loss": 0.0607, + "step": 675 + }, + { + "epoch": 0.48624348138823953, + "grad_norm": 8.902859741949358, + "learning_rate": 2.9849033602022273e-06, + "loss": 0.0788, + "step": 676 + }, + { + "epoch": 0.4869627764790505, + "grad_norm": 5.137273810772606, + "learning_rate": 2.984855170994186e-06, + "loss": 0.1291, + "step": 677 + }, + { + "epoch": 0.48768207156986154, + "grad_norm": 9.110879109386415, + "learning_rate": 2.98480690538774e-06, + "loss": 0.1309, + "step": 678 + }, + { + "epoch": 0.4884013666606725, + "grad_norm": 7.998516210908917, + "learning_rate": 2.984758563385372e-06, + "loss": 0.3183, + "step": 679 + }, + { + "epoch": 0.48912066175148355, + "grad_norm": 7.2025293524855485, + "learning_rate": 2.984710144989569e-06, + "loss": 0.3035, + "step": 680 + }, + { + "epoch": 0.48983995684229453, + "grad_norm": 11.145719945762266, + "learning_rate": 2.9846616502028233e-06, + "loss": 0.2932, + "step": 681 + }, + { + "epoch": 0.49055925193310557, + "grad_norm": 13.72270979098324, + "learning_rate": 2.9846130790276292e-06, + "loss": 0.4495, + "step": 682 + }, + { + "epoch": 0.49127854702391655, + "grad_norm": 7.064784455853499, + "learning_rate": 2.984564431466486e-06, + "loss": 0.3142, + "step": 683 + }, + { + "epoch": 0.4919978421147276, + "grad_norm": 7.014948623482035, + "learning_rate": 2.984515707521897e-06, + "loss": 0.1956, + "step": 684 + }, + { + "epoch": 0.49271713720553856, + "grad_norm": 6.795210125441131, + "learning_rate": 2.9844669071963684e-06, + "loss": 0.1822, + "step": 685 + }, + { + "epoch": 0.4934364322963496, + "grad_norm": 8.325957162522203, + "learning_rate": 2.9844180304924117e-06, + "loss": 0.0935, + "step": 686 + }, + { + "epoch": 0.49415572738716057, + "grad_norm": 9.216902870286452, + "learning_rate": 2.984369077412542e-06, + "loss": 0.2075, + "step": 687 + }, + { + "epoch": 0.4948750224779716, + "grad_norm": 2.841775058998902, + "learning_rate": 2.984320047959277e-06, + "loss": 0.0145, + "step": 688 + }, + { + "epoch": 0.4955943175687826, + "grad_norm": 5.812174435888005, + "learning_rate": 2.98427094213514e-06, + "loss": 0.2834, + "step": 689 + }, + { + "epoch": 0.4963136126595936, + "grad_norm": 8.10164249230395, + "learning_rate": 2.9842217599426575e-06, + "loss": 0.214, + "step": 690 + }, + { + "epoch": 0.4970329077504046, + "grad_norm": 3.8883802970936796, + "learning_rate": 2.98417250138436e-06, + "loss": 0.0363, + "step": 691 + }, + { + "epoch": 0.49775220284121563, + "grad_norm": 8.266520114913085, + "learning_rate": 2.984123166462782e-06, + "loss": 0.1406, + "step": 692 + }, + { + "epoch": 0.4984714979320266, + "grad_norm": 6.241849146339004, + "learning_rate": 2.9840737551804614e-06, + "loss": 0.2336, + "step": 693 + }, + { + "epoch": 0.49919079302283764, + "grad_norm": 10.290181309552962, + "learning_rate": 2.9840242675399413e-06, + "loss": 0.3195, + "step": 694 + }, + { + "epoch": 0.4999100881136486, + "grad_norm": 9.230658413338618, + "learning_rate": 2.9839747035437676e-06, + "loss": 0.2065, + "step": 695 + }, + { + "epoch": 0.5006293832044596, + "grad_norm": 6.484240878839433, + "learning_rate": 2.98392506319449e-06, + "loss": 0.1148, + "step": 696 + }, + { + "epoch": 0.5013486782952706, + "grad_norm": 6.185091579068325, + "learning_rate": 2.9838753464946635e-06, + "loss": 0.3242, + "step": 697 + }, + { + "epoch": 0.5020679733860817, + "grad_norm": 3.5617342463535513, + "learning_rate": 2.983825553446845e-06, + "loss": 0.1263, + "step": 698 + }, + { + "epoch": 0.5027872684768926, + "grad_norm": 8.275063334715663, + "learning_rate": 2.983775684053597e-06, + "loss": 0.1571, + "step": 699 + }, + { + "epoch": 0.5035065635677036, + "grad_norm": 11.220184159353963, + "learning_rate": 2.9837257383174856e-06, + "loss": 0.2024, + "step": 700 + }, + { + "epoch": 0.5042258586585147, + "grad_norm": 8.877131241467795, + "learning_rate": 2.9836757162410803e-06, + "loss": 0.1502, + "step": 701 + }, + { + "epoch": 0.5049451537493257, + "grad_norm": 7.9782318968943144, + "learning_rate": 2.983625617826955e-06, + "loss": 0.4078, + "step": 702 + }, + { + "epoch": 0.5056644488401366, + "grad_norm": 7.498834138261986, + "learning_rate": 2.983575443077687e-06, + "loss": 0.1716, + "step": 703 + }, + { + "epoch": 0.5063837439309476, + "grad_norm": 7.683376921578665, + "learning_rate": 2.9835251919958582e-06, + "loss": 0.3026, + "step": 704 + }, + { + "epoch": 0.5071030390217587, + "grad_norm": 8.30032932846491, + "learning_rate": 2.983474864584054e-06, + "loss": 0.2691, + "step": 705 + }, + { + "epoch": 0.5078223341125697, + "grad_norm": 4.6011062591079135, + "learning_rate": 2.983424460844864e-06, + "loss": 0.1886, + "step": 706 + }, + { + "epoch": 0.5085416292033806, + "grad_norm": 7.5362654330016, + "learning_rate": 2.983373980780881e-06, + "loss": 0.2406, + "step": 707 + }, + { + "epoch": 0.5092609242941917, + "grad_norm": 7.103149578313491, + "learning_rate": 2.9833234243947035e-06, + "loss": 0.2909, + "step": 708 + }, + { + "epoch": 0.5099802193850027, + "grad_norm": 7.0080138150005675, + "learning_rate": 2.9832727916889317e-06, + "loss": 0.226, + "step": 709 + }, + { + "epoch": 0.5106995144758137, + "grad_norm": 7.438000679237786, + "learning_rate": 2.9832220826661706e-06, + "loss": 0.2728, + "step": 710 + }, + { + "epoch": 0.5114188095666247, + "grad_norm": 5.9851638874817406, + "learning_rate": 2.98317129732903e-06, + "loss": 0.2456, + "step": 711 + }, + { + "epoch": 0.5121381046574357, + "grad_norm": 8.225734653045484, + "learning_rate": 2.983120435680122e-06, + "loss": 0.1945, + "step": 712 + }, + { + "epoch": 0.5128573997482467, + "grad_norm": 5.482244552245788, + "learning_rate": 2.9830694977220643e-06, + "loss": 0.1894, + "step": 713 + }, + { + "epoch": 0.5135766948390578, + "grad_norm": 10.237687884137133, + "learning_rate": 2.9830184834574777e-06, + "loss": 0.2621, + "step": 714 + }, + { + "epoch": 0.5142959899298687, + "grad_norm": 7.576925742496242, + "learning_rate": 2.982967392888987e-06, + "loss": 0.3155, + "step": 715 + }, + { + "epoch": 0.5150152850206797, + "grad_norm": 7.294886815052957, + "learning_rate": 2.98291622601922e-06, + "loss": 0.2994, + "step": 716 + }, + { + "epoch": 0.5157345801114908, + "grad_norm": 7.306832655659561, + "learning_rate": 2.9828649828508104e-06, + "loss": 0.2122, + "step": 717 + }, + { + "epoch": 0.5164538752023018, + "grad_norm": 7.26810226698831, + "learning_rate": 2.982813663386394e-06, + "loss": 0.1442, + "step": 718 + }, + { + "epoch": 0.5171731702931127, + "grad_norm": 10.433100264454152, + "learning_rate": 2.982762267628612e-06, + "loss": 0.1599, + "step": 719 + }, + { + "epoch": 0.5178924653839238, + "grad_norm": 9.097836449521381, + "learning_rate": 2.982710795580108e-06, + "loss": 0.1606, + "step": 720 + }, + { + "epoch": 0.5186117604747348, + "grad_norm": 7.726615576774665, + "learning_rate": 2.982659247243531e-06, + "loss": 0.3383, + "step": 721 + }, + { + "epoch": 0.5193310555655458, + "grad_norm": 10.375030166002839, + "learning_rate": 2.9826076226215332e-06, + "loss": 0.1489, + "step": 722 + }, + { + "epoch": 0.5200503506563567, + "grad_norm": 4.060363415490424, + "learning_rate": 2.9825559217167704e-06, + "loss": 0.0198, + "step": 723 + }, + { + "epoch": 0.5207696457471678, + "grad_norm": 4.569597776815956, + "learning_rate": 2.982504144531903e-06, + "loss": 0.0489, + "step": 724 + }, + { + "epoch": 0.5214889408379788, + "grad_norm": 5.166227494881769, + "learning_rate": 2.982452291069595e-06, + "loss": 0.1906, + "step": 725 + }, + { + "epoch": 0.5222082359287897, + "grad_norm": 6.530324433683541, + "learning_rate": 2.9824003613325138e-06, + "loss": 0.1675, + "step": 726 + }, + { + "epoch": 0.5229275310196008, + "grad_norm": 3.5966463961911264, + "learning_rate": 2.9823483553233324e-06, + "loss": 0.0334, + "step": 727 + }, + { + "epoch": 0.5236468261104118, + "grad_norm": 6.514705195264471, + "learning_rate": 2.982296273044725e-06, + "loss": 0.1166, + "step": 728 + }, + { + "epoch": 0.5243661212012228, + "grad_norm": 7.176127604553889, + "learning_rate": 2.982244114499373e-06, + "loss": 0.2241, + "step": 729 + }, + { + "epoch": 0.5250854162920338, + "grad_norm": 10.866936803204483, + "learning_rate": 2.982191879689959e-06, + "loss": 0.1729, + "step": 730 + }, + { + "epoch": 0.5258047113828448, + "grad_norm": 9.303566206043383, + "learning_rate": 2.982139568619171e-06, + "loss": 0.3529, + "step": 731 + }, + { + "epoch": 0.5265240064736558, + "grad_norm": 8.556510060440488, + "learning_rate": 2.9820871812897007e-06, + "loss": 0.3519, + "step": 732 + }, + { + "epoch": 0.5272433015644669, + "grad_norm": 4.548847054868503, + "learning_rate": 2.9820347177042427e-06, + "loss": 0.0885, + "step": 733 + }, + { + "epoch": 0.5279625966552778, + "grad_norm": 6.4970436363222355, + "learning_rate": 2.981982177865497e-06, + "loss": 0.227, + "step": 734 + }, + { + "epoch": 0.5286818917460888, + "grad_norm": 7.420866239821348, + "learning_rate": 2.981929561776167e-06, + "loss": 0.2299, + "step": 735 + }, + { + "epoch": 0.5294011868368999, + "grad_norm": 4.732821810252163, + "learning_rate": 2.981876869438959e-06, + "loss": 0.0814, + "step": 736 + }, + { + "epoch": 0.5301204819277109, + "grad_norm": 7.684468584751193, + "learning_rate": 2.981824100856585e-06, + "loss": 0.2261, + "step": 737 + }, + { + "epoch": 0.5308397770185218, + "grad_norm": 11.498093492987481, + "learning_rate": 2.9817712560317602e-06, + "loss": 0.2003, + "step": 738 + }, + { + "epoch": 0.5315590721093328, + "grad_norm": 5.899400287135329, + "learning_rate": 2.9817183349672028e-06, + "loss": 0.1327, + "step": 739 + }, + { + "epoch": 0.5322783672001439, + "grad_norm": 7.28054074370964, + "learning_rate": 2.981665337665636e-06, + "loss": 0.3803, + "step": 740 + }, + { + "epoch": 0.5329976622909549, + "grad_norm": 9.39912153050433, + "learning_rate": 2.9816122641297866e-06, + "loss": 0.3053, + "step": 741 + }, + { + "epoch": 0.5337169573817658, + "grad_norm": 5.43629824544262, + "learning_rate": 2.981559114362385e-06, + "loss": 0.1666, + "step": 742 + }, + { + "epoch": 0.5344362524725769, + "grad_norm": 8.028951891863958, + "learning_rate": 2.981505888366166e-06, + "loss": 0.1775, + "step": 743 + }, + { + "epoch": 0.5351555475633879, + "grad_norm": 7.897529684917291, + "learning_rate": 2.981452586143869e-06, + "loss": 0.172, + "step": 744 + }, + { + "epoch": 0.5358748426541989, + "grad_norm": 7.422495662263877, + "learning_rate": 2.9813992076982357e-06, + "loss": 0.2211, + "step": 745 + }, + { + "epoch": 0.5365941377450099, + "grad_norm": 4.731913158161135, + "learning_rate": 2.9813457530320123e-06, + "loss": 0.1039, + "step": 746 + }, + { + "epoch": 0.5373134328358209, + "grad_norm": 7.63727860675573, + "learning_rate": 2.9812922221479497e-06, + "loss": 0.1062, + "step": 747 + }, + { + "epoch": 0.5380327279266319, + "grad_norm": 8.234338080748223, + "learning_rate": 2.981238615048802e-06, + "loss": 0.1273, + "step": 748 + }, + { + "epoch": 0.538752023017443, + "grad_norm": 8.317128213380808, + "learning_rate": 2.9811849317373274e-06, + "loss": 0.2148, + "step": 749 + }, + { + "epoch": 0.5394713181082539, + "grad_norm": 12.143393161110994, + "learning_rate": 2.981131172216288e-06, + "loss": 0.2847, + "step": 750 + }, + { + "epoch": 0.5401906131990649, + "grad_norm": 7.808658804256654, + "learning_rate": 2.9810773364884493e-06, + "loss": 0.0629, + "step": 751 + }, + { + "epoch": 0.540909908289876, + "grad_norm": 7.742930271868104, + "learning_rate": 2.981023424556582e-06, + "loss": 0.1126, + "step": 752 + }, + { + "epoch": 0.5416292033806869, + "grad_norm": 5.016080151648933, + "learning_rate": 2.9809694364234597e-06, + "loss": 0.1071, + "step": 753 + }, + { + "epoch": 0.5423484984714979, + "grad_norm": 6.053067099455596, + "learning_rate": 2.98091537209186e-06, + "loss": 0.0862, + "step": 754 + }, + { + "epoch": 0.5430677935623089, + "grad_norm": 9.692485088274085, + "learning_rate": 2.980861231564565e-06, + "loss": 0.243, + "step": 755 + }, + { + "epoch": 0.54378708865312, + "grad_norm": 6.457491962704044, + "learning_rate": 2.9808070148443594e-06, + "loss": 0.1179, + "step": 756 + }, + { + "epoch": 0.5445063837439309, + "grad_norm": 8.732374837188866, + "learning_rate": 2.980752721934034e-06, + "loss": 0.2488, + "step": 757 + }, + { + "epoch": 0.5452256788347419, + "grad_norm": 5.924270881380956, + "learning_rate": 2.9806983528363816e-06, + "loss": 0.1444, + "step": 758 + }, + { + "epoch": 0.545944973925553, + "grad_norm": 7.5988289184306, + "learning_rate": 2.9806439075541993e-06, + "loss": 0.1845, + "step": 759 + }, + { + "epoch": 0.546664269016364, + "grad_norm": 11.24620708238816, + "learning_rate": 2.980589386090289e-06, + "loss": 0.0838, + "step": 760 + }, + { + "epoch": 0.5473835641071749, + "grad_norm": 5.958027477253964, + "learning_rate": 2.980534788447456e-06, + "loss": 0.1508, + "step": 761 + }, + { + "epoch": 0.548102859197986, + "grad_norm": 13.125376727031792, + "learning_rate": 2.9804801146285088e-06, + "loss": 0.2799, + "step": 762 + }, + { + "epoch": 0.548822154288797, + "grad_norm": 8.536323223419584, + "learning_rate": 2.980425364636261e-06, + "loss": 0.1704, + "step": 763 + }, + { + "epoch": 0.549541449379608, + "grad_norm": 9.20526378210383, + "learning_rate": 2.980370538473529e-06, + "loss": 0.2532, + "step": 764 + }, + { + "epoch": 0.550260744470419, + "grad_norm": 4.356231652665598, + "learning_rate": 2.9803156361431346e-06, + "loss": 0.0946, + "step": 765 + }, + { + "epoch": 0.55098003956123, + "grad_norm": 3.5342297318161093, + "learning_rate": 2.980260657647902e-06, + "loss": 0.0499, + "step": 766 + }, + { + "epoch": 0.551699334652041, + "grad_norm": 5.79384531558842, + "learning_rate": 2.9802056029906597e-06, + "loss": 0.1506, + "step": 767 + }, + { + "epoch": 0.552418629742852, + "grad_norm": 9.612019951464557, + "learning_rate": 2.980150472174241e-06, + "loss": 0.1162, + "step": 768 + }, + { + "epoch": 0.553137924833663, + "grad_norm": 6.795021971884667, + "learning_rate": 2.980095265201482e-06, + "loss": 0.2568, + "step": 769 + }, + { + "epoch": 0.553857219924474, + "grad_norm": 4.046837386431047, + "learning_rate": 2.9800399820752235e-06, + "loss": 0.0829, + "step": 770 + }, + { + "epoch": 0.554576515015285, + "grad_norm": 9.140823444754814, + "learning_rate": 2.9799846227983097e-06, + "loss": 0.2293, + "step": 771 + }, + { + "epoch": 0.5552958101060961, + "grad_norm": 4.868268078666935, + "learning_rate": 2.979929187373589e-06, + "loss": 0.0318, + "step": 772 + }, + { + "epoch": 0.556015105196907, + "grad_norm": 9.610275083769052, + "learning_rate": 2.979873675803914e-06, + "loss": 0.261, + "step": 773 + }, + { + "epoch": 0.556734400287718, + "grad_norm": 8.981453072200969, + "learning_rate": 2.9798180880921405e-06, + "loss": 0.2829, + "step": 774 + }, + { + "epoch": 0.5574536953785291, + "grad_norm": 5.700400150105251, + "learning_rate": 2.9797624242411286e-06, + "loss": 0.2019, + "step": 775 + }, + { + "epoch": 0.55817299046934, + "grad_norm": 10.113887925402263, + "learning_rate": 2.979706684253742e-06, + "loss": 0.2332, + "step": 776 + }, + { + "epoch": 0.558892285560151, + "grad_norm": 4.808140392186802, + "learning_rate": 2.9796508681328488e-06, + "loss": 0.1976, + "step": 777 + }, + { + "epoch": 0.5596115806509621, + "grad_norm": 7.986932477784339, + "learning_rate": 2.9795949758813216e-06, + "loss": 0.3967, + "step": 778 + }, + { + "epoch": 0.5603308757417731, + "grad_norm": 9.85334741423081, + "learning_rate": 2.9795390075020353e-06, + "loss": 0.3556, + "step": 779 + }, + { + "epoch": 0.561050170832584, + "grad_norm": 2.610636021212666, + "learning_rate": 2.97948296299787e-06, + "loss": 0.0159, + "step": 780 + }, + { + "epoch": 0.561769465923395, + "grad_norm": 5.897553693290661, + "learning_rate": 2.9794268423717085e-06, + "loss": 0.0621, + "step": 781 + }, + { + "epoch": 0.5624887610142061, + "grad_norm": 7.251537052285208, + "learning_rate": 2.9793706456264397e-06, + "loss": 0.1637, + "step": 782 + }, + { + "epoch": 0.5632080561050171, + "grad_norm": 8.99624237501475, + "learning_rate": 2.9793143727649534e-06, + "loss": 0.1249, + "step": 783 + }, + { + "epoch": 0.563927351195828, + "grad_norm": 6.819524668479388, + "learning_rate": 2.9792580237901465e-06, + "loss": 0.075, + "step": 784 + }, + { + "epoch": 0.5646466462866391, + "grad_norm": 3.9832100822175374, + "learning_rate": 2.979201598704917e-06, + "loss": 0.1153, + "step": 785 + }, + { + "epoch": 0.5653659413774501, + "grad_norm": 9.265727486230267, + "learning_rate": 2.979145097512169e-06, + "loss": 0.2155, + "step": 786 + }, + { + "epoch": 0.5660852364682611, + "grad_norm": 5.731188520557445, + "learning_rate": 2.979088520214809e-06, + "loss": 0.171, + "step": 787 + }, + { + "epoch": 0.5668045315590721, + "grad_norm": 6.487421516949131, + "learning_rate": 2.9790318668157477e-06, + "loss": 0.166, + "step": 788 + }, + { + "epoch": 0.5675238266498831, + "grad_norm": 10.45535100630016, + "learning_rate": 2.9789751373179013e-06, + "loss": 0.2556, + "step": 789 + }, + { + "epoch": 0.5682431217406941, + "grad_norm": 5.256325166345312, + "learning_rate": 2.978918331724188e-06, + "loss": 0.0858, + "step": 790 + }, + { + "epoch": 0.5689624168315052, + "grad_norm": 5.737190551099546, + "learning_rate": 2.9788614500375296e-06, + "loss": 0.1384, + "step": 791 + }, + { + "epoch": 0.5696817119223161, + "grad_norm": 9.911757841096794, + "learning_rate": 2.978804492260854e-06, + "loss": 0.1805, + "step": 792 + }, + { + "epoch": 0.5704010070131271, + "grad_norm": 6.815015945738683, + "learning_rate": 2.9787474583970907e-06, + "loss": 0.1975, + "step": 793 + }, + { + "epoch": 0.5711203021039382, + "grad_norm": 10.433350453804477, + "learning_rate": 2.9786903484491753e-06, + "loss": 0.1525, + "step": 794 + }, + { + "epoch": 0.5718395971947492, + "grad_norm": 6.323978532782773, + "learning_rate": 2.9786331624200453e-06, + "loss": 0.1871, + "step": 795 + }, + { + "epoch": 0.5725588922855601, + "grad_norm": 7.396082338969181, + "learning_rate": 2.978575900312644e-06, + "loss": 0.2419, + "step": 796 + }, + { + "epoch": 0.5732781873763712, + "grad_norm": 8.032791141097622, + "learning_rate": 2.978518562129917e-06, + "loss": 0.1349, + "step": 797 + }, + { + "epoch": 0.5739974824671822, + "grad_norm": 6.510710205003883, + "learning_rate": 2.978461147874814e-06, + "loss": 0.0453, + "step": 798 + }, + { + "epoch": 0.5747167775579932, + "grad_norm": 6.8503162322583115, + "learning_rate": 2.97840365755029e-06, + "loss": 0.2867, + "step": 799 + }, + { + "epoch": 0.5754360726488041, + "grad_norm": 2.70073630363353, + "learning_rate": 2.9783460911593023e-06, + "loss": 0.0702, + "step": 800 + }, + { + "epoch": 0.5761553677396152, + "grad_norm": 2.9048919548146626, + "learning_rate": 2.9782884487048135e-06, + "loss": 0.0469, + "step": 801 + }, + { + "epoch": 0.5768746628304262, + "grad_norm": 5.166261801389186, + "learning_rate": 2.9782307301897888e-06, + "loss": 0.1059, + "step": 802 + }, + { + "epoch": 0.5775939579212371, + "grad_norm": 7.376611675269217, + "learning_rate": 2.978172935617198e-06, + "loss": 0.1832, + "step": 803 + }, + { + "epoch": 0.5783132530120482, + "grad_norm": 5.681383195694462, + "learning_rate": 2.9781150649900146e-06, + "loss": 0.1817, + "step": 804 + }, + { + "epoch": 0.5790325481028592, + "grad_norm": 3.723639290013787, + "learning_rate": 2.9780571183112164e-06, + "loss": 0.0189, + "step": 805 + }, + { + "epoch": 0.5797518431936702, + "grad_norm": 10.700512122340205, + "learning_rate": 2.977999095583785e-06, + "loss": 0.3174, + "step": 806 + }, + { + "epoch": 0.5804711382844812, + "grad_norm": 5.529048279432231, + "learning_rate": 2.9779409968107058e-06, + "loss": 0.1629, + "step": 807 + }, + { + "epoch": 0.5811904333752922, + "grad_norm": 5.842538322821677, + "learning_rate": 2.9778828219949674e-06, + "loss": 0.06, + "step": 808 + }, + { + "epoch": 0.5819097284661032, + "grad_norm": 9.259137428455585, + "learning_rate": 2.9778245711395643e-06, + "loss": 0.4163, + "step": 809 + }, + { + "epoch": 0.5826290235569143, + "grad_norm": 10.802316835435011, + "learning_rate": 2.977766244247492e-06, + "loss": 0.1261, + "step": 810 + }, + { + "epoch": 0.5833483186477252, + "grad_norm": 10.72991981273546, + "learning_rate": 2.9777078413217525e-06, + "loss": 0.2708, + "step": 811 + }, + { + "epoch": 0.5840676137385362, + "grad_norm": 7.103068474529098, + "learning_rate": 2.9776493623653505e-06, + "loss": 0.2065, + "step": 812 + }, + { + "epoch": 0.5847869088293473, + "grad_norm": 8.449802954978395, + "learning_rate": 2.9775908073812947e-06, + "loss": 0.2483, + "step": 813 + }, + { + "epoch": 0.5855062039201583, + "grad_norm": 7.945204474271803, + "learning_rate": 2.977532176372598e-06, + "loss": 0.1614, + "step": 814 + }, + { + "epoch": 0.5862254990109692, + "grad_norm": 6.46974024027023, + "learning_rate": 2.9774734693422776e-06, + "loss": 0.1636, + "step": 815 + }, + { + "epoch": 0.5869447941017802, + "grad_norm": 2.5219976288419357, + "learning_rate": 2.9774146862933535e-06, + "loss": 0.011, + "step": 816 + }, + { + "epoch": 0.5876640891925913, + "grad_norm": 7.828295205251189, + "learning_rate": 2.9773558272288497e-06, + "loss": 0.1442, + "step": 817 + }, + { + "epoch": 0.5883833842834023, + "grad_norm": 3.81562387033852, + "learning_rate": 2.977296892151796e-06, + "loss": 0.0766, + "step": 818 + }, + { + "epoch": 0.5891026793742132, + "grad_norm": 9.429459668067874, + "learning_rate": 2.9772378810652234e-06, + "loss": 0.1673, + "step": 819 + }, + { + "epoch": 0.5898219744650243, + "grad_norm": 5.713974117643566, + "learning_rate": 2.977178793972168e-06, + "loss": 0.0452, + "step": 820 + }, + { + "epoch": 0.5905412695558353, + "grad_norm": 6.732103452438126, + "learning_rate": 2.9771196308756717e-06, + "loss": 0.1426, + "step": 821 + }, + { + "epoch": 0.5912605646466463, + "grad_norm": 4.709989179152442, + "learning_rate": 2.9770603917787763e-06, + "loss": 0.1111, + "step": 822 + }, + { + "epoch": 0.5919798597374573, + "grad_norm": 6.1052474311868155, + "learning_rate": 2.9770010766845315e-06, + "loss": 0.0889, + "step": 823 + }, + { + "epoch": 0.5926991548282683, + "grad_norm": 5.044198680883185, + "learning_rate": 2.9769416855959886e-06, + "loss": 0.1284, + "step": 824 + }, + { + "epoch": 0.5934184499190793, + "grad_norm": 4.620031748227179, + "learning_rate": 2.9768822185162034e-06, + "loss": 0.1813, + "step": 825 + }, + { + "epoch": 0.5941377450098904, + "grad_norm": 4.955822576738682, + "learning_rate": 2.9768226754482352e-06, + "loss": 0.0611, + "step": 826 + }, + { + "epoch": 0.5948570401007013, + "grad_norm": 7.06242278617012, + "learning_rate": 2.9767630563951475e-06, + "loss": 0.1843, + "step": 827 + }, + { + "epoch": 0.5955763351915123, + "grad_norm": 5.004874387211047, + "learning_rate": 2.9767033613600085e-06, + "loss": 0.1357, + "step": 828 + }, + { + "epoch": 0.5962956302823234, + "grad_norm": 7.795250779728666, + "learning_rate": 2.9766435903458897e-06, + "loss": 0.2441, + "step": 829 + }, + { + "epoch": 0.5970149253731343, + "grad_norm": 4.433821476306185, + "learning_rate": 2.9765837433558652e-06, + "loss": 0.1733, + "step": 830 + }, + { + "epoch": 0.5977342204639453, + "grad_norm": 3.720381732391177, + "learning_rate": 2.9765238203930155e-06, + "loss": 0.1126, + "step": 831 + }, + { + "epoch": 0.5984535155547563, + "grad_norm": 8.256773045292633, + "learning_rate": 2.976463821460423e-06, + "loss": 0.1267, + "step": 832 + }, + { + "epoch": 0.5991728106455674, + "grad_norm": 6.335327053294731, + "learning_rate": 2.9764037465611757e-06, + "loss": 0.1256, + "step": 833 + }, + { + "epoch": 0.5998921057363783, + "grad_norm": 5.176784166666724, + "learning_rate": 2.9763435956983635e-06, + "loss": 0.028, + "step": 834 + }, + { + "epoch": 0.6006114008271893, + "grad_norm": 5.120191727793925, + "learning_rate": 2.9762833688750816e-06, + "loss": 0.0464, + "step": 835 + }, + { + "epoch": 0.6013306959180004, + "grad_norm": 6.275983829145228, + "learning_rate": 2.976223066094429e-06, + "loss": 0.1378, + "step": 836 + }, + { + "epoch": 0.6020499910088114, + "grad_norm": 8.702513434315492, + "learning_rate": 2.976162687359508e-06, + "loss": 0.1712, + "step": 837 + }, + { + "epoch": 0.6027692860996223, + "grad_norm": 8.700678119525492, + "learning_rate": 2.9761022326734253e-06, + "loss": 0.0744, + "step": 838 + }, + { + "epoch": 0.6034885811904334, + "grad_norm": 9.153608113492444, + "learning_rate": 2.976041702039292e-06, + "loss": 0.3031, + "step": 839 + }, + { + "epoch": 0.6042078762812444, + "grad_norm": 10.418573599689886, + "learning_rate": 2.975981095460222e-06, + "loss": 0.4991, + "step": 840 + }, + { + "epoch": 0.6049271713720554, + "grad_norm": 7.871958835957954, + "learning_rate": 2.975920412939333e-06, + "loss": 0.357, + "step": 841 + }, + { + "epoch": 0.6056464664628664, + "grad_norm": 5.115200490467494, + "learning_rate": 2.975859654479748e-06, + "loss": 0.2311, + "step": 842 + }, + { + "epoch": 0.6063657615536774, + "grad_norm": 5.9669621928778565, + "learning_rate": 2.9757988200845935e-06, + "loss": 0.1251, + "step": 843 + }, + { + "epoch": 0.6070850566444884, + "grad_norm": 6.396034729264032, + "learning_rate": 2.9757379097569988e-06, + "loss": 0.1222, + "step": 844 + }, + { + "epoch": 0.6078043517352995, + "grad_norm": 2.6562509712514792, + "learning_rate": 2.975676923500098e-06, + "loss": 0.0217, + "step": 845 + }, + { + "epoch": 0.6085236468261104, + "grad_norm": 7.02411881784209, + "learning_rate": 2.975615861317029e-06, + "loss": 0.2655, + "step": 846 + }, + { + "epoch": 0.6092429419169214, + "grad_norm": 4.529085862499067, + "learning_rate": 2.9755547232109334e-06, + "loss": 0.0962, + "step": 847 + }, + { + "epoch": 0.6099622370077324, + "grad_norm": 4.750791297610307, + "learning_rate": 2.9754935091849575e-06, + "loss": 0.0283, + "step": 848 + }, + { + "epoch": 0.6106815320985435, + "grad_norm": 8.73875925573561, + "learning_rate": 2.9754322192422497e-06, + "loss": 0.281, + "step": 849 + }, + { + "epoch": 0.6114008271893544, + "grad_norm": 1.627604198611702, + "learning_rate": 2.9753708533859654e-06, + "loss": 0.0255, + "step": 850 + }, + { + "epoch": 0.6121201222801654, + "grad_norm": 3.921026272107377, + "learning_rate": 2.9753094116192595e-06, + "loss": 0.0459, + "step": 851 + }, + { + "epoch": 0.6128394173709765, + "grad_norm": 5.204394312262345, + "learning_rate": 2.975247893945295e-06, + "loss": 0.2747, + "step": 852 + }, + { + "epoch": 0.6135587124617875, + "grad_norm": 5.117453766528016, + "learning_rate": 2.9751863003672364e-06, + "loss": 0.1247, + "step": 853 + }, + { + "epoch": 0.6142780075525984, + "grad_norm": 2.2165642479184364, + "learning_rate": 2.9751246308882537e-06, + "loss": 0.0085, + "step": 854 + }, + { + "epoch": 0.6149973026434095, + "grad_norm": 9.728527315951023, + "learning_rate": 2.975062885511519e-06, + "loss": 0.3202, + "step": 855 + }, + { + "epoch": 0.6157165977342205, + "grad_norm": 7.023117301343883, + "learning_rate": 2.975001064240209e-06, + "loss": 0.0763, + "step": 856 + }, + { + "epoch": 0.6164358928250314, + "grad_norm": 11.218005745567654, + "learning_rate": 2.9749391670775056e-06, + "loss": 0.5647, + "step": 857 + }, + { + "epoch": 0.6171551879158425, + "grad_norm": 6.146475892297033, + "learning_rate": 2.9748771940265923e-06, + "loss": 0.1852, + "step": 858 + }, + { + "epoch": 0.6178744830066535, + "grad_norm": 6.063987943425242, + "learning_rate": 2.9748151450906584e-06, + "loss": 0.0667, + "step": 859 + }, + { + "epoch": 0.6185937780974645, + "grad_norm": 12.564102885214416, + "learning_rate": 2.9747530202728968e-06, + "loss": 0.1499, + "step": 860 + }, + { + "epoch": 0.6193130731882754, + "grad_norm": 5.217187746869808, + "learning_rate": 2.974690819576503e-06, + "loss": 0.2194, + "step": 861 + }, + { + "epoch": 0.6200323682790865, + "grad_norm": 8.542969665607426, + "learning_rate": 2.974628543004678e-06, + "loss": 0.2519, + "step": 862 + }, + { + "epoch": 0.6207516633698975, + "grad_norm": 6.71096918047829, + "learning_rate": 2.9745661905606255e-06, + "loss": 0.1985, + "step": 863 + }, + { + "epoch": 0.6214709584607085, + "grad_norm": 8.426627127116994, + "learning_rate": 2.9745037622475542e-06, + "loss": 0.2507, + "step": 864 + }, + { + "epoch": 0.6221902535515195, + "grad_norm": 9.730786077324774, + "learning_rate": 2.9744412580686764e-06, + "loss": 0.3127, + "step": 865 + }, + { + "epoch": 0.6229095486423305, + "grad_norm": 5.119390609090124, + "learning_rate": 2.974378678027207e-06, + "loss": 0.1108, + "step": 866 + }, + { + "epoch": 0.6236288437331415, + "grad_norm": 8.018984599126975, + "learning_rate": 2.9743160221263667e-06, + "loss": 0.1288, + "step": 867 + }, + { + "epoch": 0.6243481388239526, + "grad_norm": 4.840556373181076, + "learning_rate": 2.974253290369379e-06, + "loss": 0.108, + "step": 868 + }, + { + "epoch": 0.6250674339147635, + "grad_norm": 6.134456768729974, + "learning_rate": 2.974190482759471e-06, + "loss": 0.1538, + "step": 869 + }, + { + "epoch": 0.6257867290055745, + "grad_norm": 4.498009589164801, + "learning_rate": 2.974127599299875e-06, + "loss": 0.185, + "step": 870 + }, + { + "epoch": 0.6265060240963856, + "grad_norm": 4.333757459896507, + "learning_rate": 2.9740646399938265e-06, + "loss": 0.0424, + "step": 871 + }, + { + "epoch": 0.6272253191871966, + "grad_norm": 9.4075492372385, + "learning_rate": 2.974001604844565e-06, + "loss": 0.3262, + "step": 872 + }, + { + "epoch": 0.6279446142780075, + "grad_norm": 7.915385965207062, + "learning_rate": 2.973938493855333e-06, + "loss": 0.279, + "step": 873 + }, + { + "epoch": 0.6286639093688186, + "grad_norm": 7.425017485693498, + "learning_rate": 2.973875307029378e-06, + "loss": 0.1751, + "step": 874 + }, + { + "epoch": 0.6293832044596296, + "grad_norm": 5.25227791466232, + "learning_rate": 2.973812044369951e-06, + "loss": 0.1113, + "step": 875 + }, + { + "epoch": 0.6301024995504406, + "grad_norm": 8.690057578540404, + "learning_rate": 2.9737487058803073e-06, + "loss": 0.1286, + "step": 876 + }, + { + "epoch": 0.6308217946412515, + "grad_norm": 3.788961434337117, + "learning_rate": 2.9736852915637056e-06, + "loss": 0.1133, + "step": 877 + }, + { + "epoch": 0.6315410897320626, + "grad_norm": 6.688914066306362, + "learning_rate": 2.973621801423409e-06, + "loss": 0.1359, + "step": 878 + }, + { + "epoch": 0.6322603848228736, + "grad_norm": 5.650706561607885, + "learning_rate": 2.9735582354626833e-06, + "loss": 0.1359, + "step": 879 + }, + { + "epoch": 0.6329796799136845, + "grad_norm": 5.536085481886316, + "learning_rate": 2.9734945936848e-06, + "loss": 0.1386, + "step": 880 + }, + { + "epoch": 0.6336989750044956, + "grad_norm": 6.869733995664741, + "learning_rate": 2.9734308760930334e-06, + "loss": 0.1221, + "step": 881 + }, + { + "epoch": 0.6344182700953066, + "grad_norm": 6.164015336665599, + "learning_rate": 2.973367082690661e-06, + "loss": 0.2132, + "step": 882 + }, + { + "epoch": 0.6351375651861176, + "grad_norm": 4.285198217814613, + "learning_rate": 2.973303213480966e-06, + "loss": 0.0337, + "step": 883 + }, + { + "epoch": 0.6358568602769286, + "grad_norm": 13.904838648691733, + "learning_rate": 2.973239268467235e-06, + "loss": 0.0781, + "step": 884 + }, + { + "epoch": 0.6365761553677396, + "grad_norm": 6.666734623160479, + "learning_rate": 2.973175247652757e-06, + "loss": 0.2278, + "step": 885 + }, + { + "epoch": 0.6372954504585506, + "grad_norm": 6.62265366931964, + "learning_rate": 2.9731111510408264e-06, + "loss": 0.1156, + "step": 886 + }, + { + "epoch": 0.6380147455493617, + "grad_norm": 9.471533073267807, + "learning_rate": 2.973046978634741e-06, + "loss": 0.2777, + "step": 887 + }, + { + "epoch": 0.6387340406401726, + "grad_norm": 4.824796222940127, + "learning_rate": 2.972982730437803e-06, + "loss": 0.1629, + "step": 888 + }, + { + "epoch": 0.6394533357309836, + "grad_norm": 5.367577098199019, + "learning_rate": 2.9729184064533176e-06, + "loss": 0.2324, + "step": 889 + }, + { + "epoch": 0.6401726308217947, + "grad_norm": 10.36031445648042, + "learning_rate": 2.9728540066845947e-06, + "loss": 0.1892, + "step": 890 + }, + { + "epoch": 0.6408919259126057, + "grad_norm": 3.9358579549312864, + "learning_rate": 2.9727895311349476e-06, + "loss": 0.1133, + "step": 891 + }, + { + "epoch": 0.6416112210034166, + "grad_norm": 6.990521432262684, + "learning_rate": 2.9727249798076934e-06, + "loss": 0.2543, + "step": 892 + }, + { + "epoch": 0.6423305160942276, + "grad_norm": 7.700750845499684, + "learning_rate": 2.9726603527061537e-06, + "loss": 0.1537, + "step": 893 + }, + { + "epoch": 0.6430498111850387, + "grad_norm": 7.9271304139400955, + "learning_rate": 2.9725956498336544e-06, + "loss": 0.2247, + "step": 894 + }, + { + "epoch": 0.6437691062758497, + "grad_norm": 5.8354026064268085, + "learning_rate": 2.972530871193523e-06, + "loss": 0.1517, + "step": 895 + }, + { + "epoch": 0.6444884013666606, + "grad_norm": 6.865431154496454, + "learning_rate": 2.9724660167890937e-06, + "loss": 0.1779, + "step": 896 + }, + { + "epoch": 0.6452076964574717, + "grad_norm": 6.300216577933797, + "learning_rate": 2.9724010866237034e-06, + "loss": 0.0863, + "step": 897 + }, + { + "epoch": 0.6459269915482827, + "grad_norm": 8.92952838354427, + "learning_rate": 2.972336080700692e-06, + "loss": 0.3282, + "step": 898 + }, + { + "epoch": 0.6466462866390937, + "grad_norm": 6.962271718972027, + "learning_rate": 2.972270999023405e-06, + "loss": 0.1949, + "step": 899 + }, + { + "epoch": 0.6473655817299047, + "grad_norm": 8.933220531831731, + "learning_rate": 2.97220584159519e-06, + "loss": 0.1558, + "step": 900 + }, + { + "epoch": 0.6480848768207157, + "grad_norm": 5.447335503944896, + "learning_rate": 2.9721406084194e-06, + "loss": 0.0331, + "step": 901 + }, + { + "epoch": 0.6488041719115267, + "grad_norm": 6.005076152896434, + "learning_rate": 2.972075299499392e-06, + "loss": 0.1277, + "step": 902 + }, + { + "epoch": 0.6495234670023378, + "grad_norm": 4.884670744944036, + "learning_rate": 2.972009914838526e-06, + "loss": 0.135, + "step": 903 + }, + { + "epoch": 0.6502427620931487, + "grad_norm": 7.780318510269028, + "learning_rate": 2.9719444544401653e-06, + "loss": 0.2154, + "step": 904 + }, + { + "epoch": 0.6509620571839597, + "grad_norm": 6.667906910236401, + "learning_rate": 2.9718789183076786e-06, + "loss": 0.2764, + "step": 905 + }, + { + "epoch": 0.6516813522747708, + "grad_norm": 3.5288210181500297, + "learning_rate": 2.971813306444438e-06, + "loss": 0.1276, + "step": 906 + }, + { + "epoch": 0.6524006473655817, + "grad_norm": 4.6994708361789534, + "learning_rate": 2.9717476188538185e-06, + "loss": 0.2257, + "step": 907 + }, + { + "epoch": 0.6531199424563927, + "grad_norm": 8.552437912521276, + "learning_rate": 2.971681855539201e-06, + "loss": 0.2624, + "step": 908 + }, + { + "epoch": 0.6538392375472037, + "grad_norm": 5.324840877686902, + "learning_rate": 2.9716160165039682e-06, + "loss": 0.1073, + "step": 909 + }, + { + "epoch": 0.6545585326380148, + "grad_norm": 6.083879695968263, + "learning_rate": 2.971550101751508e-06, + "loss": 0.0718, + "step": 910 + }, + { + "epoch": 0.6552778277288257, + "grad_norm": 6.576718341399474, + "learning_rate": 2.9714841112852124e-06, + "loss": 0.3142, + "step": 911 + }, + { + "epoch": 0.6559971228196367, + "grad_norm": 7.884330445452147, + "learning_rate": 2.971418045108476e-06, + "loss": 0.0916, + "step": 912 + }, + { + "epoch": 0.6567164179104478, + "grad_norm": 4.9547657609957225, + "learning_rate": 2.9713519032246977e-06, + "loss": 0.0619, + "step": 913 + }, + { + "epoch": 0.6574357130012588, + "grad_norm": 7.268345932188917, + "learning_rate": 2.9712856856372812e-06, + "loss": 0.2041, + "step": 914 + }, + { + "epoch": 0.6581550080920697, + "grad_norm": 9.412585766618037, + "learning_rate": 2.971219392349634e-06, + "loss": 0.1887, + "step": 915 + }, + { + "epoch": 0.6588743031828808, + "grad_norm": 5.810528225012237, + "learning_rate": 2.9711530233651656e-06, + "loss": 0.1326, + "step": 916 + }, + { + "epoch": 0.6595935982736918, + "grad_norm": 4.490857380383742, + "learning_rate": 2.971086578687292e-06, + "loss": 0.1755, + "step": 917 + }, + { + "epoch": 0.6603128933645028, + "grad_norm": 8.427661227436248, + "learning_rate": 2.9710200583194315e-06, + "loss": 0.2563, + "step": 918 + }, + { + "epoch": 0.6610321884553138, + "grad_norm": 6.352343960941773, + "learning_rate": 2.970953462265007e-06, + "loss": 0.2035, + "step": 919 + }, + { + "epoch": 0.6617514835461248, + "grad_norm": 5.8647878017999275, + "learning_rate": 2.9708867905274444e-06, + "loss": 0.2122, + "step": 920 + }, + { + "epoch": 0.6624707786369358, + "grad_norm": 6.084389482201497, + "learning_rate": 2.9708200431101744e-06, + "loss": 0.213, + "step": 921 + }, + { + "epoch": 0.6631900737277469, + "grad_norm": 6.170068691394691, + "learning_rate": 2.970753220016631e-06, + "loss": 0.2141, + "step": 922 + }, + { + "epoch": 0.6639093688185578, + "grad_norm": 4.002915223438293, + "learning_rate": 2.970686321250253e-06, + "loss": 0.2609, + "step": 923 + }, + { + "epoch": 0.6646286639093688, + "grad_norm": 7.4291739362946645, + "learning_rate": 2.970619346814482e-06, + "loss": 0.1457, + "step": 924 + }, + { + "epoch": 0.6653479590001798, + "grad_norm": 10.923704001605786, + "learning_rate": 2.970552296712764e-06, + "loss": 0.3077, + "step": 925 + }, + { + "epoch": 0.6660672540909909, + "grad_norm": 4.567310973115804, + "learning_rate": 2.970485170948549e-06, + "loss": 0.1155, + "step": 926 + }, + { + "epoch": 0.6667865491818018, + "grad_norm": 15.33365099139537, + "learning_rate": 2.9704179695252903e-06, + "loss": 0.0732, + "step": 927 + }, + { + "epoch": 0.6675058442726128, + "grad_norm": 4.055209818056239, + "learning_rate": 2.970350692446446e-06, + "loss": 0.1111, + "step": 928 + }, + { + "epoch": 0.6682251393634239, + "grad_norm": 8.32518136007484, + "learning_rate": 2.9702833397154773e-06, + "loss": 0.186, + "step": 929 + }, + { + "epoch": 0.6689444344542349, + "grad_norm": 6.070507782562587, + "learning_rate": 2.9702159113358496e-06, + "loss": 0.3945, + "step": 930 + }, + { + "epoch": 0.6696637295450458, + "grad_norm": 10.56763916855245, + "learning_rate": 2.970148407311033e-06, + "loss": 0.1321, + "step": 931 + }, + { + "epoch": 0.6703830246358569, + "grad_norm": 7.362035488523589, + "learning_rate": 2.9700808276444995e-06, + "loss": 0.0956, + "step": 932 + }, + { + "epoch": 0.6711023197266679, + "grad_norm": 2.405992336768348, + "learning_rate": 2.970013172339727e-06, + "loss": 0.0675, + "step": 933 + }, + { + "epoch": 0.6718216148174788, + "grad_norm": 6.233934300851908, + "learning_rate": 2.969945441400196e-06, + "loss": 0.156, + "step": 934 + }, + { + "epoch": 0.6725409099082899, + "grad_norm": 5.287259405651, + "learning_rate": 2.969877634829392e-06, + "loss": 0.0951, + "step": 935 + }, + { + "epoch": 0.6732602049991009, + "grad_norm": 8.737951734122264, + "learning_rate": 2.9698097526308034e-06, + "loss": 0.4278, + "step": 936 + }, + { + "epoch": 0.6739795000899119, + "grad_norm": 11.565358945609637, + "learning_rate": 2.969741794807923e-06, + "loss": 0.1419, + "step": 937 + }, + { + "epoch": 0.6746987951807228, + "grad_norm": 13.482488887119393, + "learning_rate": 2.9696737613642467e-06, + "loss": 0.1875, + "step": 938 + }, + { + "epoch": 0.6754180902715339, + "grad_norm": 4.447649668678088, + "learning_rate": 2.9696056523032756e-06, + "loss": 0.0627, + "step": 939 + }, + { + "epoch": 0.6761373853623449, + "grad_norm": 8.016031755063839, + "learning_rate": 2.969537467628514e-06, + "loss": 0.1417, + "step": 940 + }, + { + "epoch": 0.676856680453156, + "grad_norm": 7.756605673502039, + "learning_rate": 2.9694692073434696e-06, + "loss": 0.0826, + "step": 941 + }, + { + "epoch": 0.6775759755439669, + "grad_norm": 9.046690224668627, + "learning_rate": 2.9694008714516555e-06, + "loss": 0.3833, + "step": 942 + }, + { + "epoch": 0.6782952706347779, + "grad_norm": 7.484253094813111, + "learning_rate": 2.9693324599565867e-06, + "loss": 0.0641, + "step": 943 + }, + { + "epoch": 0.6790145657255889, + "grad_norm": 6.483536334863294, + "learning_rate": 2.969263972861784e-06, + "loss": 0.2174, + "step": 944 + }, + { + "epoch": 0.6797338608164, + "grad_norm": 7.901059115931261, + "learning_rate": 2.9691954101707703e-06, + "loss": 0.208, + "step": 945 + }, + { + "epoch": 0.6804531559072109, + "grad_norm": 8.175716995140329, + "learning_rate": 2.9691267718870738e-06, + "loss": 0.2317, + "step": 946 + }, + { + "epoch": 0.6811724509980219, + "grad_norm": 7.685723437258521, + "learning_rate": 2.9690580580142257e-06, + "loss": 0.1976, + "step": 947 + }, + { + "epoch": 0.681891746088833, + "grad_norm": 4.707952007348748, + "learning_rate": 2.9689892685557615e-06, + "loss": 0.2099, + "step": 948 + }, + { + "epoch": 0.682611041179644, + "grad_norm": 8.336041941938515, + "learning_rate": 2.9689204035152214e-06, + "loss": 0.1955, + "step": 949 + }, + { + "epoch": 0.6833303362704549, + "grad_norm": 7.2925717327929, + "learning_rate": 2.9688514628961473e-06, + "loss": 0.1584, + "step": 950 + }, + { + "epoch": 0.684049631361266, + "grad_norm": 6.065612648113627, + "learning_rate": 2.9687824467020873e-06, + "loss": 0.0913, + "step": 951 + }, + { + "epoch": 0.684768926452077, + "grad_norm": 6.032179881947198, + "learning_rate": 2.9687133549365923e-06, + "loss": 0.1195, + "step": 952 + }, + { + "epoch": 0.685488221542888, + "grad_norm": 6.738343429863798, + "learning_rate": 2.9686441876032163e-06, + "loss": 0.3167, + "step": 953 + }, + { + "epoch": 0.6862075166336989, + "grad_norm": 5.740668653914783, + "learning_rate": 2.968574944705519e-06, + "loss": 0.1593, + "step": 954 + }, + { + "epoch": 0.68692681172451, + "grad_norm": 4.059127673729532, + "learning_rate": 2.968505626247063e-06, + "loss": 0.0734, + "step": 955 + }, + { + "epoch": 0.687646106815321, + "grad_norm": 8.467989583057847, + "learning_rate": 2.968436232231414e-06, + "loss": 0.217, + "step": 956 + }, + { + "epoch": 0.688365401906132, + "grad_norm": 4.754709716730067, + "learning_rate": 2.9683667626621436e-06, + "loss": 0.1607, + "step": 957 + }, + { + "epoch": 0.689084696996943, + "grad_norm": 6.492539802730643, + "learning_rate": 2.9682972175428257e-06, + "loss": 0.1135, + "step": 958 + }, + { + "epoch": 0.689803992087754, + "grad_norm": 7.037692384877635, + "learning_rate": 2.9682275968770386e-06, + "loss": 0.2422, + "step": 959 + }, + { + "epoch": 0.690523287178565, + "grad_norm": 6.151695768339403, + "learning_rate": 2.9681579006683636e-06, + "loss": 0.1268, + "step": 960 + }, + { + "epoch": 0.691242582269376, + "grad_norm": 7.652421229038361, + "learning_rate": 2.9680881289203874e-06, + "loss": 0.3431, + "step": 961 + }, + { + "epoch": 0.691961877360187, + "grad_norm": 3.764330521601031, + "learning_rate": 2.9680182816367006e-06, + "loss": 0.1311, + "step": 962 + }, + { + "epoch": 0.692681172450998, + "grad_norm": 4.229019754615401, + "learning_rate": 2.9679483588208955e-06, + "loss": 0.1396, + "step": 963 + }, + { + "epoch": 0.6934004675418091, + "grad_norm": 5.224071629996026, + "learning_rate": 2.9678783604765707e-06, + "loss": 0.1206, + "step": 964 + }, + { + "epoch": 0.69411976263262, + "grad_norm": 4.011431670303693, + "learning_rate": 2.9678082866073273e-06, + "loss": 0.0713, + "step": 965 + }, + { + "epoch": 0.694839057723431, + "grad_norm": 6.126105544714853, + "learning_rate": 2.9677381372167707e-06, + "loss": 0.1258, + "step": 966 + }, + { + "epoch": 0.695558352814242, + "grad_norm": 7.805957238731541, + "learning_rate": 2.967667912308511e-06, + "loss": 0.1286, + "step": 967 + }, + { + "epoch": 0.6962776479050531, + "grad_norm": 6.460562531551328, + "learning_rate": 2.9675976118861605e-06, + "loss": 0.0887, + "step": 968 + }, + { + "epoch": 0.696996942995864, + "grad_norm": 8.118452439188117, + "learning_rate": 2.967527235953337e-06, + "loss": 0.1177, + "step": 969 + }, + { + "epoch": 0.697716238086675, + "grad_norm": 5.769810171818604, + "learning_rate": 2.96745678451366e-06, + "loss": 0.1082, + "step": 970 + }, + { + "epoch": 0.6984355331774861, + "grad_norm": 4.401102495960162, + "learning_rate": 2.9673862575707564e-06, + "loss": 0.0795, + "step": 971 + }, + { + "epoch": 0.6991548282682971, + "grad_norm": 8.63568154024474, + "learning_rate": 2.9673156551282536e-06, + "loss": 0.127, + "step": 972 + }, + { + "epoch": 0.699874123359108, + "grad_norm": 6.455892407809294, + "learning_rate": 2.9672449771897842e-06, + "loss": 0.2006, + "step": 973 + }, + { + "epoch": 0.7005934184499191, + "grad_norm": 3.5765667942005503, + "learning_rate": 2.967174223758986e-06, + "loss": 0.1018, + "step": 974 + }, + { + "epoch": 0.7013127135407301, + "grad_norm": 3.334368291701813, + "learning_rate": 2.9671033948394975e-06, + "loss": 0.0491, + "step": 975 + }, + { + "epoch": 0.7020320086315411, + "grad_norm": 8.253089659471735, + "learning_rate": 2.967032490434964e-06, + "loss": 0.2462, + "step": 976 + }, + { + "epoch": 0.7027513037223521, + "grad_norm": 2.8823983697196263, + "learning_rate": 2.966961510549034e-06, + "loss": 0.0637, + "step": 977 + }, + { + "epoch": 0.7034705988131631, + "grad_norm": 34.893804146111194, + "learning_rate": 2.966890455185359e-06, + "loss": 0.3012, + "step": 978 + }, + { + "epoch": 0.7041898939039741, + "grad_norm": 19.958986078167154, + "learning_rate": 2.966819324347595e-06, + "loss": 0.1786, + "step": 979 + }, + { + "epoch": 0.7049091889947852, + "grad_norm": 8.434388695810744, + "learning_rate": 2.966748118039402e-06, + "loss": 0.1356, + "step": 980 + }, + { + "epoch": 0.7056284840855961, + "grad_norm": 7.3825319122220785, + "learning_rate": 2.966676836264443e-06, + "loss": 0.2168, + "step": 981 + }, + { + "epoch": 0.7063477791764071, + "grad_norm": 6.90429604963317, + "learning_rate": 2.9666054790263865e-06, + "loss": 0.118, + "step": 982 + }, + { + "epoch": 0.7070670742672182, + "grad_norm": 6.50748962531561, + "learning_rate": 2.9665340463289037e-06, + "loss": 0.3254, + "step": 983 + }, + { + "epoch": 0.7077863693580291, + "grad_norm": 4.983237971721256, + "learning_rate": 2.9664625381756693e-06, + "loss": 0.2475, + "step": 984 + }, + { + "epoch": 0.7085056644488401, + "grad_norm": 9.890964593091509, + "learning_rate": 2.9663909545703633e-06, + "loss": 0.226, + "step": 985 + }, + { + "epoch": 0.7092249595396511, + "grad_norm": 4.4264168456726605, + "learning_rate": 2.966319295516669e-06, + "loss": 0.1635, + "step": 986 + }, + { + "epoch": 0.7099442546304622, + "grad_norm": 5.838681578080727, + "learning_rate": 2.9662475610182723e-06, + "loss": 0.0493, + "step": 987 + }, + { + "epoch": 0.7106635497212731, + "grad_norm": 7.884719276335242, + "learning_rate": 2.9661757510788646e-06, + "loss": 0.1754, + "step": 988 + }, + { + "epoch": 0.7113828448120841, + "grad_norm": 6.507672641984555, + "learning_rate": 2.966103865702141e-06, + "loss": 0.0645, + "step": 989 + }, + { + "epoch": 0.7121021399028952, + "grad_norm": 7.225249366538003, + "learning_rate": 2.9660319048917997e-06, + "loss": 0.1235, + "step": 990 + }, + { + "epoch": 0.7128214349937062, + "grad_norm": 3.7333508759405913, + "learning_rate": 2.9659598686515434e-06, + "loss": 0.122, + "step": 991 + }, + { + "epoch": 0.7135407300845171, + "grad_norm": 10.042644778962964, + "learning_rate": 2.9658877569850785e-06, + "loss": 0.1271, + "step": 992 + }, + { + "epoch": 0.7142600251753282, + "grad_norm": 7.766648915957607, + "learning_rate": 2.965815569896115e-06, + "loss": 0.0961, + "step": 993 + }, + { + "epoch": 0.7149793202661392, + "grad_norm": 5.279417321782498, + "learning_rate": 2.965743307388368e-06, + "loss": 0.1893, + "step": 994 + }, + { + "epoch": 0.7156986153569502, + "grad_norm": 7.560644437226907, + "learning_rate": 2.965670969465554e-06, + "loss": 0.2037, + "step": 995 + }, + { + "epoch": 0.7164179104477612, + "grad_norm": 6.114245524146761, + "learning_rate": 2.965598556131396e-06, + "loss": 0.0396, + "step": 996 + }, + { + "epoch": 0.7171372055385722, + "grad_norm": 6.2985475339173895, + "learning_rate": 2.9655260673896195e-06, + "loss": 0.2074, + "step": 997 + }, + { + "epoch": 0.7178565006293832, + "grad_norm": 5.282943824069809, + "learning_rate": 2.965453503243954e-06, + "loss": 0.1521, + "step": 998 + }, + { + "epoch": 0.7185757957201943, + "grad_norm": 8.014710357070566, + "learning_rate": 2.9653808636981336e-06, + "loss": 0.2785, + "step": 999 + }, + { + "epoch": 0.7192950908110052, + "grad_norm": 8.650382345832208, + "learning_rate": 2.9653081487558953e-06, + "loss": 0.3086, + "step": 1000 + }, + { + "epoch": 0.7200143859018162, + "grad_norm": 5.003521045734504, + "learning_rate": 2.96523535842098e-06, + "loss": 0.0893, + "step": 1001 + }, + { + "epoch": 0.7207336809926272, + "grad_norm": 5.751385819014117, + "learning_rate": 2.965162492697134e-06, + "loss": 0.2563, + "step": 1002 + }, + { + "epoch": 0.7214529760834383, + "grad_norm": 7.441038980787459, + "learning_rate": 2.9650895515881057e-06, + "loss": 0.1702, + "step": 1003 + }, + { + "epoch": 0.7221722711742492, + "grad_norm": 4.836643831737609, + "learning_rate": 2.9650165350976476e-06, + "loss": 0.1199, + "step": 1004 + }, + { + "epoch": 0.7228915662650602, + "grad_norm": 4.146165859090649, + "learning_rate": 2.9649434432295175e-06, + "loss": 0.0687, + "step": 1005 + }, + { + "epoch": 0.7236108613558713, + "grad_norm": 7.924222140472695, + "learning_rate": 2.9648702759874753e-06, + "loss": 0.1596, + "step": 1006 + }, + { + "epoch": 0.7243301564466823, + "grad_norm": 5.4462806625069975, + "learning_rate": 2.964797033375286e-06, + "loss": 0.2539, + "step": 1007 + }, + { + "epoch": 0.7250494515374932, + "grad_norm": 6.551134818408238, + "learning_rate": 2.9647237153967175e-06, + "loss": 0.1955, + "step": 1008 + }, + { + "epoch": 0.7257687466283043, + "grad_norm": 8.860054031426118, + "learning_rate": 2.964650322055543e-06, + "loss": 0.3214, + "step": 1009 + }, + { + "epoch": 0.7264880417191153, + "grad_norm": 6.168108824487009, + "learning_rate": 2.964576853355539e-06, + "loss": 0.0882, + "step": 1010 + }, + { + "epoch": 0.7272073368099262, + "grad_norm": 5.965482796743988, + "learning_rate": 2.964503309300484e-06, + "loss": 0.1507, + "step": 1011 + }, + { + "epoch": 0.7279266319007373, + "grad_norm": 6.853415871889261, + "learning_rate": 2.964429689894163e-06, + "loss": 0.1678, + "step": 1012 + }, + { + "epoch": 0.7286459269915483, + "grad_norm": 7.8826983989593975, + "learning_rate": 2.964355995140364e-06, + "loss": 0.2033, + "step": 1013 + }, + { + "epoch": 0.7293652220823593, + "grad_norm": 2.310130006174262, + "learning_rate": 2.9642822250428784e-06, + "loss": 0.0551, + "step": 1014 + }, + { + "epoch": 0.7300845171731702, + "grad_norm": 6.319449571703531, + "learning_rate": 2.964208379605502e-06, + "loss": 0.3185, + "step": 1015 + }, + { + "epoch": 0.7308038122639813, + "grad_norm": 3.9879871398239364, + "learning_rate": 2.9641344588320337e-06, + "loss": 0.132, + "step": 1016 + }, + { + "epoch": 0.7315231073547923, + "grad_norm": 4.5887037458182105, + "learning_rate": 2.964060462726278e-06, + "loss": 0.0406, + "step": 1017 + }, + { + "epoch": 0.7322424024456033, + "grad_norm": 9.666589232054042, + "learning_rate": 2.9639863912920406e-06, + "loss": 0.1545, + "step": 1018 + }, + { + "epoch": 0.7329616975364143, + "grad_norm": 11.836243335873432, + "learning_rate": 2.9639122445331336e-06, + "loss": 0.135, + "step": 1019 + }, + { + "epoch": 0.7336809926272253, + "grad_norm": 5.42392486236328, + "learning_rate": 2.963838022453372e-06, + "loss": 0.164, + "step": 1020 + }, + { + "epoch": 0.7344002877180363, + "grad_norm": 5.274419931289659, + "learning_rate": 2.9637637250565745e-06, + "loss": 0.2822, + "step": 1021 + }, + { + "epoch": 0.7351195828088474, + "grad_norm": 9.797947072139278, + "learning_rate": 2.963689352346564e-06, + "loss": 0.5928, + "step": 1022 + }, + { + "epoch": 0.7358388778996583, + "grad_norm": 8.378388821768572, + "learning_rate": 2.9636149043271666e-06, + "loss": 0.2254, + "step": 1023 + }, + { + "epoch": 0.7365581729904693, + "grad_norm": 8.304864011786217, + "learning_rate": 2.963540381002213e-06, + "loss": 0.2131, + "step": 1024 + }, + { + "epoch": 0.7372774680812804, + "grad_norm": 5.377043117794858, + "learning_rate": 2.9634657823755374e-06, + "loss": 0.1024, + "step": 1025 + }, + { + "epoch": 0.7379967631720914, + "grad_norm": 6.220054664975505, + "learning_rate": 2.9633911084509788e-06, + "loss": 0.0535, + "step": 1026 + }, + { + "epoch": 0.7387160582629023, + "grad_norm": 0.9637168646863399, + "learning_rate": 2.9633163592323786e-06, + "loss": 0.0065, + "step": 1027 + }, + { + "epoch": 0.7394353533537134, + "grad_norm": 8.054302386051065, + "learning_rate": 2.9632415347235825e-06, + "loss": 0.1028, + "step": 1028 + }, + { + "epoch": 0.7401546484445244, + "grad_norm": 3.0726704388718353, + "learning_rate": 2.9631666349284414e-06, + "loss": 0.0907, + "step": 1029 + }, + { + "epoch": 0.7408739435353354, + "grad_norm": 8.51683505148813, + "learning_rate": 2.963091659850808e-06, + "loss": 0.2076, + "step": 1030 + }, + { + "epoch": 0.7415932386261463, + "grad_norm": 6.667369843402183, + "learning_rate": 2.963016609494541e-06, + "loss": 0.1198, + "step": 1031 + }, + { + "epoch": 0.7423125337169574, + "grad_norm": 5.317912112099469, + "learning_rate": 2.9629414838635004e-06, + "loss": 0.0828, + "step": 1032 + }, + { + "epoch": 0.7430318288077684, + "grad_norm": 2.645375849464506, + "learning_rate": 2.962866282961553e-06, + "loss": 0.0802, + "step": 1033 + }, + { + "epoch": 0.7437511238985794, + "grad_norm": 8.688885817653823, + "learning_rate": 2.9627910067925666e-06, + "loss": 0.1758, + "step": 1034 + }, + { + "epoch": 0.7444704189893904, + "grad_norm": 8.254817497594031, + "learning_rate": 2.9627156553604157e-06, + "loss": 0.1529, + "step": 1035 + }, + { + "epoch": 0.7451897140802014, + "grad_norm": 6.204186959347255, + "learning_rate": 2.9626402286689765e-06, + "loss": 0.1675, + "step": 1036 + }, + { + "epoch": 0.7459090091710124, + "grad_norm": 4.408973375756506, + "learning_rate": 2.9625647267221294e-06, + "loss": 0.0916, + "step": 1037 + }, + { + "epoch": 0.7466283042618234, + "grad_norm": 6.056189613910105, + "learning_rate": 2.9624891495237606e-06, + "loss": 0.2593, + "step": 1038 + }, + { + "epoch": 0.7473475993526344, + "grad_norm": 6.390137114374384, + "learning_rate": 2.962413497077757e-06, + "loss": 0.2372, + "step": 1039 + }, + { + "epoch": 0.7480668944434454, + "grad_norm": 6.641398191765912, + "learning_rate": 2.9623377693880123e-06, + "loss": 0.1913, + "step": 1040 + }, + { + "epoch": 0.7487861895342565, + "grad_norm": 6.969302015391716, + "learning_rate": 2.962261966458422e-06, + "loss": 0.1709, + "step": 1041 + }, + { + "epoch": 0.7495054846250674, + "grad_norm": 4.949634384482189, + "learning_rate": 2.9621860882928868e-06, + "loss": 0.1111, + "step": 1042 + }, + { + "epoch": 0.7502247797158784, + "grad_norm": 8.44134191394853, + "learning_rate": 2.9621101348953105e-06, + "loss": 0.1495, + "step": 1043 + }, + { + "epoch": 0.7509440748066895, + "grad_norm": 8.032308034360867, + "learning_rate": 2.962034106269601e-06, + "loss": 0.1796, + "step": 1044 + }, + { + "epoch": 0.7516633698975005, + "grad_norm": 5.1075318898939015, + "learning_rate": 2.96195800241967e-06, + "loss": 0.1875, + "step": 1045 + }, + { + "epoch": 0.7523826649883114, + "grad_norm": 9.949516066867346, + "learning_rate": 2.9618818233494335e-06, + "loss": 0.1622, + "step": 1046 + }, + { + "epoch": 0.7531019600791224, + "grad_norm": 7.077940367846719, + "learning_rate": 2.9618055690628112e-06, + "loss": 0.101, + "step": 1047 + }, + { + "epoch": 0.7538212551699335, + "grad_norm": 7.770866220795333, + "learning_rate": 2.9617292395637265e-06, + "loss": 0.3442, + "step": 1048 + }, + { + "epoch": 0.7545405502607445, + "grad_norm": 4.687233301268084, + "learning_rate": 2.9616528348561057e-06, + "loss": 0.2032, + "step": 1049 + }, + { + "epoch": 0.7552598453515554, + "grad_norm": 5.3693367091754665, + "learning_rate": 2.961576354943881e-06, + "loss": 0.2438, + "step": 1050 + }, + { + "epoch": 0.7559791404423665, + "grad_norm": 5.873247532550017, + "learning_rate": 2.9614997998309878e-06, + "loss": 0.2456, + "step": 1051 + }, + { + "epoch": 0.7566984355331775, + "grad_norm": 6.4936719305034405, + "learning_rate": 2.961423169521363e-06, + "loss": 0.3128, + "step": 1052 + }, + { + "epoch": 0.7574177306239885, + "grad_norm": 4.460512181817813, + "learning_rate": 2.961346464018952e-06, + "loss": 0.094, + "step": 1053 + }, + { + "epoch": 0.7581370257147995, + "grad_norm": 6.5616676589913006, + "learning_rate": 2.961269683327699e-06, + "loss": 0.1021, + "step": 1054 + }, + { + "epoch": 0.7588563208056105, + "grad_norm": 7.666176125891536, + "learning_rate": 2.9611928274515567e-06, + "loss": 0.1959, + "step": 1055 + }, + { + "epoch": 0.7595756158964215, + "grad_norm": 4.8175082478123485, + "learning_rate": 2.9611158963944776e-06, + "loss": 0.0594, + "step": 1056 + }, + { + "epoch": 0.7602949109872326, + "grad_norm": 4.1431375213170405, + "learning_rate": 2.961038890160421e-06, + "loss": 0.1013, + "step": 1057 + }, + { + "epoch": 0.7610142060780435, + "grad_norm": 4.763195705949494, + "learning_rate": 2.9609618087533486e-06, + "loss": 0.2288, + "step": 1058 + }, + { + "epoch": 0.7617335011688545, + "grad_norm": 5.651407027708578, + "learning_rate": 2.9608846521772265e-06, + "loss": 0.1345, + "step": 1059 + }, + { + "epoch": 0.7624527962596656, + "grad_norm": 7.988657506027217, + "learning_rate": 2.9608074204360245e-06, + "loss": 0.0894, + "step": 1060 + }, + { + "epoch": 0.7631720913504765, + "grad_norm": 4.799091978005488, + "learning_rate": 2.9607301135337165e-06, + "loss": 0.0689, + "step": 1061 + }, + { + "epoch": 0.7638913864412875, + "grad_norm": 7.254078601061779, + "learning_rate": 2.96065273147428e-06, + "loss": 0.1273, + "step": 1062 + }, + { + "epoch": 0.7646106815320985, + "grad_norm": 5.505255175721506, + "learning_rate": 2.9605752742616963e-06, + "loss": 0.1091, + "step": 1063 + }, + { + "epoch": 0.7653299766229096, + "grad_norm": 9.390324692548925, + "learning_rate": 2.960497741899951e-06, + "loss": 0.2214, + "step": 1064 + }, + { + "epoch": 0.7660492717137205, + "grad_norm": 8.265178676861693, + "learning_rate": 2.9604201343930326e-06, + "loss": 0.2469, + "step": 1065 + }, + { + "epoch": 0.7667685668045315, + "grad_norm": 5.013709944651262, + "learning_rate": 2.9603424517449343e-06, + "loss": 0.0646, + "step": 1066 + }, + { + "epoch": 0.7674878618953426, + "grad_norm": 8.230599844197473, + "learning_rate": 2.960264693959654e-06, + "loss": 0.2249, + "step": 1067 + }, + { + "epoch": 0.7682071569861536, + "grad_norm": 4.821036551476318, + "learning_rate": 2.9601868610411914e-06, + "loss": 0.1896, + "step": 1068 + }, + { + "epoch": 0.7689264520769645, + "grad_norm": 6.817500977639897, + "learning_rate": 2.960108952993552e-06, + "loss": 0.1728, + "step": 1069 + }, + { + "epoch": 0.7696457471677756, + "grad_norm": 5.993676325974895, + "learning_rate": 2.960030969820743e-06, + "loss": 0.1427, + "step": 1070 + }, + { + "epoch": 0.7703650422585866, + "grad_norm": 5.1225152323778715, + "learning_rate": 2.9599529115267787e-06, + "loss": 0.2073, + "step": 1071 + }, + { + "epoch": 0.7710843373493976, + "grad_norm": 4.727738733597606, + "learning_rate": 2.9598747781156734e-06, + "loss": 0.1028, + "step": 1072 + }, + { + "epoch": 0.7718036324402086, + "grad_norm": 4.8080381627304805, + "learning_rate": 2.9597965695914486e-06, + "loss": 0.1531, + "step": 1073 + }, + { + "epoch": 0.7725229275310196, + "grad_norm": 7.1750051303221545, + "learning_rate": 2.959718285958127e-06, + "loss": 0.0973, + "step": 1074 + }, + { + "epoch": 0.7732422226218306, + "grad_norm": 7.560782095923482, + "learning_rate": 2.959639927219738e-06, + "loss": 0.3099, + "step": 1075 + }, + { + "epoch": 0.7739615177126417, + "grad_norm": 6.081471630530151, + "learning_rate": 2.959561493380312e-06, + "loss": 0.2704, + "step": 1076 + }, + { + "epoch": 0.7746808128034526, + "grad_norm": 4.5573471916570005, + "learning_rate": 2.959482984443885e-06, + "loss": 0.1634, + "step": 1077 + }, + { + "epoch": 0.7754001078942636, + "grad_norm": 8.024352829564638, + "learning_rate": 2.959404400414496e-06, + "loss": 0.121, + "step": 1078 + }, + { + "epoch": 0.7761194029850746, + "grad_norm": 6.826364976830265, + "learning_rate": 2.959325741296189e-06, + "loss": 0.213, + "step": 1079 + }, + { + "epoch": 0.7768386980758857, + "grad_norm": 4.616172305121968, + "learning_rate": 2.959247007093011e-06, + "loss": 0.0142, + "step": 1080 + }, + { + "epoch": 0.7775579931666966, + "grad_norm": 3.903262585727129, + "learning_rate": 2.9591681978090127e-06, + "loss": 0.1081, + "step": 1081 + }, + { + "epoch": 0.7782772882575076, + "grad_norm": 7.339701277038209, + "learning_rate": 2.9590893134482493e-06, + "loss": 0.1958, + "step": 1082 + }, + { + "epoch": 0.7789965833483187, + "grad_norm": 7.0135713771378425, + "learning_rate": 2.959010354014779e-06, + "loss": 0.0795, + "step": 1083 + }, + { + "epoch": 0.7797158784391297, + "grad_norm": 3.777432169863555, + "learning_rate": 2.9589313195126654e-06, + "loss": 0.1546, + "step": 1084 + }, + { + "epoch": 0.7804351735299406, + "grad_norm": 4.856435890238801, + "learning_rate": 2.958852209945974e-06, + "loss": 0.1952, + "step": 1085 + }, + { + "epoch": 0.7811544686207517, + "grad_norm": 8.20527872966151, + "learning_rate": 2.958773025318775e-06, + "loss": 0.0835, + "step": 1086 + }, + { + "epoch": 0.7818737637115627, + "grad_norm": 5.514808434154136, + "learning_rate": 2.9586937656351437e-06, + "loss": 0.1738, + "step": 1087 + }, + { + "epoch": 0.7825930588023736, + "grad_norm": 4.659218147016898, + "learning_rate": 2.958614430899157e-06, + "loss": 0.0753, + "step": 1088 + }, + { + "epoch": 0.7833123538931847, + "grad_norm": 2.847833195278624, + "learning_rate": 2.958535021114898e-06, + "loss": 0.0229, + "step": 1089 + }, + { + "epoch": 0.7840316489839957, + "grad_norm": 4.964037366847555, + "learning_rate": 2.958455536286451e-06, + "loss": 0.1589, + "step": 1090 + }, + { + "epoch": 0.7847509440748067, + "grad_norm": 4.055532056707014, + "learning_rate": 2.958375976417907e-06, + "loss": 0.0392, + "step": 1091 + }, + { + "epoch": 0.7854702391656176, + "grad_norm": 7.985326591241292, + "learning_rate": 2.958296341513359e-06, + "loss": 0.1535, + "step": 1092 + }, + { + "epoch": 0.7861895342564287, + "grad_norm": 3.0336896592445393, + "learning_rate": 2.9582166315769037e-06, + "loss": 0.038, + "step": 1093 + }, + { + "epoch": 0.7869088293472397, + "grad_norm": 4.4161714083243355, + "learning_rate": 2.9581368466126434e-06, + "loss": 0.0506, + "step": 1094 + }, + { + "epoch": 0.7876281244380507, + "grad_norm": 6.301016191414249, + "learning_rate": 2.9580569866246827e-06, + "loss": 0.1047, + "step": 1095 + }, + { + "epoch": 0.7883474195288617, + "grad_norm": 5.4218925181388835, + "learning_rate": 2.95797705161713e-06, + "loss": 0.1682, + "step": 1096 + }, + { + "epoch": 0.7890667146196727, + "grad_norm": 5.861239821465975, + "learning_rate": 2.9578970415940986e-06, + "loss": 0.1167, + "step": 1097 + }, + { + "epoch": 0.7897860097104837, + "grad_norm": 7.475721342617539, + "learning_rate": 2.9578169565597058e-06, + "loss": 0.0846, + "step": 1098 + }, + { + "epoch": 0.7905053048012948, + "grad_norm": 7.172766655699195, + "learning_rate": 2.9577367965180705e-06, + "loss": 0.0873, + "step": 1099 + }, + { + "epoch": 0.7912245998921057, + "grad_norm": 8.105874256968812, + "learning_rate": 2.957656561473319e-06, + "loss": 0.0662, + "step": 1100 + }, + { + "epoch": 0.7919438949829167, + "grad_norm": 6.239076982398492, + "learning_rate": 2.957576251429578e-06, + "loss": 0.1863, + "step": 1101 + }, + { + "epoch": 0.7926631900737278, + "grad_norm": 5.69335042861376, + "learning_rate": 2.9574958663909803e-06, + "loss": 0.1408, + "step": 1102 + }, + { + "epoch": 0.7933824851645388, + "grad_norm": 5.003234054276349, + "learning_rate": 2.957415406361662e-06, + "loss": 0.028, + "step": 1103 + }, + { + "epoch": 0.7941017802553497, + "grad_norm": 4.592670187945757, + "learning_rate": 2.9573348713457623e-06, + "loss": 0.2002, + "step": 1104 + }, + { + "epoch": 0.7948210753461608, + "grad_norm": 6.293272077818915, + "learning_rate": 2.9572542613474254e-06, + "loss": 0.2168, + "step": 1105 + }, + { + "epoch": 0.7955403704369718, + "grad_norm": 12.009341416844594, + "learning_rate": 2.9571735763707985e-06, + "loss": 0.3254, + "step": 1106 + }, + { + "epoch": 0.7962596655277828, + "grad_norm": 7.263073206419407, + "learning_rate": 2.9570928164200333e-06, + "loss": 0.0756, + "step": 1107 + }, + { + "epoch": 0.7969789606185937, + "grad_norm": 6.098330065257, + "learning_rate": 2.9570119814992844e-06, + "loss": 0.3136, + "step": 1108 + }, + { + "epoch": 0.7976982557094048, + "grad_norm": 7.046862916564739, + "learning_rate": 2.9569310716127116e-06, + "loss": 0.2601, + "step": 1109 + }, + { + "epoch": 0.7984175508002158, + "grad_norm": 5.88888767432492, + "learning_rate": 2.9568500867644777e-06, + "loss": 0.2498, + "step": 1110 + }, + { + "epoch": 0.7991368458910268, + "grad_norm": 6.403418222612225, + "learning_rate": 2.9567690269587493e-06, + "loss": 0.161, + "step": 1111 + }, + { + "epoch": 0.7998561409818378, + "grad_norm": 7.526621169944265, + "learning_rate": 2.956687892199697e-06, + "loss": 0.2115, + "step": 1112 + }, + { + "epoch": 0.8005754360726488, + "grad_norm": 8.018745664550192, + "learning_rate": 2.956606682491496e-06, + "loss": 0.1093, + "step": 1113 + }, + { + "epoch": 0.8012947311634598, + "grad_norm": 6.200905213411397, + "learning_rate": 2.956525397838324e-06, + "loss": 0.1934, + "step": 1114 + }, + { + "epoch": 0.8020140262542708, + "grad_norm": 7.932886285925514, + "learning_rate": 2.9564440382443636e-06, + "loss": 0.2301, + "step": 1115 + }, + { + "epoch": 0.8027333213450818, + "grad_norm": 4.949344856984922, + "learning_rate": 2.9563626037138003e-06, + "loss": 0.3011, + "step": 1116 + }, + { + "epoch": 0.8034526164358928, + "grad_norm": 8.274999836950926, + "learning_rate": 2.956281094250825e-06, + "loss": 0.3443, + "step": 1117 + }, + { + "epoch": 0.8041719115267039, + "grad_norm": 5.756304489457235, + "learning_rate": 2.9561995098596304e-06, + "loss": 0.1179, + "step": 1118 + }, + { + "epoch": 0.8048912066175148, + "grad_norm": 6.890620747454537, + "learning_rate": 2.956117850544415e-06, + "loss": 0.1811, + "step": 1119 + }, + { + "epoch": 0.8056105017083258, + "grad_norm": 9.648966240120986, + "learning_rate": 2.95603611630938e-06, + "loss": 0.3437, + "step": 1120 + }, + { + "epoch": 0.8063297967991369, + "grad_norm": 6.604232863540325, + "learning_rate": 2.955954307158731e-06, + "loss": 0.1327, + "step": 1121 + }, + { + "epoch": 0.8070490918899479, + "grad_norm": 4.577272025294506, + "learning_rate": 2.955872423096677e-06, + "loss": 0.0532, + "step": 1122 + }, + { + "epoch": 0.8077683869807588, + "grad_norm": 5.682910705275785, + "learning_rate": 2.955790464127431e-06, + "loss": 0.2226, + "step": 1123 + }, + { + "epoch": 0.8084876820715698, + "grad_norm": 6.64412933635135, + "learning_rate": 2.9557084302552104e-06, + "loss": 0.1609, + "step": 1124 + }, + { + "epoch": 0.8092069771623809, + "grad_norm": 6.443648576287447, + "learning_rate": 2.955626321484235e-06, + "loss": 0.0856, + "step": 1125 + }, + { + "epoch": 0.8099262722531919, + "grad_norm": 7.326989682730623, + "learning_rate": 2.9555441378187304e-06, + "loss": 0.1617, + "step": 1126 + }, + { + "epoch": 0.8106455673440028, + "grad_norm": 7.9020294740056585, + "learning_rate": 2.9554618792629248e-06, + "loss": 0.4006, + "step": 1127 + }, + { + "epoch": 0.8113648624348139, + "grad_norm": 4.302728084841077, + "learning_rate": 2.9553795458210506e-06, + "loss": 0.1214, + "step": 1128 + }, + { + "epoch": 0.8120841575256249, + "grad_norm": 7.649938144231672, + "learning_rate": 2.9552971374973443e-06, + "loss": 0.1833, + "step": 1129 + }, + { + "epoch": 0.8128034526164359, + "grad_norm": 6.8285741943298675, + "learning_rate": 2.955214654296045e-06, + "loss": 0.2469, + "step": 1130 + }, + { + "epoch": 0.8135227477072469, + "grad_norm": 8.6710081624358, + "learning_rate": 2.955132096221397e-06, + "loss": 0.0963, + "step": 1131 + }, + { + "epoch": 0.8142420427980579, + "grad_norm": 8.997418489820435, + "learning_rate": 2.955049463277649e-06, + "loss": 0.2333, + "step": 1132 + }, + { + "epoch": 0.8149613378888689, + "grad_norm": 7.300476127066846, + "learning_rate": 2.9549667554690507e-06, + "loss": 0.1037, + "step": 1133 + }, + { + "epoch": 0.81568063297968, + "grad_norm": 6.147210142800407, + "learning_rate": 2.95488397279986e-06, + "loss": 0.1754, + "step": 1134 + }, + { + "epoch": 0.8163999280704909, + "grad_norm": 6.078805507443698, + "learning_rate": 2.9548011152743345e-06, + "loss": 0.2416, + "step": 1135 + }, + { + "epoch": 0.8171192231613019, + "grad_norm": 7.486754463637567, + "learning_rate": 2.954718182896737e-06, + "loss": 0.3225, + "step": 1136 + }, + { + "epoch": 0.817838518252113, + "grad_norm": 4.44992158630488, + "learning_rate": 2.954635175671337e-06, + "loss": 0.0599, + "step": 1137 + }, + { + "epoch": 0.818557813342924, + "grad_norm": 9.187427134684947, + "learning_rate": 2.954552093602402e-06, + "loss": 0.2059, + "step": 1138 + }, + { + "epoch": 0.8192771084337349, + "grad_norm": 3.794189057153781, + "learning_rate": 2.9544689366942094e-06, + "loss": 0.0773, + "step": 1139 + }, + { + "epoch": 0.819996403524546, + "grad_norm": 7.126594883887073, + "learning_rate": 2.9543857049510366e-06, + "loss": 0.1345, + "step": 1140 + }, + { + "epoch": 0.820715698615357, + "grad_norm": 6.5230481639395315, + "learning_rate": 2.954302398377166e-06, + "loss": 0.2994, + "step": 1141 + }, + { + "epoch": 0.8214349937061679, + "grad_norm": 5.126007158127512, + "learning_rate": 2.954219016976884e-06, + "loss": 0.162, + "step": 1142 + }, + { + "epoch": 0.8221542887969789, + "grad_norm": 3.7723544701317153, + "learning_rate": 2.9541355607544807e-06, + "loss": 0.19, + "step": 1143 + }, + { + "epoch": 0.82287358388779, + "grad_norm": 6.178592525153852, + "learning_rate": 2.9540520297142505e-06, + "loss": 0.0789, + "step": 1144 + }, + { + "epoch": 0.823592878978601, + "grad_norm": 3.672396832378457, + "learning_rate": 2.9539684238604905e-06, + "loss": 0.0887, + "step": 1145 + }, + { + "epoch": 0.8243121740694119, + "grad_norm": 6.795273015308563, + "learning_rate": 2.953884743197503e-06, + "loss": 0.1575, + "step": 1146 + }, + { + "epoch": 0.825031469160223, + "grad_norm": 6.8400877172759715, + "learning_rate": 2.9538009877295934e-06, + "loss": 0.2467, + "step": 1147 + }, + { + "epoch": 0.825750764251034, + "grad_norm": 7.9529661138274435, + "learning_rate": 2.9537171574610706e-06, + "loss": 0.2735, + "step": 1148 + }, + { + "epoch": 0.826470059341845, + "grad_norm": 6.646315973163267, + "learning_rate": 2.9536332523962483e-06, + "loss": 0.2352, + "step": 1149 + }, + { + "epoch": 0.827189354432656, + "grad_norm": 5.4355372209105095, + "learning_rate": 2.953549272539443e-06, + "loss": 0.2303, + "step": 1150 + }, + { + "epoch": 0.827908649523467, + "grad_norm": 4.649327710734051, + "learning_rate": 2.9534652178949764e-06, + "loss": 0.1088, + "step": 1151 + }, + { + "epoch": 0.828627944614278, + "grad_norm": 7.301757477076145, + "learning_rate": 2.953381088467173e-06, + "loss": 0.1146, + "step": 1152 + }, + { + "epoch": 0.8293472397050891, + "grad_norm": 6.484628002096242, + "learning_rate": 2.953296884260361e-06, + "loss": 0.1852, + "step": 1153 + }, + { + "epoch": 0.8300665347959, + "grad_norm": 6.809996187292308, + "learning_rate": 2.953212605278873e-06, + "loss": 0.051, + "step": 1154 + }, + { + "epoch": 0.830785829886711, + "grad_norm": 4.737229044265182, + "learning_rate": 2.9531282515270457e-06, + "loss": 0.1251, + "step": 1155 + }, + { + "epoch": 0.831505124977522, + "grad_norm": 6.21572016099917, + "learning_rate": 2.953043823009219e-06, + "loss": 0.1774, + "step": 1156 + }, + { + "epoch": 0.8322244200683331, + "grad_norm": 3.982063537099724, + "learning_rate": 2.952959319729737e-06, + "loss": 0.1102, + "step": 1157 + }, + { + "epoch": 0.832943715159144, + "grad_norm": 4.725101822974682, + "learning_rate": 2.9528747416929465e-06, + "loss": 0.105, + "step": 1158 + }, + { + "epoch": 0.833663010249955, + "grad_norm": 9.06050821331901, + "learning_rate": 2.952790088903201e-06, + "loss": 0.3101, + "step": 1159 + }, + { + "epoch": 0.8343823053407661, + "grad_norm": 3.292327735901539, + "learning_rate": 2.9527053613648548e-06, + "loss": 0.097, + "step": 1160 + }, + { + "epoch": 0.8351016004315771, + "grad_norm": 5.202336927832246, + "learning_rate": 2.952620559082268e-06, + "loss": 0.2143, + "step": 1161 + }, + { + "epoch": 0.835820895522388, + "grad_norm": 6.689117450786752, + "learning_rate": 2.952535682059803e-06, + "loss": 0.1148, + "step": 1162 + }, + { + "epoch": 0.8365401906131991, + "grad_norm": 3.639434170382141, + "learning_rate": 2.9524507303018274e-06, + "loss": 0.1436, + "step": 1163 + }, + { + "epoch": 0.8372594857040101, + "grad_norm": 5.337383021001455, + "learning_rate": 2.9523657038127117e-06, + "loss": 0.1328, + "step": 1164 + }, + { + "epoch": 0.837978780794821, + "grad_norm": 10.150920536661177, + "learning_rate": 2.9522806025968315e-06, + "loss": 0.1365, + "step": 1165 + }, + { + "epoch": 0.838698075885632, + "grad_norm": 7.101209791133696, + "learning_rate": 2.952195426658565e-06, + "loss": 0.1619, + "step": 1166 + }, + { + "epoch": 0.8394173709764431, + "grad_norm": 7.04230801618488, + "learning_rate": 2.9521101760022943e-06, + "loss": 0.2199, + "step": 1167 + }, + { + "epoch": 0.8401366660672541, + "grad_norm": 9.453457447970827, + "learning_rate": 2.952024850632406e-06, + "loss": 0.3058, + "step": 1168 + }, + { + "epoch": 0.840855961158065, + "grad_norm": 4.428357603479249, + "learning_rate": 2.95193945055329e-06, + "loss": 0.0438, + "step": 1169 + }, + { + "epoch": 0.8415752562488761, + "grad_norm": 4.189763361018842, + "learning_rate": 2.9518539757693406e-06, + "loss": 0.1537, + "step": 1170 + }, + { + "epoch": 0.8422945513396871, + "grad_norm": 3.3079856837768338, + "learning_rate": 2.9517684262849557e-06, + "loss": 0.0736, + "step": 1171 + }, + { + "epoch": 0.8430138464304981, + "grad_norm": 2.643169964527249, + "learning_rate": 2.9516828021045363e-06, + "loss": 0.0104, + "step": 1172 + }, + { + "epoch": 0.8437331415213091, + "grad_norm": 6.765432324989906, + "learning_rate": 2.951597103232489e-06, + "loss": 0.131, + "step": 1173 + }, + { + "epoch": 0.8444524366121201, + "grad_norm": 3.418371881465293, + "learning_rate": 2.951511329673222e-06, + "loss": 0.0769, + "step": 1174 + }, + { + "epoch": 0.8451717317029311, + "grad_norm": 4.966100371899219, + "learning_rate": 2.9514254814311493e-06, + "loss": 0.1442, + "step": 1175 + }, + { + "epoch": 0.8458910267937422, + "grad_norm": 5.023802103418847, + "learning_rate": 2.9513395585106873e-06, + "loss": 0.0917, + "step": 1176 + }, + { + "epoch": 0.8466103218845531, + "grad_norm": 4.664124893028925, + "learning_rate": 2.9512535609162577e-06, + "loss": 0.1302, + "step": 1177 + }, + { + "epoch": 0.8473296169753641, + "grad_norm": 6.295702829080911, + "learning_rate": 2.951167488652285e-06, + "loss": 0.0948, + "step": 1178 + }, + { + "epoch": 0.8480489120661752, + "grad_norm": 7.916290328605307, + "learning_rate": 2.951081341723197e-06, + "loss": 0.1991, + "step": 1179 + }, + { + "epoch": 0.8487682071569862, + "grad_norm": 7.6712222152259715, + "learning_rate": 2.950995120133427e-06, + "loss": 0.1496, + "step": 1180 + }, + { + "epoch": 0.8494875022477971, + "grad_norm": 9.276953839431325, + "learning_rate": 2.950908823887411e-06, + "loss": 0.2795, + "step": 1181 + }, + { + "epoch": 0.8502067973386082, + "grad_norm": 1.9217269220028907, + "learning_rate": 2.950822452989589e-06, + "loss": 0.0253, + "step": 1182 + }, + { + "epoch": 0.8509260924294192, + "grad_norm": 8.14480139577424, + "learning_rate": 2.950736007444404e-06, + "loss": 0.1772, + "step": 1183 + }, + { + "epoch": 0.8516453875202302, + "grad_norm": 8.885296371982806, + "learning_rate": 2.950649487256306e-06, + "loss": 0.114, + "step": 1184 + }, + { + "epoch": 0.8523646826110411, + "grad_norm": 8.971215417365684, + "learning_rate": 2.950562892429745e-06, + "loss": 0.0933, + "step": 1185 + }, + { + "epoch": 0.8530839777018522, + "grad_norm": 2.573547917683082, + "learning_rate": 2.9504762229691767e-06, + "loss": 0.0371, + "step": 1186 + }, + { + "epoch": 0.8538032727926632, + "grad_norm": 3.9643645332769712, + "learning_rate": 2.9503894788790604e-06, + "loss": 0.0691, + "step": 1187 + }, + { + "epoch": 0.8545225678834742, + "grad_norm": 5.7179828786626175, + "learning_rate": 2.9503026601638592e-06, + "loss": 0.1094, + "step": 1188 + }, + { + "epoch": 0.8552418629742852, + "grad_norm": 1.2052738335468935, + "learning_rate": 2.9502157668280407e-06, + "loss": 0.003, + "step": 1189 + }, + { + "epoch": 0.8559611580650962, + "grad_norm": 3.6236871979980614, + "learning_rate": 2.9501287988760747e-06, + "loss": 0.0241, + "step": 1190 + }, + { + "epoch": 0.8566804531559072, + "grad_norm": 7.552550722934375, + "learning_rate": 2.950041756312437e-06, + "loss": 0.2191, + "step": 1191 + }, + { + "epoch": 0.8573997482467182, + "grad_norm": 8.600707780290314, + "learning_rate": 2.9499546391416053e-06, + "loss": 0.2345, + "step": 1192 + }, + { + "epoch": 0.8581190433375292, + "grad_norm": 5.803646814690306, + "learning_rate": 2.9498674473680617e-06, + "loss": 0.0759, + "step": 1193 + }, + { + "epoch": 0.8588383384283402, + "grad_norm": 6.604878421124933, + "learning_rate": 2.9497801809962927e-06, + "loss": 0.0672, + "step": 1194 + }, + { + "epoch": 0.8595576335191513, + "grad_norm": 6.378136126701569, + "learning_rate": 2.9496928400307886e-06, + "loss": 0.196, + "step": 1195 + }, + { + "epoch": 0.8602769286099622, + "grad_norm": 5.707386542098853, + "learning_rate": 2.9496054244760434e-06, + "loss": 0.0935, + "step": 1196 + }, + { + "epoch": 0.8609962237007732, + "grad_norm": 6.677928575334813, + "learning_rate": 2.949517934336554e-06, + "loss": 0.1042, + "step": 1197 + }, + { + "epoch": 0.8617155187915843, + "grad_norm": 5.3079440530726885, + "learning_rate": 2.949430369616822e-06, + "loss": 0.3231, + "step": 1198 + }, + { + "epoch": 0.8624348138823953, + "grad_norm": 4.190781666623576, + "learning_rate": 2.949342730321354e-06, + "loss": 0.1149, + "step": 1199 + }, + { + "epoch": 0.8631541089732062, + "grad_norm": 6.1539370694544475, + "learning_rate": 2.949255016454658e-06, + "loss": 0.0918, + "step": 1200 + }, + { + "epoch": 0.8638734040640172, + "grad_norm": 3.4290825650816505, + "learning_rate": 2.9491672280212467e-06, + "loss": 0.1636, + "step": 1201 + }, + { + "epoch": 0.8645926991548283, + "grad_norm": 9.99921442925207, + "learning_rate": 2.9490793650256383e-06, + "loss": 0.1481, + "step": 1202 + }, + { + "epoch": 0.8653119942456393, + "grad_norm": 5.949960649467944, + "learning_rate": 2.9489914274723526e-06, + "loss": 0.0723, + "step": 1203 + }, + { + "epoch": 0.8660312893364502, + "grad_norm": 7.583766986501179, + "learning_rate": 2.948903415365914e-06, + "loss": 0.1765, + "step": 1204 + }, + { + "epoch": 0.8667505844272613, + "grad_norm": 9.160160172140433, + "learning_rate": 2.948815328710852e-06, + "loss": 0.0937, + "step": 1205 + }, + { + "epoch": 0.8674698795180723, + "grad_norm": 5.684521603683679, + "learning_rate": 2.948727167511697e-06, + "loss": 0.2544, + "step": 1206 + }, + { + "epoch": 0.8681891746088833, + "grad_norm": 4.213944374192409, + "learning_rate": 2.9486389317729865e-06, + "loss": 0.0375, + "step": 1207 + }, + { + "epoch": 0.8689084696996943, + "grad_norm": 5.973771362392137, + "learning_rate": 2.9485506214992602e-06, + "loss": 0.1503, + "step": 1208 + }, + { + "epoch": 0.8696277647905053, + "grad_norm": 6.228585309902353, + "learning_rate": 2.9484622366950613e-06, + "loss": 0.0494, + "step": 1209 + }, + { + "epoch": 0.8703470598813163, + "grad_norm": 6.940746207053184, + "learning_rate": 2.948373777364938e-06, + "loss": 0.0685, + "step": 1210 + }, + { + "epoch": 0.8710663549721274, + "grad_norm": 4.388336320038271, + "learning_rate": 2.9482852435134406e-06, + "loss": 0.0836, + "step": 1211 + }, + { + "epoch": 0.8717856500629383, + "grad_norm": 2.5279300724029734, + "learning_rate": 2.9481966351451257e-06, + "loss": 0.0068, + "step": 1212 + }, + { + "epoch": 0.8725049451537493, + "grad_norm": 4.259963930781713, + "learning_rate": 2.9481079522645515e-06, + "loss": 0.0721, + "step": 1213 + }, + { + "epoch": 0.8732242402445604, + "grad_norm": 6.216823784857963, + "learning_rate": 2.948019194876281e-06, + "loss": 0.3805, + "step": 1214 + }, + { + "epoch": 0.8739435353353714, + "grad_norm": 6.62381001420388, + "learning_rate": 2.9479303629848808e-06, + "loss": 0.0861, + "step": 1215 + }, + { + "epoch": 0.8746628304261823, + "grad_norm": 4.2549517734651, + "learning_rate": 2.9478414565949216e-06, + "loss": 0.0706, + "step": 1216 + }, + { + "epoch": 0.8753821255169933, + "grad_norm": 5.935522314523385, + "learning_rate": 2.9477524757109783e-06, + "loss": 0.1294, + "step": 1217 + }, + { + "epoch": 0.8761014206078044, + "grad_norm": 6.954651229777788, + "learning_rate": 2.9476634203376286e-06, + "loss": 0.2319, + "step": 1218 + }, + { + "epoch": 0.8768207156986153, + "grad_norm": 4.200440348501926, + "learning_rate": 2.9475742904794546e-06, + "loss": 0.0946, + "step": 1219 + }, + { + "epoch": 0.8775400107894263, + "grad_norm": 6.456132926050342, + "learning_rate": 2.947485086141042e-06, + "loss": 0.2028, + "step": 1220 + }, + { + "epoch": 0.8782593058802374, + "grad_norm": 5.677931268460433, + "learning_rate": 2.947395807326981e-06, + "loss": 0.2854, + "step": 1221 + }, + { + "epoch": 0.8789786009710484, + "grad_norm": 5.203662927823084, + "learning_rate": 2.9473064540418647e-06, + "loss": 0.2042, + "step": 1222 + }, + { + "epoch": 0.8796978960618593, + "grad_norm": 4.362549025321659, + "learning_rate": 2.947217026290291e-06, + "loss": 0.1679, + "step": 1223 + }, + { + "epoch": 0.8804171911526704, + "grad_norm": 6.652739407814444, + "learning_rate": 2.94712752407686e-06, + "loss": 0.0691, + "step": 1224 + }, + { + "epoch": 0.8811364862434814, + "grad_norm": 5.045016718414958, + "learning_rate": 2.9470379474061786e-06, + "loss": 0.0467, + "step": 1225 + }, + { + "epoch": 0.8818557813342924, + "grad_norm": 10.990865894028701, + "learning_rate": 2.946948296282854e-06, + "loss": 0.2178, + "step": 1226 + }, + { + "epoch": 0.8825750764251034, + "grad_norm": 5.601601655540781, + "learning_rate": 2.9468585707114996e-06, + "loss": 0.2552, + "step": 1227 + }, + { + "epoch": 0.8832943715159144, + "grad_norm": 6.521488628112908, + "learning_rate": 2.946768770696732e-06, + "loss": 0.1975, + "step": 1228 + }, + { + "epoch": 0.8840136666067254, + "grad_norm": 6.806559308235417, + "learning_rate": 2.946678896243171e-06, + "loss": 0.1087, + "step": 1229 + }, + { + "epoch": 0.8847329616975365, + "grad_norm": 3.8790748262950028, + "learning_rate": 2.946588947355442e-06, + "loss": 0.0668, + "step": 1230 + }, + { + "epoch": 0.8854522567883474, + "grad_norm": 4.823969467566996, + "learning_rate": 2.9464989240381715e-06, + "loss": 0.0816, + "step": 1231 + }, + { + "epoch": 0.8861715518791584, + "grad_norm": 5.323535724191767, + "learning_rate": 2.946408826295992e-06, + "loss": 0.2467, + "step": 1232 + }, + { + "epoch": 0.8868908469699694, + "grad_norm": 7.013989091142057, + "learning_rate": 2.94631865413354e-06, + "loss": 0.2513, + "step": 1233 + }, + { + "epoch": 0.8876101420607805, + "grad_norm": 5.465247838719756, + "learning_rate": 2.946228407555454e-06, + "loss": 0.1457, + "step": 1234 + }, + { + "epoch": 0.8883294371515914, + "grad_norm": 4.679006903274349, + "learning_rate": 2.9461380865663774e-06, + "loss": 0.1078, + "step": 1235 + }, + { + "epoch": 0.8890487322424024, + "grad_norm": 5.483062868460178, + "learning_rate": 2.946047691170958e-06, + "loss": 0.1565, + "step": 1236 + }, + { + "epoch": 0.8897680273332135, + "grad_norm": 5.406812191515783, + "learning_rate": 2.9459572213738464e-06, + "loss": 0.035, + "step": 1237 + }, + { + "epoch": 0.8904873224240245, + "grad_norm": 2.5888970798615603, + "learning_rate": 2.945866677179698e-06, + "loss": 0.0136, + "step": 1238 + }, + { + "epoch": 0.8912066175148354, + "grad_norm": 3.9216443470909277, + "learning_rate": 2.9457760585931695e-06, + "loss": 0.1292, + "step": 1239 + }, + { + "epoch": 0.8919259126056465, + "grad_norm": 1.0890010511649675, + "learning_rate": 2.945685365618926e-06, + "loss": 0.0035, + "step": 1240 + }, + { + "epoch": 0.8926452076964575, + "grad_norm": 5.750388507468079, + "learning_rate": 2.9455945982616323e-06, + "loss": 0.2842, + "step": 1241 + }, + { + "epoch": 0.8933645027872684, + "grad_norm": 9.726848843463296, + "learning_rate": 2.9455037565259584e-06, + "loss": 0.0856, + "step": 1242 + }, + { + "epoch": 0.8940837978780795, + "grad_norm": 4.062250826333195, + "learning_rate": 2.945412840416579e-06, + "loss": 0.0195, + "step": 1243 + }, + { + "epoch": 0.8948030929688905, + "grad_norm": 8.346903018705726, + "learning_rate": 2.945321849938172e-06, + "loss": 0.2444, + "step": 1244 + }, + { + "epoch": 0.8955223880597015, + "grad_norm": 6.2177094945650095, + "learning_rate": 2.9452307850954182e-06, + "loss": 0.1567, + "step": 1245 + }, + { + "epoch": 0.8962416831505124, + "grad_norm": 6.6402104096089065, + "learning_rate": 2.9451396458930037e-06, + "loss": 0.191, + "step": 1246 + }, + { + "epoch": 0.8969609782413235, + "grad_norm": 3.4509943019419875, + "learning_rate": 2.9450484323356174e-06, + "loss": 0.0691, + "step": 1247 + }, + { + "epoch": 0.8976802733321345, + "grad_norm": 5.251082501208291, + "learning_rate": 2.9449571444279525e-06, + "loss": 0.1794, + "step": 1248 + }, + { + "epoch": 0.8983995684229455, + "grad_norm": 6.132516370067139, + "learning_rate": 2.944865782174706e-06, + "loss": 0.2095, + "step": 1249 + }, + { + "epoch": 0.8991188635137565, + "grad_norm": 7.915475872403965, + "learning_rate": 2.9447743455805794e-06, + "loss": 0.1762, + "step": 1250 + }, + { + "epoch": 0.8998381586045675, + "grad_norm": 6.770286576991967, + "learning_rate": 2.9446828346502754e-06, + "loss": 0.1382, + "step": 1251 + }, + { + "epoch": 0.9005574536953785, + "grad_norm": 6.971956049358906, + "learning_rate": 2.944591249388504e-06, + "loss": 0.2458, + "step": 1252 + }, + { + "epoch": 0.9012767487861896, + "grad_norm": 4.784110524046328, + "learning_rate": 2.944499589799977e-06, + "loss": 0.2185, + "step": 1253 + }, + { + "epoch": 0.9019960438770005, + "grad_norm": 5.236202692358694, + "learning_rate": 2.9444078558894104e-06, + "loss": 0.0482, + "step": 1254 + }, + { + "epoch": 0.9027153389678115, + "grad_norm": 4.625465142948313, + "learning_rate": 2.944316047661524e-06, + "loss": 0.1741, + "step": 1255 + }, + { + "epoch": 0.9034346340586226, + "grad_norm": 4.24548194353502, + "learning_rate": 2.944224165121041e-06, + "loss": 0.1101, + "step": 1256 + }, + { + "epoch": 0.9041539291494336, + "grad_norm": 8.385866965279517, + "learning_rate": 2.94413220827269e-06, + "loss": 0.1371, + "step": 1257 + }, + { + "epoch": 0.9048732242402445, + "grad_norm": 8.72447130184926, + "learning_rate": 2.9440401771212016e-06, + "loss": 0.1705, + "step": 1258 + }, + { + "epoch": 0.9055925193310556, + "grad_norm": 9.930171442491735, + "learning_rate": 2.943948071671311e-06, + "loss": 0.1952, + "step": 1259 + }, + { + "epoch": 0.9063118144218666, + "grad_norm": 6.506973826894642, + "learning_rate": 2.943855891927758e-06, + "loss": 0.3228, + "step": 1260 + }, + { + "epoch": 0.9070311095126776, + "grad_norm": 8.808876841120776, + "learning_rate": 2.943763637895284e-06, + "loss": 0.3505, + "step": 1261 + }, + { + "epoch": 0.9077504046034885, + "grad_norm": 6.216960179004871, + "learning_rate": 2.9436713095786365e-06, + "loss": 0.1968, + "step": 1262 + }, + { + "epoch": 0.9084696996942996, + "grad_norm": 4.13135181764248, + "learning_rate": 2.9435789069825663e-06, + "loss": 0.0744, + "step": 1263 + }, + { + "epoch": 0.9091889947851106, + "grad_norm": 5.358020926579018, + "learning_rate": 2.943486430111827e-06, + "loss": 0.053, + "step": 1264 + }, + { + "epoch": 0.9099082898759216, + "grad_norm": 7.161167042971359, + "learning_rate": 2.9433938789711763e-06, + "loss": 0.2794, + "step": 1265 + }, + { + "epoch": 0.9106275849667326, + "grad_norm": 8.033414353489725, + "learning_rate": 2.943301253565378e-06, + "loss": 0.1243, + "step": 1266 + }, + { + "epoch": 0.9113468800575436, + "grad_norm": 3.9047823003403646, + "learning_rate": 2.9432085538991954e-06, + "loss": 0.0507, + "step": 1267 + }, + { + "epoch": 0.9120661751483546, + "grad_norm": 5.256928859374807, + "learning_rate": 2.9431157799773998e-06, + "loss": 0.1042, + "step": 1268 + }, + { + "epoch": 0.9127854702391656, + "grad_norm": 4.162360614440319, + "learning_rate": 2.9430229318047638e-06, + "loss": 0.0951, + "step": 1269 + }, + { + "epoch": 0.9135047653299766, + "grad_norm": 3.6772985401229925, + "learning_rate": 2.9429300093860646e-06, + "loss": 0.0914, + "step": 1270 + }, + { + "epoch": 0.9142240604207876, + "grad_norm": 5.549020067543673, + "learning_rate": 2.942837012726084e-06, + "loss": 0.0981, + "step": 1271 + }, + { + "epoch": 0.9149433555115987, + "grad_norm": 5.554176913634114, + "learning_rate": 2.942743941829606e-06, + "loss": 0.1003, + "step": 1272 + }, + { + "epoch": 0.9156626506024096, + "grad_norm": 5.843817959520465, + "learning_rate": 2.9426507967014195e-06, + "loss": 0.2313, + "step": 1273 + }, + { + "epoch": 0.9163819456932206, + "grad_norm": 2.45356412532244, + "learning_rate": 2.942557577346317e-06, + "loss": 0.0304, + "step": 1274 + }, + { + "epoch": 0.9171012407840317, + "grad_norm": 6.255491047619371, + "learning_rate": 2.942464283769095e-06, + "loss": 0.2095, + "step": 1275 + }, + { + "epoch": 0.9178205358748427, + "grad_norm": 3.467736850193894, + "learning_rate": 2.942370915974553e-06, + "loss": 0.0842, + "step": 1276 + }, + { + "epoch": 0.9185398309656536, + "grad_norm": 8.9891249190695, + "learning_rate": 2.942277473967496e-06, + "loss": 0.1419, + "step": 1277 + }, + { + "epoch": 0.9192591260564646, + "grad_norm": 6.123182947956828, + "learning_rate": 2.9421839577527303e-06, + "loss": 0.264, + "step": 1278 + }, + { + "epoch": 0.9199784211472757, + "grad_norm": 4.592559339279738, + "learning_rate": 2.942090367335069e-06, + "loss": 0.1805, + "step": 1279 + }, + { + "epoch": 0.9206977162380867, + "grad_norm": 3.202767606028297, + "learning_rate": 2.9419967027193267e-06, + "loss": 0.0861, + "step": 1280 + }, + { + "epoch": 0.9214170113288976, + "grad_norm": 3.990044003740237, + "learning_rate": 2.9419029639103224e-06, + "loss": 0.0984, + "step": 1281 + }, + { + "epoch": 0.9221363064197087, + "grad_norm": 6.3767280819076175, + "learning_rate": 2.94180915091288e-06, + "loss": 0.2484, + "step": 1282 + }, + { + "epoch": 0.9228556015105197, + "grad_norm": 2.395642048952248, + "learning_rate": 2.941715263731825e-06, + "loss": 0.0637, + "step": 1283 + }, + { + "epoch": 0.9235748966013307, + "grad_norm": 4.72161436848947, + "learning_rate": 2.9416213023719897e-06, + "loss": 0.1272, + "step": 1284 + }, + { + "epoch": 0.9242941916921417, + "grad_norm": 5.226652822577646, + "learning_rate": 2.941527266838207e-06, + "loss": 0.1075, + "step": 1285 + }, + { + "epoch": 0.9250134867829527, + "grad_norm": 9.29996368039424, + "learning_rate": 2.941433157135316e-06, + "loss": 0.2204, + "step": 1286 + }, + { + "epoch": 0.9257327818737637, + "grad_norm": 3.7379044946969633, + "learning_rate": 2.9413389732681587e-06, + "loss": 0.0442, + "step": 1287 + }, + { + "epoch": 0.9264520769645748, + "grad_norm": 4.56554867619894, + "learning_rate": 2.9412447152415814e-06, + "loss": 0.104, + "step": 1288 + }, + { + "epoch": 0.9271713720553857, + "grad_norm": 5.042743762018386, + "learning_rate": 2.941150383060433e-06, + "loss": 0.1619, + "step": 1289 + }, + { + "epoch": 0.9278906671461967, + "grad_norm": 4.87798369899283, + "learning_rate": 2.9410559767295678e-06, + "loss": 0.2154, + "step": 1290 + }, + { + "epoch": 0.9286099622370078, + "grad_norm": 3.575254223545318, + "learning_rate": 2.9409614962538426e-06, + "loss": 0.1993, + "step": 1291 + }, + { + "epoch": 0.9293292573278188, + "grad_norm": 5.75267092343592, + "learning_rate": 2.9408669416381194e-06, + "loss": 0.2261, + "step": 1292 + }, + { + "epoch": 0.9300485524186297, + "grad_norm": 5.991164848122601, + "learning_rate": 2.9407723128872626e-06, + "loss": 0.2486, + "step": 1293 + }, + { + "epoch": 0.9307678475094407, + "grad_norm": 7.90291872679686, + "learning_rate": 2.940677610006141e-06, + "loss": 0.0533, + "step": 1294 + }, + { + "epoch": 0.9314871426002518, + "grad_norm": 5.052033192637415, + "learning_rate": 2.940582832999627e-06, + "loss": 0.1146, + "step": 1295 + }, + { + "epoch": 0.9322064376910627, + "grad_norm": 4.6764772861453165, + "learning_rate": 2.9404879818725976e-06, + "loss": 0.2488, + "step": 1296 + }, + { + "epoch": 0.9329257327818737, + "grad_norm": 3.854389510193265, + "learning_rate": 2.940393056629933e-06, + "loss": 0.1619, + "step": 1297 + }, + { + "epoch": 0.9336450278726848, + "grad_norm": 5.171191248763344, + "learning_rate": 2.9402980572765174e-06, + "loss": 0.1611, + "step": 1298 + }, + { + "epoch": 0.9343643229634958, + "grad_norm": 4.043876190598902, + "learning_rate": 2.9402029838172375e-06, + "loss": 0.0957, + "step": 1299 + }, + { + "epoch": 0.9350836180543067, + "grad_norm": 4.263483569515798, + "learning_rate": 2.9401078362569865e-06, + "loss": 0.0244, + "step": 1300 + }, + { + "epoch": 0.9358029131451178, + "grad_norm": 8.249234951837312, + "learning_rate": 2.9400126146006595e-06, + "loss": 0.2394, + "step": 1301 + }, + { + "epoch": 0.9365222082359288, + "grad_norm": 7.578433066199969, + "learning_rate": 2.939917318853155e-06, + "loss": 0.1882, + "step": 1302 + }, + { + "epoch": 0.9372415033267398, + "grad_norm": 7.045618904804449, + "learning_rate": 2.9398219490193773e-06, + "loss": 0.0711, + "step": 1303 + }, + { + "epoch": 0.9379607984175508, + "grad_norm": 6.621477538424895, + "learning_rate": 2.939726505104232e-06, + "loss": 0.2995, + "step": 1304 + }, + { + "epoch": 0.9386800935083618, + "grad_norm": 4.87428226046337, + "learning_rate": 2.9396309871126317e-06, + "loss": 0.1381, + "step": 1305 + }, + { + "epoch": 0.9393993885991728, + "grad_norm": 3.9012964794409446, + "learning_rate": 2.9395353950494893e-06, + "loss": 0.0381, + "step": 1306 + }, + { + "epoch": 0.9401186836899839, + "grad_norm": 4.528116812568083, + "learning_rate": 2.939439728919724e-06, + "loss": 0.107, + "step": 1307 + }, + { + "epoch": 0.9408379787807948, + "grad_norm": 8.494676761221033, + "learning_rate": 2.939343988728257e-06, + "loss": 0.3946, + "step": 1308 + }, + { + "epoch": 0.9415572738716058, + "grad_norm": 3.663528762661194, + "learning_rate": 2.9392481744800163e-06, + "loss": 0.0892, + "step": 1309 + }, + { + "epoch": 0.9422765689624168, + "grad_norm": 6.238754791163149, + "learning_rate": 2.93915228617993e-06, + "loss": 0.0526, + "step": 1310 + }, + { + "epoch": 0.9429958640532279, + "grad_norm": 4.424049831857157, + "learning_rate": 2.9390563238329324e-06, + "loss": 0.0841, + "step": 1311 + }, + { + "epoch": 0.9437151591440388, + "grad_norm": 0.6020758718317281, + "learning_rate": 2.9389602874439607e-06, + "loss": 0.0019, + "step": 1312 + }, + { + "epoch": 0.9444344542348498, + "grad_norm": 6.779203769438188, + "learning_rate": 2.9388641770179557e-06, + "loss": 0.2314, + "step": 1313 + }, + { + "epoch": 0.9451537493256609, + "grad_norm": 6.937282478794683, + "learning_rate": 2.938767992559863e-06, + "loss": 0.1591, + "step": 1314 + }, + { + "epoch": 0.9458730444164719, + "grad_norm": 6.234079077051618, + "learning_rate": 2.938671734074632e-06, + "loss": 0.1834, + "step": 1315 + }, + { + "epoch": 0.9465923395072828, + "grad_norm": 6.841309775424993, + "learning_rate": 2.938575401567215e-06, + "loss": 0.2178, + "step": 1316 + }, + { + "epoch": 0.9473116345980939, + "grad_norm": 9.602528586623908, + "learning_rate": 2.9384789950425673e-06, + "loss": 0.2306, + "step": 1317 + }, + { + "epoch": 0.9480309296889049, + "grad_norm": 6.005471809326496, + "learning_rate": 2.9383825145056503e-06, + "loss": 0.0187, + "step": 1318 + }, + { + "epoch": 0.9487502247797159, + "grad_norm": 2.657277897645906, + "learning_rate": 2.9382859599614284e-06, + "loss": 0.0502, + "step": 1319 + }, + { + "epoch": 0.9494695198705269, + "grad_norm": 8.775551327657592, + "learning_rate": 2.9381893314148687e-06, + "loss": 0.2498, + "step": 1320 + }, + { + "epoch": 0.9501888149613379, + "grad_norm": 5.698308629174225, + "learning_rate": 2.9380926288709437e-06, + "loss": 0.1002, + "step": 1321 + }, + { + "epoch": 0.9509081100521489, + "grad_norm": 5.1383003657534765, + "learning_rate": 2.9379958523346283e-06, + "loss": 0.1909, + "step": 1322 + }, + { + "epoch": 0.9516274051429598, + "grad_norm": 8.798398440645089, + "learning_rate": 2.9378990018109014e-06, + "loss": 0.0802, + "step": 1323 + }, + { + "epoch": 0.9523467002337709, + "grad_norm": 4.914679586563055, + "learning_rate": 2.937802077304747e-06, + "loss": 0.1796, + "step": 1324 + }, + { + "epoch": 0.9530659953245819, + "grad_norm": 8.312203314162511, + "learning_rate": 2.937705078821152e-06, + "loss": 0.095, + "step": 1325 + }, + { + "epoch": 0.953785290415393, + "grad_norm": 3.6172836222284883, + "learning_rate": 2.937608006365107e-06, + "loss": 0.0435, + "step": 1326 + }, + { + "epoch": 0.9545045855062039, + "grad_norm": 5.043999173881439, + "learning_rate": 2.9375108599416062e-06, + "loss": 0.1614, + "step": 1327 + }, + { + "epoch": 0.9552238805970149, + "grad_norm": 6.349023434348644, + "learning_rate": 2.9374136395556477e-06, + "loss": 0.103, + "step": 1328 + }, + { + "epoch": 0.9559431756878259, + "grad_norm": 5.276008530527816, + "learning_rate": 2.937316345212235e-06, + "loss": 0.1666, + "step": 1329 + }, + { + "epoch": 0.956662470778637, + "grad_norm": 6.163055949919098, + "learning_rate": 2.9372189769163726e-06, + "loss": 0.1866, + "step": 1330 + }, + { + "epoch": 0.9573817658694479, + "grad_norm": 3.929252611917299, + "learning_rate": 2.937121534673071e-06, + "loss": 0.0842, + "step": 1331 + }, + { + "epoch": 0.9581010609602589, + "grad_norm": 9.238766847364214, + "learning_rate": 2.937024018487344e-06, + "loss": 0.2874, + "step": 1332 + }, + { + "epoch": 0.95882035605107, + "grad_norm": 4.538855438518856, + "learning_rate": 2.936926428364208e-06, + "loss": 0.1823, + "step": 1333 + }, + { + "epoch": 0.959539651141881, + "grad_norm": 2.543646867943277, + "learning_rate": 2.936828764308685e-06, + "loss": 0.1089, + "step": 1334 + }, + { + "epoch": 0.9602589462326919, + "grad_norm": 4.442979294694012, + "learning_rate": 2.9367310263258e-06, + "loss": 0.1321, + "step": 1335 + }, + { + "epoch": 0.960978241323503, + "grad_norm": 5.115615296348857, + "learning_rate": 2.936633214420582e-06, + "loss": 0.0692, + "step": 1336 + }, + { + "epoch": 0.961697536414314, + "grad_norm": 4.844818887072438, + "learning_rate": 2.936535328598063e-06, + "loss": 0.2071, + "step": 1337 + }, + { + "epoch": 0.962416831505125, + "grad_norm": 5.909470102807963, + "learning_rate": 2.936437368863279e-06, + "loss": 0.1876, + "step": 1338 + }, + { + "epoch": 0.963136126595936, + "grad_norm": 8.972473629729977, + "learning_rate": 2.9363393352212715e-06, + "loss": 0.3448, + "step": 1339 + }, + { + "epoch": 0.963855421686747, + "grad_norm": 5.185373699215868, + "learning_rate": 2.9362412276770835e-06, + "loss": 0.1129, + "step": 1340 + }, + { + "epoch": 0.964574716777558, + "grad_norm": 2.1134832222454696, + "learning_rate": 2.9361430462357625e-06, + "loss": 0.0257, + "step": 1341 + }, + { + "epoch": 0.965294011868369, + "grad_norm": 3.2378638520482026, + "learning_rate": 2.9360447909023615e-06, + "loss": 0.0285, + "step": 1342 + }, + { + "epoch": 0.96601330695918, + "grad_norm": 3.2743449072117907, + "learning_rate": 2.935946461681935e-06, + "loss": 0.0225, + "step": 1343 + }, + { + "epoch": 0.966732602049991, + "grad_norm": 4.9159760654342435, + "learning_rate": 2.9358480585795416e-06, + "loss": 0.1496, + "step": 1344 + }, + { + "epoch": 0.967451897140802, + "grad_norm": 1.8003457505524643, + "learning_rate": 2.935749581600246e-06, + "loss": 0.0152, + "step": 1345 + }, + { + "epoch": 0.968171192231613, + "grad_norm": 3.3679058981093135, + "learning_rate": 2.9356510307491135e-06, + "loss": 0.0885, + "step": 1346 + }, + { + "epoch": 0.968890487322424, + "grad_norm": 3.2417301776507714, + "learning_rate": 2.9355524060312155e-06, + "loss": 0.0298, + "step": 1347 + }, + { + "epoch": 0.969609782413235, + "grad_norm": 6.817446565799031, + "learning_rate": 2.935453707451626e-06, + "loss": 0.2995, + "step": 1348 + }, + { + "epoch": 0.9703290775040461, + "grad_norm": 6.2293619177312225, + "learning_rate": 2.9353549350154236e-06, + "loss": 0.1859, + "step": 1349 + }, + { + "epoch": 0.971048372594857, + "grad_norm": 4.873471336724789, + "learning_rate": 2.93525608872769e-06, + "loss": 0.1807, + "step": 1350 + }, + { + "epoch": 0.971767667685668, + "grad_norm": 7.62690882822264, + "learning_rate": 2.935157168593511e-06, + "loss": 0.034, + "step": 1351 + }, + { + "epoch": 0.9724869627764791, + "grad_norm": 5.999300008855233, + "learning_rate": 2.9350581746179764e-06, + "loss": 0.1917, + "step": 1352 + }, + { + "epoch": 0.9732062578672901, + "grad_norm": 4.680982964250305, + "learning_rate": 2.9349591068061797e-06, + "loss": 0.0774, + "step": 1353 + }, + { + "epoch": 0.973925552958101, + "grad_norm": 5.22320571231796, + "learning_rate": 2.9348599651632177e-06, + "loss": 0.157, + "step": 1354 + }, + { + "epoch": 0.974644848048912, + "grad_norm": 5.632876424360375, + "learning_rate": 2.934760749694192e-06, + "loss": 0.2553, + "step": 1355 + }, + { + "epoch": 0.9753641431397231, + "grad_norm": 6.096409888625927, + "learning_rate": 2.934661460404207e-06, + "loss": 0.1513, + "step": 1356 + }, + { + "epoch": 0.9760834382305341, + "grad_norm": 4.187139932755058, + "learning_rate": 2.9345620972983715e-06, + "loss": 0.0236, + "step": 1357 + }, + { + "epoch": 0.976802733321345, + "grad_norm": 5.64497313491945, + "learning_rate": 2.9344626603817972e-06, + "loss": 0.1223, + "step": 1358 + }, + { + "epoch": 0.9775220284121561, + "grad_norm": 11.614650134794717, + "learning_rate": 2.9343631496596014e-06, + "loss": 0.2071, + "step": 1359 + }, + { + "epoch": 0.9782413235029671, + "grad_norm": 5.196556178766472, + "learning_rate": 2.9342635651369033e-06, + "loss": 0.1035, + "step": 1360 + }, + { + "epoch": 0.9789606185937781, + "grad_norm": 8.2711162132498, + "learning_rate": 2.9341639068188275e-06, + "loss": 0.1057, + "step": 1361 + }, + { + "epoch": 0.9796799136845891, + "grad_norm": 5.906295648441473, + "learning_rate": 2.934064174710501e-06, + "loss": 0.1331, + "step": 1362 + }, + { + "epoch": 0.9803992087754001, + "grad_norm": 5.165080378088079, + "learning_rate": 2.933964368817055e-06, + "loss": 0.04, + "step": 1363 + }, + { + "epoch": 0.9811185038662111, + "grad_norm": 3.4052673277614267, + "learning_rate": 2.933864489143625e-06, + "loss": 0.0135, + "step": 1364 + }, + { + "epoch": 0.9818377989570222, + "grad_norm": 4.38275133226788, + "learning_rate": 2.9337645356953496e-06, + "loss": 0.124, + "step": 1365 + }, + { + "epoch": 0.9825570940478331, + "grad_norm": 5.412869994564405, + "learning_rate": 2.9336645084773725e-06, + "loss": 0.1408, + "step": 1366 + }, + { + "epoch": 0.9832763891386441, + "grad_norm": 5.028367265133365, + "learning_rate": 2.93356440749484e-06, + "loss": 0.2062, + "step": 1367 + }, + { + "epoch": 0.9839956842294552, + "grad_norm": 1.3784656980872196, + "learning_rate": 2.9334642327529016e-06, + "loss": 0.0093, + "step": 1368 + }, + { + "epoch": 0.9847149793202662, + "grad_norm": 5.78242422254892, + "learning_rate": 2.9333639842567123e-06, + "loss": 0.1425, + "step": 1369 + }, + { + "epoch": 0.9854342744110771, + "grad_norm": 7.063427648312734, + "learning_rate": 2.93326366201143e-06, + "loss": 0.2509, + "step": 1370 + }, + { + "epoch": 0.9861535695018881, + "grad_norm": 5.499604929505757, + "learning_rate": 2.9331632660222156e-06, + "loss": 0.0416, + "step": 1371 + }, + { + "epoch": 0.9868728645926992, + "grad_norm": 5.263638780634471, + "learning_rate": 2.933062796294236e-06, + "loss": 0.1209, + "step": 1372 + }, + { + "epoch": 0.9875921596835101, + "grad_norm": 2.6796228443468735, + "learning_rate": 2.9329622528326596e-06, + "loss": 0.0206, + "step": 1373 + }, + { + "epoch": 0.9883114547743211, + "grad_norm": 4.00930218762538, + "learning_rate": 2.9328616356426597e-06, + "loss": 0.1217, + "step": 1374 + }, + { + "epoch": 0.9890307498651322, + "grad_norm": 6.193822314405915, + "learning_rate": 2.932760944729414e-06, + "loss": 0.2707, + "step": 1375 + }, + { + "epoch": 0.9897500449559432, + "grad_norm": 6.603153096357186, + "learning_rate": 2.9326601800981015e-06, + "loss": 0.1886, + "step": 1376 + }, + { + "epoch": 0.9904693400467541, + "grad_norm": 5.059592451445477, + "learning_rate": 2.9325593417539088e-06, + "loss": 0.1363, + "step": 1377 + }, + { + "epoch": 0.9911886351375652, + "grad_norm": 4.818312944074029, + "learning_rate": 2.9324584297020228e-06, + "loss": 0.176, + "step": 1378 + }, + { + "epoch": 0.9919079302283762, + "grad_norm": 4.182412138337019, + "learning_rate": 2.9323574439476354e-06, + "loss": 0.1295, + "step": 1379 + }, + { + "epoch": 0.9926272253191872, + "grad_norm": 5.4954889159296805, + "learning_rate": 2.9322563844959438e-06, + "loss": 0.0343, + "step": 1380 + }, + { + "epoch": 0.9933465204099982, + "grad_norm": 4.712358215324916, + "learning_rate": 2.9321552513521472e-06, + "loss": 0.0926, + "step": 1381 + }, + { + "epoch": 0.9940658155008092, + "grad_norm": 7.396615357648573, + "learning_rate": 2.9320540445214483e-06, + "loss": 0.1132, + "step": 1382 + }, + { + "epoch": 0.9947851105916202, + "grad_norm": 7.973474165631346, + "learning_rate": 2.931952764009055e-06, + "loss": 0.2409, + "step": 1383 + }, + { + "epoch": 0.9955044056824313, + "grad_norm": 8.781700307611532, + "learning_rate": 2.9318514098201783e-06, + "loss": 0.0689, + "step": 1384 + }, + { + "epoch": 0.9962237007732422, + "grad_norm": 8.515470094618632, + "learning_rate": 2.9317499819600332e-06, + "loss": 0.1517, + "step": 1385 + }, + { + "epoch": 0.9969429958640532, + "grad_norm": 4.958434198085519, + "learning_rate": 2.9316484804338377e-06, + "loss": 0.0607, + "step": 1386 + }, + { + "epoch": 0.9976622909548643, + "grad_norm": 7.187995371338799, + "learning_rate": 2.931546905246815e-06, + "loss": 0.0821, + "step": 1387 + }, + { + "epoch": 0.9983815860456753, + "grad_norm": 4.784326353445957, + "learning_rate": 2.9314452564041913e-06, + "loss": 0.0671, + "step": 1388 + }, + { + "epoch": 0.9991008811364862, + "grad_norm": 7.122956004731571, + "learning_rate": 2.9313435339111957e-06, + "loss": 0.0955, + "step": 1389 + }, + { + "epoch": 0.9998201762272972, + "grad_norm": 4.884861312124538, + "learning_rate": 2.9312417377730633e-06, + "loss": 0.1574, + "step": 1390 + }, + { + "epoch": 1.0005394713181082, + "grad_norm": 2.3247312465883416, + "learning_rate": 2.9311398679950304e-06, + "loss": 0.0058, + "step": 1391 + }, + { + "epoch": 1.0012587664089192, + "grad_norm": 4.477142552949364, + "learning_rate": 2.9310379245823397e-06, + "loss": 0.0387, + "step": 1392 + }, + { + "epoch": 1.0019780614997302, + "grad_norm": 5.628018056361989, + "learning_rate": 2.9309359075402347e-06, + "loss": 0.1208, + "step": 1393 + }, + { + "epoch": 1.0026973565905413, + "grad_norm": 7.118604766685235, + "learning_rate": 2.9308338168739663e-06, + "loss": 0.1335, + "step": 1394 + }, + { + "epoch": 1.0034166516813523, + "grad_norm": 5.515477812340682, + "learning_rate": 2.9307316525887855e-06, + "loss": 0.0396, + "step": 1395 + }, + { + "epoch": 1.0041359467721633, + "grad_norm": 2.789903061554602, + "learning_rate": 2.9306294146899504e-06, + "loss": 0.0532, + "step": 1396 + }, + { + "epoch": 1.0048552418629744, + "grad_norm": 5.143654984850947, + "learning_rate": 2.93052710318272e-06, + "loss": 0.1297, + "step": 1397 + }, + { + "epoch": 1.0055745369537852, + "grad_norm": 5.494299138391224, + "learning_rate": 2.9304247180723595e-06, + "loss": 0.1874, + "step": 1398 + }, + { + "epoch": 1.0062938320445962, + "grad_norm": 2.4251399029339287, + "learning_rate": 2.9303222593641357e-06, + "loss": 0.0112, + "step": 1399 + }, + { + "epoch": 1.0070131271354072, + "grad_norm": 3.6563007601426576, + "learning_rate": 2.930219727063321e-06, + "loss": 0.0721, + "step": 1400 + }, + { + "epoch": 1.0077324222262183, + "grad_norm": 3.487920433125351, + "learning_rate": 2.9301171211751904e-06, + "loss": 0.1469, + "step": 1401 + }, + { + "epoch": 1.0084517173170293, + "grad_norm": 7.97052461925232, + "learning_rate": 2.9300144417050237e-06, + "loss": 0.1024, + "step": 1402 + }, + { + "epoch": 1.0091710124078404, + "grad_norm": 6.148267670307876, + "learning_rate": 2.9299116886581032e-06, + "loss": 0.0484, + "step": 1403 + }, + { + "epoch": 1.0098903074986514, + "grad_norm": 6.740993747586648, + "learning_rate": 2.9298088620397166e-06, + "loss": 0.1861, + "step": 1404 + }, + { + "epoch": 1.0106096025894624, + "grad_norm": 6.533819187784662, + "learning_rate": 2.929705961855154e-06, + "loss": 0.1007, + "step": 1405 + }, + { + "epoch": 1.0113288976802732, + "grad_norm": 3.6454979506074268, + "learning_rate": 2.9296029881097104e-06, + "loss": 0.1467, + "step": 1406 + }, + { + "epoch": 1.0120481927710843, + "grad_norm": 3.0240831331690945, + "learning_rate": 2.9294999408086825e-06, + "loss": 0.0781, + "step": 1407 + }, + { + "epoch": 1.0127674878618953, + "grad_norm": 5.737698595041257, + "learning_rate": 2.929396819957374e-06, + "loss": 0.1157, + "step": 1408 + }, + { + "epoch": 1.0134867829527063, + "grad_norm": 7.86037923693671, + "learning_rate": 2.92929362556109e-06, + "loss": 0.0637, + "step": 1409 + }, + { + "epoch": 1.0142060780435174, + "grad_norm": 1.3225909553719413, + "learning_rate": 2.9291903576251394e-06, + "loss": 0.0065, + "step": 1410 + }, + { + "epoch": 1.0149253731343284, + "grad_norm": 10.01505367850342, + "learning_rate": 2.929087016154836e-06, + "loss": 0.2248, + "step": 1411 + }, + { + "epoch": 1.0156446682251394, + "grad_norm": 3.3701173192710914, + "learning_rate": 2.928983601155497e-06, + "loss": 0.0091, + "step": 1412 + }, + { + "epoch": 1.0163639633159505, + "grad_norm": 4.647333162567701, + "learning_rate": 2.928880112632443e-06, + "loss": 0.069, + "step": 1413 + }, + { + "epoch": 1.0170832584067613, + "grad_norm": 2.105368285009397, + "learning_rate": 2.9287765505909995e-06, + "loss": 0.0105, + "step": 1414 + }, + { + "epoch": 1.0178025534975723, + "grad_norm": 7.194997295623576, + "learning_rate": 2.928672915036494e-06, + "loss": 0.2147, + "step": 1415 + }, + { + "epoch": 1.0185218485883833, + "grad_norm": 3.9545142095715553, + "learning_rate": 2.928569205974259e-06, + "loss": 0.1553, + "step": 1416 + }, + { + "epoch": 1.0192411436791944, + "grad_norm": 2.6531451827293253, + "learning_rate": 2.928465423409631e-06, + "loss": 0.0178, + "step": 1417 + }, + { + "epoch": 1.0199604387700054, + "grad_norm": 6.907830394938768, + "learning_rate": 2.928361567347949e-06, + "loss": 0.2328, + "step": 1418 + }, + { + "epoch": 1.0206797338608165, + "grad_norm": 4.547012339078066, + "learning_rate": 2.928257637794557e-06, + "loss": 0.1655, + "step": 1419 + }, + { + "epoch": 1.0213990289516275, + "grad_norm": 10.334031782391865, + "learning_rate": 2.9281536347548024e-06, + "loss": 0.098, + "step": 1420 + }, + { + "epoch": 1.0221183240424385, + "grad_norm": 7.75808816296915, + "learning_rate": 2.928049558234036e-06, + "loss": 0.1627, + "step": 1421 + }, + { + "epoch": 1.0228376191332493, + "grad_norm": 6.071793791720597, + "learning_rate": 2.9279454082376134e-06, + "loss": 0.1057, + "step": 1422 + }, + { + "epoch": 1.0235569142240604, + "grad_norm": 4.817435916806376, + "learning_rate": 2.9278411847708928e-06, + "loss": 0.1751, + "step": 1423 + }, + { + "epoch": 1.0242762093148714, + "grad_norm": 1.3726063175249814, + "learning_rate": 2.9277368878392365e-06, + "loss": 0.0248, + "step": 1424 + }, + { + "epoch": 1.0249955044056824, + "grad_norm": 5.6378240333896565, + "learning_rate": 2.927632517448011e-06, + "loss": 0.134, + "step": 1425 + }, + { + "epoch": 1.0257147994964935, + "grad_norm": 5.4055841455839255, + "learning_rate": 2.9275280736025864e-06, + "loss": 0.119, + "step": 1426 + }, + { + "epoch": 1.0264340945873045, + "grad_norm": 4.493895074694661, + "learning_rate": 2.9274235563083365e-06, + "loss": 0.2053, + "step": 1427 + }, + { + "epoch": 1.0271533896781155, + "grad_norm": 4.274873372164417, + "learning_rate": 2.927318965570639e-06, + "loss": 0.0457, + "step": 1428 + }, + { + "epoch": 1.0278726847689263, + "grad_norm": 7.900905239687231, + "learning_rate": 2.9272143013948747e-06, + "loss": 0.2692, + "step": 1429 + }, + { + "epoch": 1.0285919798597374, + "grad_norm": 3.793530770988621, + "learning_rate": 2.9271095637864294e-06, + "loss": 0.0985, + "step": 1430 + }, + { + "epoch": 1.0293112749505484, + "grad_norm": 8.240289582051897, + "learning_rate": 2.9270047527506922e-06, + "loss": 0.2147, + "step": 1431 + }, + { + "epoch": 1.0300305700413595, + "grad_norm": 4.551649957537649, + "learning_rate": 2.926899868293055e-06, + "loss": 0.0983, + "step": 1432 + }, + { + "epoch": 1.0307498651321705, + "grad_norm": 7.346732655405621, + "learning_rate": 2.926794910418915e-06, + "loss": 0.0999, + "step": 1433 + }, + { + "epoch": 1.0314691602229815, + "grad_norm": 6.0936634645168475, + "learning_rate": 2.926689879133672e-06, + "loss": 0.2177, + "step": 1434 + }, + { + "epoch": 1.0321884553137926, + "grad_norm": 6.091105478119522, + "learning_rate": 2.9265847744427307e-06, + "loss": 0.1846, + "step": 1435 + }, + { + "epoch": 1.0329077504046036, + "grad_norm": 4.057924498885607, + "learning_rate": 2.9264795963514983e-06, + "loss": 0.1894, + "step": 1436 + }, + { + "epoch": 1.0336270454954144, + "grad_norm": 4.9917784186657235, + "learning_rate": 2.926374344865386e-06, + "loss": 0.189, + "step": 1437 + }, + { + "epoch": 1.0343463405862254, + "grad_norm": 7.027690058970331, + "learning_rate": 2.92626901998981e-06, + "loss": 0.1647, + "step": 1438 + }, + { + "epoch": 1.0350656356770365, + "grad_norm": 12.398962184859455, + "learning_rate": 2.9261636217301895e-06, + "loss": 0.232, + "step": 1439 + }, + { + "epoch": 1.0357849307678475, + "grad_norm": 6.077292551731999, + "learning_rate": 2.9260581500919465e-06, + "loss": 0.0736, + "step": 1440 + }, + { + "epoch": 1.0365042258586585, + "grad_norm": 7.571167475639318, + "learning_rate": 2.9259526050805087e-06, + "loss": 0.0668, + "step": 1441 + }, + { + "epoch": 1.0372235209494696, + "grad_norm": 4.388295734597306, + "learning_rate": 2.9258469867013064e-06, + "loss": 0.0775, + "step": 1442 + }, + { + "epoch": 1.0379428160402806, + "grad_norm": 7.09977324931201, + "learning_rate": 2.9257412949597733e-06, + "loss": 0.3884, + "step": 1443 + }, + { + "epoch": 1.0386621111310916, + "grad_norm": 6.287338747096999, + "learning_rate": 2.925635529861348e-06, + "loss": 0.3921, + "step": 1444 + }, + { + "epoch": 1.0393814062219024, + "grad_norm": 5.847406918683498, + "learning_rate": 2.925529691411472e-06, + "loss": 0.2147, + "step": 1445 + }, + { + "epoch": 1.0401007013127135, + "grad_norm": 6.180789413120241, + "learning_rate": 2.925423779615591e-06, + "loss": 0.0817, + "step": 1446 + }, + { + "epoch": 1.0408199964035245, + "grad_norm": 6.7466227065199815, + "learning_rate": 2.925317794479154e-06, + "loss": 0.0836, + "step": 1447 + }, + { + "epoch": 1.0415392914943356, + "grad_norm": 7.155673514386755, + "learning_rate": 2.9252117360076146e-06, + "loss": 0.0498, + "step": 1448 + }, + { + "epoch": 1.0422585865851466, + "grad_norm": 5.530935444079844, + "learning_rate": 2.9251056042064297e-06, + "loss": 0.1739, + "step": 1449 + }, + { + "epoch": 1.0429778816759576, + "grad_norm": 4.001087623476624, + "learning_rate": 2.9249993990810598e-06, + "loss": 0.1431, + "step": 1450 + }, + { + "epoch": 1.0436971767667687, + "grad_norm": 1.0575289350855142, + "learning_rate": 2.924893120636969e-06, + "loss": 0.0061, + "step": 1451 + }, + { + "epoch": 1.0444164718575795, + "grad_norm": 3.468307518976385, + "learning_rate": 2.9247867688796257e-06, + "loss": 0.1044, + "step": 1452 + }, + { + "epoch": 1.0451357669483905, + "grad_norm": 6.469066752340097, + "learning_rate": 2.9246803438145023e-06, + "loss": 0.0339, + "step": 1453 + }, + { + "epoch": 1.0458550620392015, + "grad_norm": 6.6049055395458955, + "learning_rate": 2.9245738454470744e-06, + "loss": 0.0855, + "step": 1454 + }, + { + "epoch": 1.0465743571300126, + "grad_norm": 2.1504829050347958, + "learning_rate": 2.924467273782821e-06, + "loss": 0.0749, + "step": 1455 + }, + { + "epoch": 1.0472936522208236, + "grad_norm": 3.8041312945012264, + "learning_rate": 2.9243606288272257e-06, + "loss": 0.1165, + "step": 1456 + }, + { + "epoch": 1.0480129473116346, + "grad_norm": 7.742287880125306, + "learning_rate": 2.924253910585776e-06, + "loss": 0.1575, + "step": 1457 + }, + { + "epoch": 1.0487322424024457, + "grad_norm": 0.7523721604658536, + "learning_rate": 2.9241471190639624e-06, + "loss": 0.0028, + "step": 1458 + }, + { + "epoch": 1.0494515374932567, + "grad_norm": 3.5971517352750255, + "learning_rate": 2.924040254267279e-06, + "loss": 0.0175, + "step": 1459 + }, + { + "epoch": 1.0501708325840675, + "grad_norm": 4.0900087429273855, + "learning_rate": 2.9239333162012256e-06, + "loss": 0.0356, + "step": 1460 + }, + { + "epoch": 1.0508901276748785, + "grad_norm": 5.512019872798274, + "learning_rate": 2.923826304871303e-06, + "loss": 0.2396, + "step": 1461 + }, + { + "epoch": 1.0516094227656896, + "grad_norm": 4.543217051628713, + "learning_rate": 2.9237192202830173e-06, + "loss": 0.0946, + "step": 1462 + }, + { + "epoch": 1.0523287178565006, + "grad_norm": 5.630736402167831, + "learning_rate": 2.9236120624418786e-06, + "loss": 0.1906, + "step": 1463 + }, + { + "epoch": 1.0530480129473117, + "grad_norm": 4.3672323352831794, + "learning_rate": 2.9235048313534007e-06, + "loss": 0.0378, + "step": 1464 + }, + { + "epoch": 1.0537673080381227, + "grad_norm": 5.25257814040437, + "learning_rate": 2.9233975270230997e-06, + "loss": 0.2043, + "step": 1465 + }, + { + "epoch": 1.0544866031289337, + "grad_norm": 3.5113107174612366, + "learning_rate": 2.9232901494564976e-06, + "loss": 0.0094, + "step": 1466 + }, + { + "epoch": 1.0552058982197448, + "grad_norm": 3.717395206264605, + "learning_rate": 2.9231826986591186e-06, + "loss": 0.0953, + "step": 1467 + }, + { + "epoch": 1.0559251933105556, + "grad_norm": 4.042386265159079, + "learning_rate": 2.923075174636492e-06, + "loss": 0.1294, + "step": 1468 + }, + { + "epoch": 1.0566444884013666, + "grad_norm": 4.6509046800507825, + "learning_rate": 2.9229675773941484e-06, + "loss": 0.0603, + "step": 1469 + }, + { + "epoch": 1.0573637834921776, + "grad_norm": 0.7541291078948386, + "learning_rate": 2.9228599069376257e-06, + "loss": 0.0055, + "step": 1470 + }, + { + "epoch": 1.0580830785829887, + "grad_norm": 5.113995253185113, + "learning_rate": 2.922752163272463e-06, + "loss": 0.1465, + "step": 1471 + }, + { + "epoch": 1.0588023736737997, + "grad_norm": 5.425080566314265, + "learning_rate": 2.922644346404204e-06, + "loss": 0.1132, + "step": 1472 + }, + { + "epoch": 1.0595216687646107, + "grad_norm": 3.1178686089635117, + "learning_rate": 2.922536456338396e-06, + "loss": 0.0899, + "step": 1473 + }, + { + "epoch": 1.0602409638554218, + "grad_norm": 4.261433285723393, + "learning_rate": 2.92242849308059e-06, + "loss": 0.1209, + "step": 1474 + }, + { + "epoch": 1.0609602589462326, + "grad_norm": 5.058830564921074, + "learning_rate": 2.9223204566363416e-06, + "loss": 0.2113, + "step": 1475 + }, + { + "epoch": 1.0616795540370436, + "grad_norm": 4.372576214669597, + "learning_rate": 2.922212347011208e-06, + "loss": 0.0579, + "step": 1476 + }, + { + "epoch": 1.0623988491278546, + "grad_norm": 2.6226259871011286, + "learning_rate": 2.922104164210753e-06, + "loss": 0.0617, + "step": 1477 + }, + { + "epoch": 1.0631181442186657, + "grad_norm": 5.23400879076255, + "learning_rate": 2.9219959082405422e-06, + "loss": 0.1924, + "step": 1478 + }, + { + "epoch": 1.0638374393094767, + "grad_norm": 1.9934632681874618, + "learning_rate": 2.921887579106146e-06, + "loss": 0.0348, + "step": 1479 + }, + { + "epoch": 1.0645567344002878, + "grad_norm": 5.656383427397238, + "learning_rate": 2.921779176813137e-06, + "loss": 0.2216, + "step": 1480 + }, + { + "epoch": 1.0652760294910988, + "grad_norm": 5.449406362596943, + "learning_rate": 2.9216707013670946e-06, + "loss": 0.2164, + "step": 1481 + }, + { + "epoch": 1.0659953245819098, + "grad_norm": 4.958180971987771, + "learning_rate": 2.9215621527735986e-06, + "loss": 0.038, + "step": 1482 + }, + { + "epoch": 1.0667146196727206, + "grad_norm": 4.064581350276271, + "learning_rate": 2.921453531038234e-06, + "loss": 0.2205, + "step": 1483 + }, + { + "epoch": 1.0674339147635317, + "grad_norm": 4.031473562279539, + "learning_rate": 2.92134483616659e-06, + "loss": 0.1879, + "step": 1484 + }, + { + "epoch": 1.0681532098543427, + "grad_norm": 6.274310679738218, + "learning_rate": 2.9212360681642592e-06, + "loss": 0.2501, + "step": 1485 + }, + { + "epoch": 1.0688725049451537, + "grad_norm": 6.095003922119308, + "learning_rate": 2.921127227036838e-06, + "loss": 0.0938, + "step": 1486 + }, + { + "epoch": 1.0695918000359648, + "grad_norm": 5.232846560734211, + "learning_rate": 2.921018312789926e-06, + "loss": 0.1341, + "step": 1487 + }, + { + "epoch": 1.0703110951267758, + "grad_norm": 5.001592695540452, + "learning_rate": 2.9209093254291273e-06, + "loss": 0.1873, + "step": 1488 + }, + { + "epoch": 1.0710303902175868, + "grad_norm": 4.035141328939517, + "learning_rate": 2.9208002649600493e-06, + "loss": 0.1061, + "step": 1489 + }, + { + "epoch": 1.0717496853083979, + "grad_norm": 4.649097168360496, + "learning_rate": 2.920691131388304e-06, + "loss": 0.0712, + "step": 1490 + }, + { + "epoch": 1.0724689803992087, + "grad_norm": 4.481316157897854, + "learning_rate": 2.920581924719506e-06, + "loss": 0.0394, + "step": 1491 + }, + { + "epoch": 1.0731882754900197, + "grad_norm": 6.06525938989988, + "learning_rate": 2.920472644959274e-06, + "loss": 0.0963, + "step": 1492 + }, + { + "epoch": 1.0739075705808308, + "grad_norm": 7.427434280609504, + "learning_rate": 2.920363292113231e-06, + "loss": 0.3338, + "step": 1493 + }, + { + "epoch": 1.0746268656716418, + "grad_norm": 3.4303618319189217, + "learning_rate": 2.9202538661870032e-06, + "loss": 0.0399, + "step": 1494 + }, + { + "epoch": 1.0753461607624528, + "grad_norm": 2.731194643778459, + "learning_rate": 2.9201443671862205e-06, + "loss": 0.0376, + "step": 1495 + }, + { + "epoch": 1.0760654558532639, + "grad_norm": 3.306613701716555, + "learning_rate": 2.920034795116517e-06, + "loss": 0.0546, + "step": 1496 + }, + { + "epoch": 1.0767847509440749, + "grad_norm": 2.8839141253524576, + "learning_rate": 2.9199251499835313e-06, + "loss": 0.0267, + "step": 1497 + }, + { + "epoch": 1.077504046034886, + "grad_norm": 6.925996889433103, + "learning_rate": 2.9198154317929035e-06, + "loss": 0.2802, + "step": 1498 + }, + { + "epoch": 1.0782233411256967, + "grad_norm": 3.489081656265809, + "learning_rate": 2.9197056405502795e-06, + "loss": 0.0969, + "step": 1499 + }, + { + "epoch": 1.0789426362165078, + "grad_norm": 5.419614479280475, + "learning_rate": 2.9195957762613083e-06, + "loss": 0.0788, + "step": 1500 + }, + { + "epoch": 1.0796619313073188, + "grad_norm": 5.577798786356071, + "learning_rate": 2.9194858389316416e-06, + "loss": 0.1749, + "step": 1501 + }, + { + "epoch": 1.0803812263981298, + "grad_norm": 4.539753135639085, + "learning_rate": 2.9193758285669373e-06, + "loss": 0.0986, + "step": 1502 + }, + { + "epoch": 1.0811005214889409, + "grad_norm": 4.9256584195434066, + "learning_rate": 2.9192657451728547e-06, + "loss": 0.1503, + "step": 1503 + }, + { + "epoch": 1.081819816579752, + "grad_norm": 3.3280021763159007, + "learning_rate": 2.9191555887550585e-06, + "loss": 0.1057, + "step": 1504 + }, + { + "epoch": 1.082539111670563, + "grad_norm": 3.263113229654212, + "learning_rate": 2.9190453593192156e-06, + "loss": 0.0222, + "step": 1505 + }, + { + "epoch": 1.0832584067613737, + "grad_norm": 3.839295795800084, + "learning_rate": 2.918935056870998e-06, + "loss": 0.131, + "step": 1506 + }, + { + "epoch": 1.0839777018521848, + "grad_norm": 3.4267481683413163, + "learning_rate": 2.9188246814160808e-06, + "loss": 0.0918, + "step": 1507 + }, + { + "epoch": 1.0846969969429958, + "grad_norm": 8.627027513030217, + "learning_rate": 2.9187142329601437e-06, + "loss": 0.3319, + "step": 1508 + }, + { + "epoch": 1.0854162920338069, + "grad_norm": 5.16670900444957, + "learning_rate": 2.918603711508868e-06, + "loss": 0.1224, + "step": 1509 + }, + { + "epoch": 1.0861355871246179, + "grad_norm": 6.5169930251932655, + "learning_rate": 2.9184931170679414e-06, + "loss": 0.0295, + "step": 1510 + }, + { + "epoch": 1.086854882215429, + "grad_norm": 7.975271216831541, + "learning_rate": 2.9183824496430535e-06, + "loss": 0.0333, + "step": 1511 + }, + { + "epoch": 1.08757417730624, + "grad_norm": 3.6150544552242723, + "learning_rate": 2.918271709239899e-06, + "loss": 0.1581, + "step": 1512 + }, + { + "epoch": 1.088293472397051, + "grad_norm": 3.736523644745436, + "learning_rate": 2.9181608958641756e-06, + "loss": 0.0496, + "step": 1513 + }, + { + "epoch": 1.0890127674878618, + "grad_norm": 3.429287413290873, + "learning_rate": 2.918050009521584e-06, + "loss": 0.0189, + "step": 1514 + }, + { + "epoch": 1.0897320625786728, + "grad_norm": 5.575831110927186, + "learning_rate": 2.9179390502178307e-06, + "loss": 0.2109, + "step": 1515 + }, + { + "epoch": 1.0904513576694839, + "grad_norm": 7.013230445251674, + "learning_rate": 2.917828017958624e-06, + "loss": 0.0752, + "step": 1516 + }, + { + "epoch": 1.091170652760295, + "grad_norm": 2.9801877393944856, + "learning_rate": 2.9177169127496766e-06, + "loss": 0.1064, + "step": 1517 + }, + { + "epoch": 1.091889947851106, + "grad_norm": 3.174139546542005, + "learning_rate": 2.917605734596705e-06, + "loss": 0.021, + "step": 1518 + }, + { + "epoch": 1.092609242941917, + "grad_norm": 4.479997482667637, + "learning_rate": 2.9174944835054305e-06, + "loss": 0.1909, + "step": 1519 + }, + { + "epoch": 1.093328538032728, + "grad_norm": 5.009655545194303, + "learning_rate": 2.9173831594815766e-06, + "loss": 0.0836, + "step": 1520 + }, + { + "epoch": 1.0940478331235388, + "grad_norm": 4.451020554627839, + "learning_rate": 2.917271762530871e-06, + "loss": 0.0985, + "step": 1521 + }, + { + "epoch": 1.0947671282143498, + "grad_norm": 6.125687542248357, + "learning_rate": 2.9171602926590447e-06, + "loss": 0.1761, + "step": 1522 + }, + { + "epoch": 1.0954864233051609, + "grad_norm": 2.894633391894206, + "learning_rate": 2.917048749871834e-06, + "loss": 0.0549, + "step": 1523 + }, + { + "epoch": 1.096205718395972, + "grad_norm": 5.311210867302218, + "learning_rate": 2.9169371341749777e-06, + "loss": 0.2164, + "step": 1524 + }, + { + "epoch": 1.096925013486783, + "grad_norm": 2.7861491703189896, + "learning_rate": 2.916825445574218e-06, + "loss": 0.1039, + "step": 1525 + }, + { + "epoch": 1.097644308577594, + "grad_norm": 4.814114699571843, + "learning_rate": 2.916713684075302e-06, + "loss": 0.2066, + "step": 1526 + }, + { + "epoch": 1.098363603668405, + "grad_norm": 5.544663014767644, + "learning_rate": 2.9166018496839804e-06, + "loss": 0.3047, + "step": 1527 + }, + { + "epoch": 1.099082898759216, + "grad_norm": 4.202365550843018, + "learning_rate": 2.9164899424060063e-06, + "loss": 0.1997, + "step": 1528 + }, + { + "epoch": 1.099802193850027, + "grad_norm": 3.2860409072834855, + "learning_rate": 2.916377962247139e-06, + "loss": 0.0212, + "step": 1529 + }, + { + "epoch": 1.100521488940838, + "grad_norm": 5.201999634671689, + "learning_rate": 2.9162659092131387e-06, + "loss": 0.0553, + "step": 1530 + }, + { + "epoch": 1.101240784031649, + "grad_norm": 4.6061959025854735, + "learning_rate": 2.916153783309771e-06, + "loss": 0.0182, + "step": 1531 + }, + { + "epoch": 1.10196007912246, + "grad_norm": 6.418190686183614, + "learning_rate": 2.916041584542805e-06, + "loss": 0.1723, + "step": 1532 + }, + { + "epoch": 1.102679374213271, + "grad_norm": 4.424650205317266, + "learning_rate": 2.9159293129180142e-06, + "loss": 0.2243, + "step": 1533 + }, + { + "epoch": 1.103398669304082, + "grad_norm": 3.819294914050668, + "learning_rate": 2.9158169684411744e-06, + "loss": 0.1249, + "step": 1534 + }, + { + "epoch": 1.104117964394893, + "grad_norm": 3.609413402564569, + "learning_rate": 2.915704551118066e-06, + "loss": 0.1149, + "step": 1535 + }, + { + "epoch": 1.104837259485704, + "grad_norm": 4.902049742689961, + "learning_rate": 2.9155920609544737e-06, + "loss": 0.1057, + "step": 1536 + }, + { + "epoch": 1.105556554576515, + "grad_norm": 4.975100067862964, + "learning_rate": 2.915479497956185e-06, + "loss": 0.1452, + "step": 1537 + }, + { + "epoch": 1.106275849667326, + "grad_norm": 2.0901599900076078, + "learning_rate": 2.9153668621289903e-06, + "loss": 0.0066, + "step": 1538 + }, + { + "epoch": 1.106995144758137, + "grad_norm": 5.892322293288568, + "learning_rate": 2.9152541534786866e-06, + "loss": 0.0551, + "step": 1539 + }, + { + "epoch": 1.107714439848948, + "grad_norm": 6.262369229039697, + "learning_rate": 2.9151413720110724e-06, + "loss": 0.1587, + "step": 1540 + }, + { + "epoch": 1.108433734939759, + "grad_norm": 5.7664315663265135, + "learning_rate": 2.915028517731951e-06, + "loss": 0.1796, + "step": 1541 + }, + { + "epoch": 1.10915303003057, + "grad_norm": 7.455060344039463, + "learning_rate": 2.9149155906471275e-06, + "loss": 0.316, + "step": 1542 + }, + { + "epoch": 1.1098723251213811, + "grad_norm": 7.1115107586720665, + "learning_rate": 2.914802590762413e-06, + "loss": 0.2119, + "step": 1543 + }, + { + "epoch": 1.1105916202121922, + "grad_norm": 5.886648515140201, + "learning_rate": 2.9146895180836217e-06, + "loss": 0.1851, + "step": 1544 + }, + { + "epoch": 1.111310915303003, + "grad_norm": 2.813077492182065, + "learning_rate": 2.9145763726165717e-06, + "loss": 0.0108, + "step": 1545 + }, + { + "epoch": 1.112030210393814, + "grad_norm": 2.622905487123166, + "learning_rate": 2.914463154367084e-06, + "loss": 0.0102, + "step": 1546 + }, + { + "epoch": 1.112749505484625, + "grad_norm": 4.475410008524123, + "learning_rate": 2.9143498633409836e-06, + "loss": 0.2229, + "step": 1547 + }, + { + "epoch": 1.113468800575436, + "grad_norm": 3.86739818321325, + "learning_rate": 2.9142364995441e-06, + "loss": 0.1164, + "step": 1548 + }, + { + "epoch": 1.114188095666247, + "grad_norm": 9.08395473574008, + "learning_rate": 2.914123062982266e-06, + "loss": 0.3142, + "step": 1549 + }, + { + "epoch": 1.1149073907570581, + "grad_norm": 5.945113354008717, + "learning_rate": 2.9140095536613182e-06, + "loss": 0.1653, + "step": 1550 + }, + { + "epoch": 1.1156266858478692, + "grad_norm": 4.910511869490129, + "learning_rate": 2.913895971587097e-06, + "loss": 0.0962, + "step": 1551 + }, + { + "epoch": 1.11634598093868, + "grad_norm": 3.6194086687598443, + "learning_rate": 2.913782316765445e-06, + "loss": 0.1011, + "step": 1552 + }, + { + "epoch": 1.117065276029491, + "grad_norm": 4.9102983992662566, + "learning_rate": 2.9136685892022118e-06, + "loss": 0.1588, + "step": 1553 + }, + { + "epoch": 1.117784571120302, + "grad_norm": 11.319662486227994, + "learning_rate": 2.913554788903248e-06, + "loss": 0.1685, + "step": 1554 + }, + { + "epoch": 1.118503866211113, + "grad_norm": 10.794114707844374, + "learning_rate": 2.913440915874408e-06, + "loss": 0.1582, + "step": 1555 + }, + { + "epoch": 1.1192231613019241, + "grad_norm": 5.640046685728013, + "learning_rate": 2.9133269701215525e-06, + "loss": 0.1814, + "step": 1556 + }, + { + "epoch": 1.1199424563927352, + "grad_norm": 1.8783785292977917, + "learning_rate": 2.9132129516505437e-06, + "loss": 0.0466, + "step": 1557 + }, + { + "epoch": 1.1206617514835462, + "grad_norm": 6.649623046313859, + "learning_rate": 2.913098860467247e-06, + "loss": 0.2591, + "step": 1558 + }, + { + "epoch": 1.1213810465743572, + "grad_norm": 2.4541511087383463, + "learning_rate": 2.9129846965775336e-06, + "loss": 0.0437, + "step": 1559 + }, + { + "epoch": 1.122100341665168, + "grad_norm": 4.196139146157071, + "learning_rate": 2.9128704599872772e-06, + "loss": 0.1225, + "step": 1560 + }, + { + "epoch": 1.122819636755979, + "grad_norm": 5.6710068660214725, + "learning_rate": 2.9127561507023553e-06, + "loss": 0.1585, + "step": 1561 + }, + { + "epoch": 1.12353893184679, + "grad_norm": 3.7293700063475366, + "learning_rate": 2.9126417687286497e-06, + "loss": 0.0693, + "step": 1562 + }, + { + "epoch": 1.1242582269376011, + "grad_norm": 7.142216721266976, + "learning_rate": 2.9125273140720446e-06, + "loss": 0.2059, + "step": 1563 + }, + { + "epoch": 1.1249775220284122, + "grad_norm": 1.502967925286762, + "learning_rate": 2.91241278673843e-06, + "loss": 0.0116, + "step": 1564 + }, + { + "epoch": 1.1256968171192232, + "grad_norm": 7.837880717222229, + "learning_rate": 2.9122981867336983e-06, + "loss": 0.1559, + "step": 1565 + }, + { + "epoch": 1.1264161122100342, + "grad_norm": 1.7459042863174874, + "learning_rate": 2.9121835140637455e-06, + "loss": 0.0386, + "step": 1566 + }, + { + "epoch": 1.127135407300845, + "grad_norm": 8.00085869248901, + "learning_rate": 2.912068768734472e-06, + "loss": 0.2524, + "step": 1567 + }, + { + "epoch": 1.127854702391656, + "grad_norm": 2.338797998044746, + "learning_rate": 2.911953950751781e-06, + "loss": 0.0073, + "step": 1568 + }, + { + "epoch": 1.1285739974824671, + "grad_norm": 3.3131262445769454, + "learning_rate": 2.9118390601215815e-06, + "loss": 0.1398, + "step": 1569 + }, + { + "epoch": 1.1292932925732782, + "grad_norm": 4.161046435348234, + "learning_rate": 2.9117240968497834e-06, + "loss": 0.1839, + "step": 1570 + }, + { + "epoch": 1.1300125876640892, + "grad_norm": 2.831063301192099, + "learning_rate": 2.911609060942302e-06, + "loss": 0.0244, + "step": 1571 + }, + { + "epoch": 1.1307318827549002, + "grad_norm": 8.83505975602995, + "learning_rate": 2.911493952405057e-06, + "loss": 0.1729, + "step": 1572 + }, + { + "epoch": 1.1314511778457113, + "grad_norm": 5.834108132069269, + "learning_rate": 2.9113787712439704e-06, + "loss": 0.0456, + "step": 1573 + }, + { + "epoch": 1.1321704729365223, + "grad_norm": 3.897330672700195, + "learning_rate": 2.9112635174649682e-06, + "loss": 0.1276, + "step": 1574 + }, + { + "epoch": 1.1328897680273333, + "grad_norm": 5.150524497442603, + "learning_rate": 2.9111481910739807e-06, + "loss": 0.1107, + "step": 1575 + }, + { + "epoch": 1.1336090631181441, + "grad_norm": 5.365956942451597, + "learning_rate": 2.9110327920769416e-06, + "loss": 0.2568, + "step": 1576 + }, + { + "epoch": 1.1343283582089552, + "grad_norm": 10.071509649899602, + "learning_rate": 2.9109173204797884e-06, + "loss": 0.2087, + "step": 1577 + }, + { + "epoch": 1.1350476532997662, + "grad_norm": 9.254387644492711, + "learning_rate": 2.9108017762884624e-06, + "loss": 0.2935, + "step": 1578 + }, + { + "epoch": 1.1357669483905772, + "grad_norm": 8.097262911525858, + "learning_rate": 2.910686159508908e-06, + "loss": 0.1981, + "step": 1579 + }, + { + "epoch": 1.1364862434813883, + "grad_norm": 2.787547442070557, + "learning_rate": 2.9105704701470745e-06, + "loss": 0.0113, + "step": 1580 + }, + { + "epoch": 1.1372055385721993, + "grad_norm": 2.557056387541306, + "learning_rate": 2.910454708208914e-06, + "loss": 0.0226, + "step": 1581 + }, + { + "epoch": 1.1379248336630103, + "grad_norm": 4.709660866454045, + "learning_rate": 2.9103388737003825e-06, + "loss": 0.1623, + "step": 1582 + }, + { + "epoch": 1.1386441287538211, + "grad_norm": 5.3638953598016474, + "learning_rate": 2.910222966627441e-06, + "loss": 0.2392, + "step": 1583 + }, + { + "epoch": 1.1393634238446322, + "grad_norm": 5.617800416822645, + "learning_rate": 2.910106986996052e-06, + "loss": 0.0265, + "step": 1584 + }, + { + "epoch": 1.1400827189354432, + "grad_norm": 7.406853221393165, + "learning_rate": 2.909990934812183e-06, + "loss": 0.0512, + "step": 1585 + }, + { + "epoch": 1.1408020140262543, + "grad_norm": 5.258026027066331, + "learning_rate": 2.9098748100818054e-06, + "loss": 0.1145, + "step": 1586 + }, + { + "epoch": 1.1415213091170653, + "grad_norm": 5.309413136417267, + "learning_rate": 2.9097586128108933e-06, + "loss": 0.158, + "step": 1587 + }, + { + "epoch": 1.1422406042078763, + "grad_norm": 4.8261048926627605, + "learning_rate": 2.9096423430054265e-06, + "loss": 0.2778, + "step": 1588 + }, + { + "epoch": 1.1429598992986874, + "grad_norm": 7.625622314728041, + "learning_rate": 2.9095260006713864e-06, + "loss": 0.2095, + "step": 1589 + }, + { + "epoch": 1.1436791943894984, + "grad_norm": 5.149380534623682, + "learning_rate": 2.9094095858147593e-06, + "loss": 0.0574, + "step": 1590 + }, + { + "epoch": 1.1443984894803092, + "grad_norm": 4.041501218959745, + "learning_rate": 2.9092930984415345e-06, + "loss": 0.0429, + "step": 1591 + }, + { + "epoch": 1.1451177845711202, + "grad_norm": 8.74005471457248, + "learning_rate": 2.9091765385577062e-06, + "loss": 0.2577, + "step": 1592 + }, + { + "epoch": 1.1458370796619313, + "grad_norm": 6.140724726940976, + "learning_rate": 2.9090599061692713e-06, + "loss": 0.0325, + "step": 1593 + }, + { + "epoch": 1.1465563747527423, + "grad_norm": 8.44877172666737, + "learning_rate": 2.9089432012822302e-06, + "loss": 0.2595, + "step": 1594 + }, + { + "epoch": 1.1472756698435533, + "grad_norm": 3.3033147905789986, + "learning_rate": 2.908826423902589e-06, + "loss": 0.0529, + "step": 1595 + }, + { + "epoch": 1.1479949649343644, + "grad_norm": 3.8281522604427707, + "learning_rate": 2.9087095740363546e-06, + "loss": 0.0403, + "step": 1596 + }, + { + "epoch": 1.1487142600251754, + "grad_norm": 3.3425254006305978, + "learning_rate": 2.9085926516895396e-06, + "loss": 0.07, + "step": 1597 + }, + { + "epoch": 1.1494335551159862, + "grad_norm": 7.303284480632762, + "learning_rate": 2.90847565686816e-06, + "loss": 0.1798, + "step": 1598 + }, + { + "epoch": 1.1501528502067973, + "grad_norm": 5.195745851298197, + "learning_rate": 2.908358589578235e-06, + "loss": 0.1652, + "step": 1599 + }, + { + "epoch": 1.1508721452976083, + "grad_norm": 4.128539758666629, + "learning_rate": 2.9082414498257892e-06, + "loss": 0.2128, + "step": 1600 + }, + { + "epoch": 1.1515914403884193, + "grad_norm": 1.2779722479038704, + "learning_rate": 2.9081242376168482e-06, + "loss": 0.0285, + "step": 1601 + }, + { + "epoch": 1.1523107354792304, + "grad_norm": 9.234725656767678, + "learning_rate": 2.9080069529574434e-06, + "loss": 0.2444, + "step": 1602 + }, + { + "epoch": 1.1530300305700414, + "grad_norm": 3.0457630925252355, + "learning_rate": 2.907889595853609e-06, + "loss": 0.0849, + "step": 1603 + }, + { + "epoch": 1.1537493256608524, + "grad_norm": 6.307294480679254, + "learning_rate": 2.9077721663113837e-06, + "loss": 0.2584, + "step": 1604 + }, + { + "epoch": 1.1544686207516635, + "grad_norm": 3.7467935991572134, + "learning_rate": 2.9076546643368087e-06, + "loss": 0.0898, + "step": 1605 + }, + { + "epoch": 1.1551879158424745, + "grad_norm": 7.174854215821884, + "learning_rate": 2.9075370899359306e-06, + "loss": 0.2132, + "step": 1606 + }, + { + "epoch": 1.1559072109332853, + "grad_norm": 5.258392544012779, + "learning_rate": 2.907419443114798e-06, + "loss": 0.1903, + "step": 1607 + }, + { + "epoch": 1.1566265060240963, + "grad_norm": 4.88702945614562, + "learning_rate": 2.9073017238794647e-06, + "loss": 0.1104, + "step": 1608 + }, + { + "epoch": 1.1573458011149074, + "grad_norm": 5.9137560242249005, + "learning_rate": 2.9071839322359872e-06, + "loss": 0.0195, + "step": 1609 + }, + { + "epoch": 1.1580650962057184, + "grad_norm": 5.631158851072675, + "learning_rate": 2.9070660681904264e-06, + "loss": 0.1812, + "step": 1610 + }, + { + "epoch": 1.1587843912965294, + "grad_norm": 4.279208120920692, + "learning_rate": 2.906948131748846e-06, + "loss": 0.1262, + "step": 1611 + }, + { + "epoch": 1.1595036863873405, + "grad_norm": 6.257222694408051, + "learning_rate": 2.9068301229173147e-06, + "loss": 0.0703, + "step": 1612 + }, + { + "epoch": 1.1602229814781515, + "grad_norm": 0.2032081337469878, + "learning_rate": 2.9067120417019036e-06, + "loss": 0.0008, + "step": 1613 + }, + { + "epoch": 1.1609422765689623, + "grad_norm": 3.1377080755271, + "learning_rate": 2.906593888108689e-06, + "loss": 0.023, + "step": 1614 + }, + { + "epoch": 1.1616615716597734, + "grad_norm": 5.02787057365801, + "learning_rate": 2.906475662143749e-06, + "loss": 0.1807, + "step": 1615 + }, + { + "epoch": 1.1623808667505844, + "grad_norm": 5.467161484946802, + "learning_rate": 2.906357363813168e-06, + "loss": 0.2285, + "step": 1616 + }, + { + "epoch": 1.1631001618413954, + "grad_norm": 7.3490352375187245, + "learning_rate": 2.9062389931230312e-06, + "loss": 0.1174, + "step": 1617 + }, + { + "epoch": 1.1638194569322065, + "grad_norm": 6.908168157345872, + "learning_rate": 2.90612055007943e-06, + "loss": 0.2132, + "step": 1618 + }, + { + "epoch": 1.1645387520230175, + "grad_norm": 7.707772252927093, + "learning_rate": 2.906002034688458e-06, + "loss": 0.0638, + "step": 1619 + }, + { + "epoch": 1.1652580471138285, + "grad_norm": 4.193829636563466, + "learning_rate": 2.9058834469562132e-06, + "loss": 0.0434, + "step": 1620 + }, + { + "epoch": 1.1659773422046396, + "grad_norm": 5.047942454099802, + "learning_rate": 2.905764786888797e-06, + "loss": 0.192, + "step": 1621 + }, + { + "epoch": 1.1666966372954504, + "grad_norm": 2.1721732650697443, + "learning_rate": 2.9056460544923148e-06, + "loss": 0.061, + "step": 1622 + }, + { + "epoch": 1.1674159323862614, + "grad_norm": 4.716460384335506, + "learning_rate": 2.9055272497728752e-06, + "loss": 0.0689, + "step": 1623 + }, + { + "epoch": 1.1681352274770724, + "grad_norm": 6.096225078287153, + "learning_rate": 2.9054083727365918e-06, + "loss": 0.1649, + "step": 1624 + }, + { + "epoch": 1.1688545225678835, + "grad_norm": 2.728843418451714, + "learning_rate": 2.9052894233895803e-06, + "loss": 0.0826, + "step": 1625 + }, + { + "epoch": 1.1695738176586945, + "grad_norm": 3.5483582819149637, + "learning_rate": 2.9051704017379616e-06, + "loss": 0.1121, + "step": 1626 + }, + { + "epoch": 1.1702931127495055, + "grad_norm": 2.82822980608982, + "learning_rate": 2.9050513077878584e-06, + "loss": 0.0654, + "step": 1627 + }, + { + "epoch": 1.1710124078403166, + "grad_norm": 6.554517102944529, + "learning_rate": 2.904932141545399e-06, + "loss": 0.3161, + "step": 1628 + }, + { + "epoch": 1.1717317029311274, + "grad_norm": 4.840126659486563, + "learning_rate": 2.904812903016715e-06, + "loss": 0.0688, + "step": 1629 + }, + { + "epoch": 1.1724509980219384, + "grad_norm": 6.832470797988187, + "learning_rate": 2.9046935922079406e-06, + "loss": 0.3091, + "step": 1630 + }, + { + "epoch": 1.1731702931127495, + "grad_norm": 3.863683949245649, + "learning_rate": 2.9045742091252152e-06, + "loss": 0.0981, + "step": 1631 + }, + { + "epoch": 1.1738895882035605, + "grad_norm": 3.9288208803046483, + "learning_rate": 2.904454753774681e-06, + "loss": 0.0989, + "step": 1632 + }, + { + "epoch": 1.1746088832943715, + "grad_norm": 3.7116885465434852, + "learning_rate": 2.9043352261624846e-06, + "loss": 0.0658, + "step": 1633 + }, + { + "epoch": 1.1753281783851826, + "grad_norm": 4.45888215501642, + "learning_rate": 2.9042156262947753e-06, + "loss": 0.1439, + "step": 1634 + }, + { + "epoch": 1.1760474734759936, + "grad_norm": 5.3579535942791265, + "learning_rate": 2.9040959541777073e-06, + "loss": 0.084, + "step": 1635 + }, + { + "epoch": 1.1767667685668046, + "grad_norm": 5.018282903143844, + "learning_rate": 2.9039762098174372e-06, + "loss": 0.1068, + "step": 1636 + }, + { + "epoch": 1.1774860636576157, + "grad_norm": 2.7803505311432417, + "learning_rate": 2.903856393220127e-06, + "loss": 0.0379, + "step": 1637 + }, + { + "epoch": 1.1782053587484265, + "grad_norm": 2.0191149948054914, + "learning_rate": 2.9037365043919405e-06, + "loss": 0.0148, + "step": 1638 + }, + { + "epoch": 1.1789246538392375, + "grad_norm": 7.588252475581475, + "learning_rate": 2.9036165433390465e-06, + "loss": 0.2215, + "step": 1639 + }, + { + "epoch": 1.1796439489300485, + "grad_norm": 7.867204675293377, + "learning_rate": 2.903496510067618e-06, + "loss": 0.1564, + "step": 1640 + }, + { + "epoch": 1.1803632440208596, + "grad_norm": 4.055991528870527, + "learning_rate": 2.90337640458383e-06, + "loss": 0.0736, + "step": 1641 + }, + { + "epoch": 1.1810825391116706, + "grad_norm": 8.103835864067179, + "learning_rate": 2.9032562268938625e-06, + "loss": 0.2698, + "step": 1642 + }, + { + "epoch": 1.1818018342024816, + "grad_norm": 4.384752734563741, + "learning_rate": 2.9031359770038986e-06, + "loss": 0.0974, + "step": 1643 + }, + { + "epoch": 1.1825211292932925, + "grad_norm": 5.261961242773679, + "learning_rate": 2.9030156549201258e-06, + "loss": 0.1479, + "step": 1644 + }, + { + "epoch": 1.1832404243841035, + "grad_norm": 5.305132852860056, + "learning_rate": 2.902895260648735e-06, + "loss": 0.1416, + "step": 1645 + }, + { + "epoch": 1.1839597194749145, + "grad_norm": 3.4968084235009202, + "learning_rate": 2.9027747941959194e-06, + "loss": 0.0847, + "step": 1646 + }, + { + "epoch": 1.1846790145657256, + "grad_norm": 4.627624467589781, + "learning_rate": 2.9026542555678784e-06, + "loss": 0.1882, + "step": 1647 + }, + { + "epoch": 1.1853983096565366, + "grad_norm": 4.225286606857522, + "learning_rate": 2.902533644770814e-06, + "loss": 0.0803, + "step": 1648 + }, + { + "epoch": 1.1861176047473476, + "grad_norm": 7.82031764626809, + "learning_rate": 2.9024129618109314e-06, + "loss": 0.2254, + "step": 1649 + }, + { + "epoch": 1.1868368998381587, + "grad_norm": 1.7268987748539237, + "learning_rate": 2.9022922066944404e-06, + "loss": 0.0064, + "step": 1650 + }, + { + "epoch": 1.1875561949289697, + "grad_norm": 6.649548549476015, + "learning_rate": 2.9021713794275534e-06, + "loss": 0.2012, + "step": 1651 + }, + { + "epoch": 1.1882754900197807, + "grad_norm": 4.181289615962555, + "learning_rate": 2.9020504800164876e-06, + "loss": 0.0563, + "step": 1652 + }, + { + "epoch": 1.1889947851105915, + "grad_norm": 4.3707111153336875, + "learning_rate": 2.901929508467463e-06, + "loss": 0.0199, + "step": 1653 + }, + { + "epoch": 1.1897140802014026, + "grad_norm": 3.2651002102996634, + "learning_rate": 2.9018084647867044e-06, + "loss": 0.0697, + "step": 1654 + }, + { + "epoch": 1.1904333752922136, + "grad_norm": 3.792175290629736, + "learning_rate": 2.9016873489804397e-06, + "loss": 0.1297, + "step": 1655 + }, + { + "epoch": 1.1911526703830246, + "grad_norm": 5.088332277924237, + "learning_rate": 2.9015661610549e-06, + "loss": 0.1366, + "step": 1656 + }, + { + "epoch": 1.1918719654738357, + "grad_norm": 5.444267509245161, + "learning_rate": 2.9014449010163215e-06, + "loss": 0.1073, + "step": 1657 + }, + { + "epoch": 1.1925912605646467, + "grad_norm": 6.90261058553411, + "learning_rate": 2.9013235688709417e-06, + "loss": 0.1746, + "step": 1658 + }, + { + "epoch": 1.1933105556554577, + "grad_norm": 7.2030895963938715, + "learning_rate": 2.901202164625005e-06, + "loss": 0.0453, + "step": 1659 + }, + { + "epoch": 1.1940298507462686, + "grad_norm": 0.295911356265283, + "learning_rate": 2.901080688284757e-06, + "loss": 0.0011, + "step": 1660 + }, + { + "epoch": 1.1947491458370796, + "grad_norm": 5.592379349397983, + "learning_rate": 2.9009591398564485e-06, + "loss": 0.2711, + "step": 1661 + }, + { + "epoch": 1.1954684409278906, + "grad_norm": 3.1551209109348144, + "learning_rate": 2.9008375193463324e-06, + "loss": 0.0928, + "step": 1662 + }, + { + "epoch": 1.1961877360187017, + "grad_norm": 3.6384103828381433, + "learning_rate": 2.900715826760667e-06, + "loss": 0.0153, + "step": 1663 + }, + { + "epoch": 1.1969070311095127, + "grad_norm": 6.72422480277707, + "learning_rate": 2.9005940621057136e-06, + "loss": 0.1839, + "step": 1664 + }, + { + "epoch": 1.1976263262003237, + "grad_norm": 4.500247125539293, + "learning_rate": 2.900472225387737e-06, + "loss": 0.0792, + "step": 1665 + }, + { + "epoch": 1.1983456212911348, + "grad_norm": 4.2434099752318994, + "learning_rate": 2.9003503166130058e-06, + "loss": 0.0755, + "step": 1666 + }, + { + "epoch": 1.1990649163819458, + "grad_norm": 4.460320268967298, + "learning_rate": 2.9002283357877925e-06, + "loss": 0.0816, + "step": 1667 + }, + { + "epoch": 1.1997842114727568, + "grad_norm": 5.604233029443637, + "learning_rate": 2.900106282918373e-06, + "loss": 0.3211, + "step": 1668 + }, + { + "epoch": 1.2005035065635676, + "grad_norm": 4.807050427072252, + "learning_rate": 2.8999841580110285e-06, + "loss": 0.1609, + "step": 1669 + }, + { + "epoch": 1.2012228016543787, + "grad_norm": 4.0434982571244875, + "learning_rate": 2.899861961072041e-06, + "loss": 0.0429, + "step": 1670 + }, + { + "epoch": 1.2019420967451897, + "grad_norm": 0.2619300813377373, + "learning_rate": 2.8997396921076976e-06, + "loss": 0.0007, + "step": 1671 + }, + { + "epoch": 1.2026613918360007, + "grad_norm": 6.572932135924672, + "learning_rate": 2.8996173511242903e-06, + "loss": 0.073, + "step": 1672 + }, + { + "epoch": 1.2033806869268118, + "grad_norm": 6.528714417366292, + "learning_rate": 2.8994949381281133e-06, + "loss": 0.1391, + "step": 1673 + }, + { + "epoch": 1.2040999820176228, + "grad_norm": 3.315380520396503, + "learning_rate": 2.8993724531254655e-06, + "loss": 0.0428, + "step": 1674 + }, + { + "epoch": 1.2048192771084336, + "grad_norm": 4.1644101093246855, + "learning_rate": 2.899249896122648e-06, + "loss": 0.1788, + "step": 1675 + }, + { + "epoch": 1.2055385721992447, + "grad_norm": 3.8480625075497272, + "learning_rate": 2.899127267125967e-06, + "loss": 0.077, + "step": 1676 + }, + { + "epoch": 1.2062578672900557, + "grad_norm": 4.588933061483007, + "learning_rate": 2.8990045661417323e-06, + "loss": 0.2078, + "step": 1677 + }, + { + "epoch": 1.2069771623808667, + "grad_norm": 3.8666833205546607, + "learning_rate": 2.8988817931762568e-06, + "loss": 0.0929, + "step": 1678 + }, + { + "epoch": 1.2076964574716778, + "grad_norm": 5.676310045283477, + "learning_rate": 2.898758948235857e-06, + "loss": 0.2461, + "step": 1679 + }, + { + "epoch": 1.2084157525624888, + "grad_norm": 5.637635799687778, + "learning_rate": 2.898636031326854e-06, + "loss": 0.2119, + "step": 1680 + }, + { + "epoch": 1.2091350476532998, + "grad_norm": 4.460716115876083, + "learning_rate": 2.898513042455572e-06, + "loss": 0.2893, + "step": 1681 + }, + { + "epoch": 1.2098543427441109, + "grad_norm": 3.600088991992017, + "learning_rate": 2.898389981628339e-06, + "loss": 0.0759, + "step": 1682 + }, + { + "epoch": 1.210573637834922, + "grad_norm": 2.6545082260260515, + "learning_rate": 2.898266848851486e-06, + "loss": 0.065, + "step": 1683 + }, + { + "epoch": 1.2112929329257327, + "grad_norm": 6.046021847605433, + "learning_rate": 2.8981436441313504e-06, + "loss": 0.1142, + "step": 1684 + }, + { + "epoch": 1.2120122280165437, + "grad_norm": 2.3560799867125826, + "learning_rate": 2.8980203674742693e-06, + "loss": 0.0451, + "step": 1685 + }, + { + "epoch": 1.2127315231073548, + "grad_norm": 5.070365155320128, + "learning_rate": 2.897897018886586e-06, + "loss": 0.1977, + "step": 1686 + }, + { + "epoch": 1.2134508181981658, + "grad_norm": 6.724096286192235, + "learning_rate": 2.8977735983746467e-06, + "loss": 0.2726, + "step": 1687 + }, + { + "epoch": 1.2141701132889768, + "grad_norm": 5.338627161316427, + "learning_rate": 2.8976501059448025e-06, + "loss": 0.0556, + "step": 1688 + }, + { + "epoch": 1.2148894083797879, + "grad_norm": 4.0449252993987885, + "learning_rate": 2.897526541603407e-06, + "loss": 0.0289, + "step": 1689 + }, + { + "epoch": 1.215608703470599, + "grad_norm": 4.4576782854327535, + "learning_rate": 2.897402905356818e-06, + "loss": 0.0652, + "step": 1690 + }, + { + "epoch": 1.2163279985614097, + "grad_norm": 4.602794051716955, + "learning_rate": 2.8972791972113957e-06, + "loss": 0.0341, + "step": 1691 + }, + { + "epoch": 1.2170472936522208, + "grad_norm": 3.724446979806658, + "learning_rate": 2.897155417173506e-06, + "loss": 0.099, + "step": 1692 + }, + { + "epoch": 1.2177665887430318, + "grad_norm": 4.986198057836993, + "learning_rate": 2.897031565249518e-06, + "loss": 0.1343, + "step": 1693 + }, + { + "epoch": 1.2184858838338428, + "grad_norm": 5.981542866327953, + "learning_rate": 2.896907641445803e-06, + "loss": 0.1653, + "step": 1694 + }, + { + "epoch": 1.2192051789246539, + "grad_norm": 5.443099139719888, + "learning_rate": 2.8967836457687377e-06, + "loss": 0.101, + "step": 1695 + }, + { + "epoch": 1.2199244740154649, + "grad_norm": 7.664009635039305, + "learning_rate": 2.896659578224702e-06, + "loss": 0.1482, + "step": 1696 + }, + { + "epoch": 1.220643769106276, + "grad_norm": 3.3137968050544084, + "learning_rate": 2.8965354388200792e-06, + "loss": 0.0754, + "step": 1697 + }, + { + "epoch": 1.221363064197087, + "grad_norm": 5.193651429800962, + "learning_rate": 2.896411227561257e-06, + "loss": 0.1697, + "step": 1698 + }, + { + "epoch": 1.2220823592878978, + "grad_norm": 5.331218352793793, + "learning_rate": 2.8962869444546252e-06, + "loss": 0.1576, + "step": 1699 + }, + { + "epoch": 1.2228016543787088, + "grad_norm": 5.747028241965266, + "learning_rate": 2.896162589506579e-06, + "loss": 0.0535, + "step": 1700 + }, + { + "epoch": 1.2235209494695198, + "grad_norm": 10.478785866039598, + "learning_rate": 2.896038162723517e-06, + "loss": 0.1801, + "step": 1701 + }, + { + "epoch": 1.2242402445603309, + "grad_norm": 4.430856437026723, + "learning_rate": 2.895913664111841e-06, + "loss": 0.1079, + "step": 1702 + }, + { + "epoch": 1.224959539651142, + "grad_norm": 4.38930395769076, + "learning_rate": 2.8957890936779563e-06, + "loss": 0.1367, + "step": 1703 + }, + { + "epoch": 1.225678834741953, + "grad_norm": 7.706914834039335, + "learning_rate": 2.895664451428272e-06, + "loss": 0.292, + "step": 1704 + }, + { + "epoch": 1.226398129832764, + "grad_norm": 7.317569259186702, + "learning_rate": 2.8955397373692023e-06, + "loss": 0.1357, + "step": 1705 + }, + { + "epoch": 1.2271174249235748, + "grad_norm": 5.101754142556382, + "learning_rate": 2.8954149515071635e-06, + "loss": 0.2589, + "step": 1706 + }, + { + "epoch": 1.2278367200143858, + "grad_norm": 2.6797096333382493, + "learning_rate": 2.895290093848575e-06, + "loss": 0.0541, + "step": 1707 + }, + { + "epoch": 1.2285560151051969, + "grad_norm": 7.141590482261668, + "learning_rate": 2.8951651643998623e-06, + "loss": 0.1327, + "step": 1708 + }, + { + "epoch": 1.2292753101960079, + "grad_norm": 3.5902350157397236, + "learning_rate": 2.895040163167453e-06, + "loss": 0.0376, + "step": 1709 + }, + { + "epoch": 1.229994605286819, + "grad_norm": 3.947371518643215, + "learning_rate": 2.8949150901577784e-06, + "loss": 0.0526, + "step": 1710 + }, + { + "epoch": 1.23071390037763, + "grad_norm": 4.036717449798842, + "learning_rate": 2.894789945377273e-06, + "loss": 0.0835, + "step": 1711 + }, + { + "epoch": 1.231433195468441, + "grad_norm": 5.2361268322036665, + "learning_rate": 2.894664728832377e-06, + "loss": 0.0801, + "step": 1712 + }, + { + "epoch": 1.232152490559252, + "grad_norm": 7.4235292974507665, + "learning_rate": 2.8945394405295327e-06, + "loss": 0.1383, + "step": 1713 + }, + { + "epoch": 1.232871785650063, + "grad_norm": 6.6864392813312525, + "learning_rate": 2.8944140804751855e-06, + "loss": 0.1541, + "step": 1714 + }, + { + "epoch": 1.2335910807408739, + "grad_norm": 5.6049479536392575, + "learning_rate": 2.8942886486757865e-06, + "loss": 0.0993, + "step": 1715 + }, + { + "epoch": 1.234310375831685, + "grad_norm": 6.597159116914805, + "learning_rate": 2.894163145137789e-06, + "loss": 0.2109, + "step": 1716 + }, + { + "epoch": 1.235029670922496, + "grad_norm": 8.59232018090716, + "learning_rate": 2.89403756986765e-06, + "loss": 0.1359, + "step": 1717 + }, + { + "epoch": 1.235748966013307, + "grad_norm": 6.9354133161352065, + "learning_rate": 2.893911922871831e-06, + "loss": 0.0575, + "step": 1718 + }, + { + "epoch": 1.236468261104118, + "grad_norm": 6.398491529279321, + "learning_rate": 2.8937862041567965e-06, + "loss": 0.3083, + "step": 1719 + }, + { + "epoch": 1.237187556194929, + "grad_norm": 7.600526697434066, + "learning_rate": 2.8936604137290152e-06, + "loss": 0.158, + "step": 1720 + }, + { + "epoch": 1.2379068512857399, + "grad_norm": 6.121493554226204, + "learning_rate": 2.893534551594959e-06, + "loss": 0.1952, + "step": 1721 + }, + { + "epoch": 1.2386261463765509, + "grad_norm": 4.091168035155126, + "learning_rate": 2.893408617761104e-06, + "loss": 0.0896, + "step": 1722 + }, + { + "epoch": 1.239345441467362, + "grad_norm": 5.762578515654643, + "learning_rate": 2.8932826122339297e-06, + "loss": 0.1906, + "step": 1723 + }, + { + "epoch": 1.240064736558173, + "grad_norm": 8.26043874681714, + "learning_rate": 2.893156535019919e-06, + "loss": 0.1216, + "step": 1724 + }, + { + "epoch": 1.240784031648984, + "grad_norm": 6.749779803615349, + "learning_rate": 2.893030386125559e-06, + "loss": 0.2953, + "step": 1725 + }, + { + "epoch": 1.241503326739795, + "grad_norm": 3.848139590984846, + "learning_rate": 2.89290416555734e-06, + "loss": 0.1744, + "step": 1726 + }, + { + "epoch": 1.242222621830606, + "grad_norm": 5.583602290170202, + "learning_rate": 2.892777873321757e-06, + "loss": 0.041, + "step": 1727 + }, + { + "epoch": 1.242941916921417, + "grad_norm": 4.676847080727266, + "learning_rate": 2.892651509425307e-06, + "loss": 0.0645, + "step": 1728 + }, + { + "epoch": 1.2436612120122281, + "grad_norm": 6.433773964091845, + "learning_rate": 2.892525073874492e-06, + "loss": 0.1588, + "step": 1729 + }, + { + "epoch": 1.244380507103039, + "grad_norm": 6.444075682049979, + "learning_rate": 2.892398566675818e-06, + "loss": 0.0945, + "step": 1730 + }, + { + "epoch": 1.24509980219385, + "grad_norm": 6.680322537514869, + "learning_rate": 2.892271987835793e-06, + "loss": 0.1826, + "step": 1731 + }, + { + "epoch": 1.245819097284661, + "grad_norm": 0.3296360838826887, + "learning_rate": 2.8921453373609307e-06, + "loss": 0.0019, + "step": 1732 + }, + { + "epoch": 1.246538392375472, + "grad_norm": 2.1011532810299163, + "learning_rate": 2.8920186152577465e-06, + "loss": 0.0274, + "step": 1733 + }, + { + "epoch": 1.247257687466283, + "grad_norm": 6.976378875433132, + "learning_rate": 2.8918918215327614e-06, + "loss": 0.3516, + "step": 1734 + }, + { + "epoch": 1.247976982557094, + "grad_norm": 6.291583287655501, + "learning_rate": 2.8917649561924983e-06, + "loss": 0.2794, + "step": 1735 + }, + { + "epoch": 1.2486962776479051, + "grad_norm": 5.361044298982202, + "learning_rate": 2.8916380192434846e-06, + "loss": 0.1136, + "step": 1736 + }, + { + "epoch": 1.249415572738716, + "grad_norm": 4.922418236009025, + "learning_rate": 2.8915110106922526e-06, + "loss": 0.1511, + "step": 1737 + }, + { + "epoch": 1.250134867829527, + "grad_norm": 5.538833389035142, + "learning_rate": 2.8913839305453367e-06, + "loss": 0.085, + "step": 1738 + }, + { + "epoch": 1.250854162920338, + "grad_norm": 2.4748071806339125, + "learning_rate": 2.891256778809274e-06, + "loss": 0.0171, + "step": 1739 + }, + { + "epoch": 1.251573458011149, + "grad_norm": 3.3505386432754416, + "learning_rate": 2.891129555490608e-06, + "loss": 0.0495, + "step": 1740 + }, + { + "epoch": 1.25229275310196, + "grad_norm": 6.994173889253759, + "learning_rate": 2.891002260595885e-06, + "loss": 0.2899, + "step": 1741 + }, + { + "epoch": 1.2530120481927711, + "grad_norm": 6.602909639058268, + "learning_rate": 2.8908748941316534e-06, + "loss": 0.1698, + "step": 1742 + }, + { + "epoch": 1.2537313432835822, + "grad_norm": 6.379299761467531, + "learning_rate": 2.890747456104467e-06, + "loss": 0.1662, + "step": 1743 + }, + { + "epoch": 1.2544506383743932, + "grad_norm": 4.356813554599697, + "learning_rate": 2.8906199465208825e-06, + "loss": 0.0419, + "step": 1744 + }, + { + "epoch": 1.2551699334652042, + "grad_norm": 8.794103807862856, + "learning_rate": 2.8904923653874607e-06, + "loss": 0.3719, + "step": 1745 + }, + { + "epoch": 1.255889228556015, + "grad_norm": 8.530073217884308, + "learning_rate": 2.890364712710766e-06, + "loss": 0.1047, + "step": 1746 + }, + { + "epoch": 1.256608523646826, + "grad_norm": 2.0734435101327624, + "learning_rate": 2.8902369884973657e-06, + "loss": 0.057, + "step": 1747 + }, + { + "epoch": 1.257327818737637, + "grad_norm": 4.430281238214375, + "learning_rate": 2.890109192753832e-06, + "loss": 0.0498, + "step": 1748 + }, + { + "epoch": 1.2580471138284481, + "grad_norm": 3.641779769524025, + "learning_rate": 2.88998132548674e-06, + "loss": 0.1163, + "step": 1749 + }, + { + "epoch": 1.2587664089192592, + "grad_norm": 1.6740554612817447, + "learning_rate": 2.889853386702669e-06, + "loss": 0.0385, + "step": 1750 + }, + { + "epoch": 1.2594857040100702, + "grad_norm": 2.500383046995515, + "learning_rate": 2.889725376408201e-06, + "loss": 0.0159, + "step": 1751 + }, + { + "epoch": 1.260204999100881, + "grad_norm": 4.620359802476714, + "learning_rate": 2.8895972946099234e-06, + "loss": 0.1521, + "step": 1752 + }, + { + "epoch": 1.260924294191692, + "grad_norm": 3.90366105756383, + "learning_rate": 2.889469141314425e-06, + "loss": 0.025, + "step": 1753 + }, + { + "epoch": 1.261643589282503, + "grad_norm": 3.0382449177771913, + "learning_rate": 2.8893409165283006e-06, + "loss": 0.0666, + "step": 1754 + }, + { + "epoch": 1.2623628843733141, + "grad_norm": 3.7165377644985442, + "learning_rate": 2.889212620258147e-06, + "loss": 0.107, + "step": 1755 + }, + { + "epoch": 1.2630821794641252, + "grad_norm": 5.03734432241681, + "learning_rate": 2.8890842525105657e-06, + "loss": 0.2066, + "step": 1756 + }, + { + "epoch": 1.2638014745549362, + "grad_norm": 4.060745336182474, + "learning_rate": 2.888955813292161e-06, + "loss": 0.0557, + "step": 1757 + }, + { + "epoch": 1.2645207696457472, + "grad_norm": 9.626269648195246, + "learning_rate": 2.888827302609541e-06, + "loss": 0.1153, + "step": 1758 + }, + { + "epoch": 1.2652400647365583, + "grad_norm": 5.0309212296301355, + "learning_rate": 2.8886987204693185e-06, + "loss": 0.0466, + "step": 1759 + }, + { + "epoch": 1.2659593598273693, + "grad_norm": 3.8823097303680516, + "learning_rate": 2.888570066878109e-06, + "loss": 0.0715, + "step": 1760 + }, + { + "epoch": 1.26667865491818, + "grad_norm": 4.899662830415322, + "learning_rate": 2.8884413418425323e-06, + "loss": 0.1671, + "step": 1761 + }, + { + "epoch": 1.2673979500089911, + "grad_norm": 1.4556095411675072, + "learning_rate": 2.888312545369211e-06, + "loss": 0.0096, + "step": 1762 + }, + { + "epoch": 1.2681172450998022, + "grad_norm": 4.863421250658491, + "learning_rate": 2.888183677464772e-06, + "loss": 0.1088, + "step": 1763 + }, + { + "epoch": 1.2688365401906132, + "grad_norm": 5.069363586327778, + "learning_rate": 2.8880547381358466e-06, + "loss": 0.1343, + "step": 1764 + }, + { + "epoch": 1.2695558352814242, + "grad_norm": 6.596713193494357, + "learning_rate": 2.8879257273890674e-06, + "loss": 0.2512, + "step": 1765 + }, + { + "epoch": 1.2702751303722353, + "grad_norm": 3.6578834557298676, + "learning_rate": 2.8877966452310737e-06, + "loss": 0.0197, + "step": 1766 + }, + { + "epoch": 1.270994425463046, + "grad_norm": 7.139498237139241, + "learning_rate": 2.8876674916685066e-06, + "loss": 0.1187, + "step": 1767 + }, + { + "epoch": 1.2717137205538571, + "grad_norm": 5.442985747142592, + "learning_rate": 2.8875382667080103e-06, + "loss": 0.2379, + "step": 1768 + }, + { + "epoch": 1.2724330156446682, + "grad_norm": 2.815477592454708, + "learning_rate": 2.887408970356235e-06, + "loss": 0.0099, + "step": 1769 + }, + { + "epoch": 1.2731523107354792, + "grad_norm": 3.5539231872341364, + "learning_rate": 2.8872796026198322e-06, + "loss": 0.1148, + "step": 1770 + }, + { + "epoch": 1.2738716058262902, + "grad_norm": 5.956974322358411, + "learning_rate": 2.887150163505459e-06, + "loss": 0.1346, + "step": 1771 + }, + { + "epoch": 1.2745909009171013, + "grad_norm": 5.778114535080318, + "learning_rate": 2.8870206530197747e-06, + "loss": 0.0432, + "step": 1772 + }, + { + "epoch": 1.2753101960079123, + "grad_norm": 7.749029376734017, + "learning_rate": 2.8868910711694428e-06, + "loss": 0.499, + "step": 1773 + }, + { + "epoch": 1.2760294910987233, + "grad_norm": 8.817791650529825, + "learning_rate": 2.886761417961131e-06, + "loss": 0.3223, + "step": 1774 + }, + { + "epoch": 1.2767487861895344, + "grad_norm": 5.582186615519434, + "learning_rate": 2.8866316934015094e-06, + "loss": 0.0394, + "step": 1775 + }, + { + "epoch": 1.2774680812803454, + "grad_norm": 3.724161404478559, + "learning_rate": 2.886501897497253e-06, + "loss": 0.1146, + "step": 1776 + }, + { + "epoch": 1.2781873763711562, + "grad_norm": 4.598854300745457, + "learning_rate": 2.8863720302550403e-06, + "loss": 0.0926, + "step": 1777 + }, + { + "epoch": 1.2789066714619672, + "grad_norm": 1.443410755650339, + "learning_rate": 2.8862420916815535e-06, + "loss": 0.0036, + "step": 1778 + }, + { + "epoch": 1.2796259665527783, + "grad_norm": 6.324532017095005, + "learning_rate": 2.886112081783477e-06, + "loss": 0.1501, + "step": 1779 + }, + { + "epoch": 1.2803452616435893, + "grad_norm": 2.037291547362322, + "learning_rate": 2.885982000567501e-06, + "loss": 0.0426, + "step": 1780 + }, + { + "epoch": 1.2810645567344003, + "grad_norm": 4.535852923569214, + "learning_rate": 2.8858518480403175e-06, + "loss": 0.122, + "step": 1781 + }, + { + "epoch": 1.2817838518252114, + "grad_norm": 2.390959142873699, + "learning_rate": 2.8857216242086236e-06, + "loss": 0.0165, + "step": 1782 + }, + { + "epoch": 1.2825031469160222, + "grad_norm": 6.136026621783262, + "learning_rate": 2.8855913290791202e-06, + "loss": 0.2305, + "step": 1783 + }, + { + "epoch": 1.2832224420068332, + "grad_norm": 6.910578009349899, + "learning_rate": 2.88546096265851e-06, + "loss": 0.1542, + "step": 1784 + }, + { + "epoch": 1.2839417370976443, + "grad_norm": 6.4265249107022955, + "learning_rate": 2.8853305249535015e-06, + "loss": 0.1086, + "step": 1785 + }, + { + "epoch": 1.2846610321884553, + "grad_norm": 8.33119472210341, + "learning_rate": 2.8852000159708057e-06, + "loss": 0.1584, + "step": 1786 + }, + { + "epoch": 1.2853803272792663, + "grad_norm": 4.704317466052146, + "learning_rate": 2.8850694357171374e-06, + "loss": 0.142, + "step": 1787 + }, + { + "epoch": 1.2860996223700774, + "grad_norm": 6.978055033832254, + "learning_rate": 2.884938784199215e-06, + "loss": 0.1079, + "step": 1788 + }, + { + "epoch": 1.2868189174608884, + "grad_norm": 4.60709267559893, + "learning_rate": 2.8848080614237615e-06, + "loss": 0.0396, + "step": 1789 + }, + { + "epoch": 1.2875382125516994, + "grad_norm": 1.2713493964235183, + "learning_rate": 2.884677267397502e-06, + "loss": 0.0031, + "step": 1790 + }, + { + "epoch": 1.2882575076425105, + "grad_norm": 3.0437346723017846, + "learning_rate": 2.884546402127166e-06, + "loss": 0.0179, + "step": 1791 + }, + { + "epoch": 1.2889768027333213, + "grad_norm": 2.369345739160241, + "learning_rate": 2.8844154656194872e-06, + "loss": 0.0723, + "step": 1792 + }, + { + "epoch": 1.2896960978241323, + "grad_norm": 6.1815280427151045, + "learning_rate": 2.8842844578812026e-06, + "loss": 0.2611, + "step": 1793 + }, + { + "epoch": 1.2904153929149433, + "grad_norm": 3.369798320848323, + "learning_rate": 2.8841533789190523e-06, + "loss": 0.0337, + "step": 1794 + }, + { + "epoch": 1.2911346880057544, + "grad_norm": 4.75350687169893, + "learning_rate": 2.884022228739781e-06, + "loss": 0.1725, + "step": 1795 + }, + { + "epoch": 1.2918539830965654, + "grad_norm": 3.9552176641061934, + "learning_rate": 2.883891007350137e-06, + "loss": 0.0274, + "step": 1796 + }, + { + "epoch": 1.2925732781873764, + "grad_norm": 3.6310465645044543, + "learning_rate": 2.883759714756871e-06, + "loss": 0.1276, + "step": 1797 + }, + { + "epoch": 1.2932925732781873, + "grad_norm": 4.689954191449377, + "learning_rate": 2.8836283509667385e-06, + "loss": 0.0084, + "step": 1798 + }, + { + "epoch": 1.2940118683689983, + "grad_norm": 5.860750940850004, + "learning_rate": 2.8834969159864985e-06, + "loss": 0.2997, + "step": 1799 + }, + { + "epoch": 1.2947311634598093, + "grad_norm": 8.771977581882343, + "learning_rate": 2.8833654098229132e-06, + "loss": 0.3005, + "step": 1800 + }, + { + "epoch": 1.2954504585506204, + "grad_norm": 7.472293046286825, + "learning_rate": 2.8832338324827496e-06, + "loss": 0.1178, + "step": 1801 + }, + { + "epoch": 1.2961697536414314, + "grad_norm": 4.310086642271168, + "learning_rate": 2.883102183972777e-06, + "loss": 0.2254, + "step": 1802 + }, + { + "epoch": 1.2968890487322424, + "grad_norm": 4.707417766061669, + "learning_rate": 2.8829704642997685e-06, + "loss": 0.1275, + "step": 1803 + }, + { + "epoch": 1.2976083438230535, + "grad_norm": 10.22764751000119, + "learning_rate": 2.882838673470503e-06, + "loss": 0.2069, + "step": 1804 + }, + { + "epoch": 1.2983276389138645, + "grad_norm": 5.686348961327223, + "learning_rate": 2.8827068114917596e-06, + "loss": 0.1678, + "step": 1805 + }, + { + "epoch": 1.2990469340046755, + "grad_norm": 2.0740265001200275, + "learning_rate": 2.882574878370323e-06, + "loss": 0.0726, + "step": 1806 + }, + { + "epoch": 1.2997662290954866, + "grad_norm": 4.024038524883801, + "learning_rate": 2.8824428741129828e-06, + "loss": 0.1014, + "step": 1807 + }, + { + "epoch": 1.3004855241862974, + "grad_norm": 2.939439582226724, + "learning_rate": 2.8823107987265297e-06, + "loss": 0.0511, + "step": 1808 + }, + { + "epoch": 1.3012048192771084, + "grad_norm": 3.567666260928984, + "learning_rate": 2.8821786522177592e-06, + "loss": 0.0651, + "step": 1809 + }, + { + "epoch": 1.3019241143679194, + "grad_norm": 0.2020885875817831, + "learning_rate": 2.882046434593471e-06, + "loss": 0.0008, + "step": 1810 + }, + { + "epoch": 1.3026434094587305, + "grad_norm": 5.3511253225263475, + "learning_rate": 2.881914145860467e-06, + "loss": 0.1655, + "step": 1811 + }, + { + "epoch": 1.3033627045495415, + "grad_norm": 7.059966386638408, + "learning_rate": 2.8817817860255553e-06, + "loss": 0.2073, + "step": 1812 + }, + { + "epoch": 1.3040819996403523, + "grad_norm": 4.115207631963609, + "learning_rate": 2.881649355095544e-06, + "loss": 0.0305, + "step": 1813 + }, + { + "epoch": 1.3048012947311634, + "grad_norm": 5.305701484618844, + "learning_rate": 2.8815168530772485e-06, + "loss": 0.109, + "step": 1814 + }, + { + "epoch": 1.3055205898219744, + "grad_norm": 7.543386207758745, + "learning_rate": 2.8813842799774857e-06, + "loss": 0.3113, + "step": 1815 + }, + { + "epoch": 1.3062398849127854, + "grad_norm": 5.235520833490784, + "learning_rate": 2.881251635803077e-06, + "loss": 0.1324, + "step": 1816 + }, + { + "epoch": 1.3069591800035965, + "grad_norm": 3.283920929196038, + "learning_rate": 2.8811189205608466e-06, + "loss": 0.0564, + "step": 1817 + }, + { + "epoch": 1.3076784750944075, + "grad_norm": 3.5219511617821206, + "learning_rate": 2.8809861342576233e-06, + "loss": 0.1593, + "step": 1818 + }, + { + "epoch": 1.3083977701852185, + "grad_norm": 8.377887136886393, + "learning_rate": 2.8808532769002395e-06, + "loss": 0.3513, + "step": 1819 + }, + { + "epoch": 1.3091170652760296, + "grad_norm": 2.095217812133882, + "learning_rate": 2.88072034849553e-06, + "loss": 0.0101, + "step": 1820 + }, + { + "epoch": 1.3098363603668406, + "grad_norm": 0.12494302192704931, + "learning_rate": 2.880587349050335e-06, + "loss": 0.0004, + "step": 1821 + }, + { + "epoch": 1.3105556554576516, + "grad_norm": 5.062425812515262, + "learning_rate": 2.8804542785714975e-06, + "loss": 0.1723, + "step": 1822 + }, + { + "epoch": 1.3112749505484624, + "grad_norm": 2.8671584954488165, + "learning_rate": 2.880321137065864e-06, + "loss": 0.0644, + "step": 1823 + }, + { + "epoch": 1.3119942456392735, + "grad_norm": 7.190457460575838, + "learning_rate": 2.8801879245402853e-06, + "loss": 0.1453, + "step": 1824 + }, + { + "epoch": 1.3127135407300845, + "grad_norm": 4.555336227310286, + "learning_rate": 2.8800546410016143e-06, + "loss": 0.1951, + "step": 1825 + }, + { + "epoch": 1.3134328358208955, + "grad_norm": 2.8697308459652904, + "learning_rate": 2.87992128645671e-06, + "loss": 0.0344, + "step": 1826 + }, + { + "epoch": 1.3141521309117066, + "grad_norm": 4.465652176100414, + "learning_rate": 2.8797878609124323e-06, + "loss": 0.1618, + "step": 1827 + }, + { + "epoch": 1.3148714260025176, + "grad_norm": 1.8426529178084399, + "learning_rate": 2.879654364375648e-06, + "loss": 0.0037, + "step": 1828 + }, + { + "epoch": 1.3155907210933284, + "grad_norm": 5.041210844442453, + "learning_rate": 2.879520796853224e-06, + "loss": 0.2127, + "step": 1829 + }, + { + "epoch": 1.3163100161841395, + "grad_norm": 4.17112896426112, + "learning_rate": 2.879387158352034e-06, + "loss": 0.1656, + "step": 1830 + }, + { + "epoch": 1.3170293112749505, + "grad_norm": 5.417089385255273, + "learning_rate": 2.8792534488789528e-06, + "loss": 0.1785, + "step": 1831 + }, + { + "epoch": 1.3177486063657615, + "grad_norm": 6.594442214193899, + "learning_rate": 2.8791196684408606e-06, + "loss": 0.3024, + "step": 1832 + }, + { + "epoch": 1.3184679014565726, + "grad_norm": 2.4721537178047086, + "learning_rate": 2.8789858170446407e-06, + "loss": 0.0193, + "step": 1833 + }, + { + "epoch": 1.3191871965473836, + "grad_norm": 5.675765175033997, + "learning_rate": 2.8788518946971795e-06, + "loss": 0.1626, + "step": 1834 + }, + { + "epoch": 1.3199064916381946, + "grad_norm": 3.992369985089034, + "learning_rate": 2.878717901405368e-06, + "loss": 0.0839, + "step": 1835 + }, + { + "epoch": 1.3206257867290057, + "grad_norm": 6.138903582265613, + "learning_rate": 2.8785838371761e-06, + "loss": 0.2443, + "step": 1836 + }, + { + "epoch": 1.3213450818198167, + "grad_norm": 3.5018656836216384, + "learning_rate": 2.8784497020162735e-06, + "loss": 0.0231, + "step": 1837 + }, + { + "epoch": 1.3220643769106275, + "grad_norm": 5.687174189057592, + "learning_rate": 2.8783154959327904e-06, + "loss": 0.2432, + "step": 1838 + }, + { + "epoch": 1.3227836720014385, + "grad_norm": 5.2718872033794435, + "learning_rate": 2.878181218932555e-06, + "loss": 0.1147, + "step": 1839 + }, + { + "epoch": 1.3235029670922496, + "grad_norm": 6.607650048238397, + "learning_rate": 2.878046871022477e-06, + "loss": 0.2408, + "step": 1840 + }, + { + "epoch": 1.3242222621830606, + "grad_norm": 5.059278886193367, + "learning_rate": 2.877912452209468e-06, + "loss": 0.158, + "step": 1841 + }, + { + "epoch": 1.3249415572738716, + "grad_norm": 5.742009902032591, + "learning_rate": 2.877777962500445e-06, + "loss": 0.2005, + "step": 1842 + }, + { + "epoch": 1.3256608523646827, + "grad_norm": 2.883964825618812, + "learning_rate": 2.877643401902327e-06, + "loss": 0.0677, + "step": 1843 + }, + { + "epoch": 1.3263801474554935, + "grad_norm": 0.9963092949581706, + "learning_rate": 2.8775087704220374e-06, + "loss": 0.007, + "step": 1844 + }, + { + "epoch": 1.3270994425463045, + "grad_norm": 2.112298784312431, + "learning_rate": 2.8773740680665037e-06, + "loss": 0.054, + "step": 1845 + }, + { + "epoch": 1.3278187376371156, + "grad_norm": 2.374280353637511, + "learning_rate": 2.877239294842656e-06, + "loss": 0.0071, + "step": 1846 + }, + { + "epoch": 1.3285380327279266, + "grad_norm": 7.016661202511424, + "learning_rate": 2.877104450757429e-06, + "loss": 0.1912, + "step": 1847 + }, + { + "epoch": 1.3292573278187376, + "grad_norm": 3.237087173742285, + "learning_rate": 2.8769695358177606e-06, + "loss": 0.0782, + "step": 1848 + }, + { + "epoch": 1.3299766229095487, + "grad_norm": 6.190447468857955, + "learning_rate": 2.8768345500305926e-06, + "loss": 0.1052, + "step": 1849 + }, + { + "epoch": 1.3306959180003597, + "grad_norm": 4.884468546707741, + "learning_rate": 2.8766994934028697e-06, + "loss": 0.0965, + "step": 1850 + }, + { + "epoch": 1.3314152130911707, + "grad_norm": 5.051099855875771, + "learning_rate": 2.8765643659415414e-06, + "loss": 0.1169, + "step": 1851 + }, + { + "epoch": 1.3321345081819818, + "grad_norm": 6.455809895305699, + "learning_rate": 2.8764291676535597e-06, + "loss": 0.0151, + "step": 1852 + }, + { + "epoch": 1.3328538032727928, + "grad_norm": 5.785708648703478, + "learning_rate": 2.8762938985458814e-06, + "loss": 0.1337, + "step": 1853 + }, + { + "epoch": 1.3335730983636036, + "grad_norm": 3.931423117882026, + "learning_rate": 2.8761585586254655e-06, + "loss": 0.1127, + "step": 1854 + }, + { + "epoch": 1.3342923934544146, + "grad_norm": 4.761794607288105, + "learning_rate": 2.876023147899277e-06, + "loss": 0.1858, + "step": 1855 + }, + { + "epoch": 1.3350116885452257, + "grad_norm": 3.006424913906359, + "learning_rate": 2.875887666374281e-06, + "loss": 0.0475, + "step": 1856 + }, + { + "epoch": 1.3357309836360367, + "grad_norm": 7.130340436290732, + "learning_rate": 2.87575211405745e-06, + "loss": 0.0653, + "step": 1857 + }, + { + "epoch": 1.3364502787268477, + "grad_norm": 5.813711605396539, + "learning_rate": 2.875616490955757e-06, + "loss": 0.1385, + "step": 1858 + }, + { + "epoch": 1.3371695738176588, + "grad_norm": 8.33374339637154, + "learning_rate": 2.8754807970761812e-06, + "loss": 0.1908, + "step": 1859 + }, + { + "epoch": 1.3378888689084696, + "grad_norm": 6.39867126613376, + "learning_rate": 2.8753450324257036e-06, + "loss": 0.1393, + "step": 1860 + }, + { + "epoch": 1.3386081639992806, + "grad_norm": 3.4348843090382966, + "learning_rate": 2.87520919701131e-06, + "loss": 0.1019, + "step": 1861 + }, + { + "epoch": 1.3393274590900917, + "grad_norm": 4.684351452048262, + "learning_rate": 2.8750732908399887e-06, + "loss": 0.0786, + "step": 1862 + }, + { + "epoch": 1.3400467541809027, + "grad_norm": 6.398053898938119, + "learning_rate": 2.874937313918733e-06, + "loss": 0.0547, + "step": 1863 + }, + { + "epoch": 1.3407660492717137, + "grad_norm": 3.3814315374114825, + "learning_rate": 2.874801266254539e-06, + "loss": 0.0197, + "step": 1864 + }, + { + "epoch": 1.3414853443625248, + "grad_norm": 6.8620285275705, + "learning_rate": 2.8746651478544064e-06, + "loss": 0.2157, + "step": 1865 + }, + { + "epoch": 1.3422046394533358, + "grad_norm": 7.384899978037323, + "learning_rate": 2.8745289587253384e-06, + "loss": 0.168, + "step": 1866 + }, + { + "epoch": 1.3429239345441468, + "grad_norm": 6.078623593013814, + "learning_rate": 2.874392698874343e-06, + "loss": 0.1552, + "step": 1867 + }, + { + "epoch": 1.3436432296349579, + "grad_norm": 4.592180192317263, + "learning_rate": 2.8742563683084304e-06, + "loss": 0.1839, + "step": 1868 + }, + { + "epoch": 1.3443625247257687, + "grad_norm": 7.5951922177785045, + "learning_rate": 2.874119967034615e-06, + "loss": 0.3214, + "step": 1869 + }, + { + "epoch": 1.3450818198165797, + "grad_norm": 10.287606520018116, + "learning_rate": 2.873983495059915e-06, + "loss": 0.0433, + "step": 1870 + }, + { + "epoch": 1.3458011149073907, + "grad_norm": 3.396971632359829, + "learning_rate": 2.873846952391353e-06, + "loss": 0.0855, + "step": 1871 + }, + { + "epoch": 1.3465204099982018, + "grad_norm": 2.5977139927443296, + "learning_rate": 2.873710339035953e-06, + "loss": 0.0748, + "step": 1872 + }, + { + "epoch": 1.3472397050890128, + "grad_norm": 9.532789292087031, + "learning_rate": 2.8735736550007447e-06, + "loss": 0.1301, + "step": 1873 + }, + { + "epoch": 1.3479590001798238, + "grad_norm": 0.5868089599182268, + "learning_rate": 2.8734369002927607e-06, + "loss": 0.0027, + "step": 1874 + }, + { + "epoch": 1.3486782952706347, + "grad_norm": 5.32084968799656, + "learning_rate": 2.873300074919037e-06, + "loss": 0.0328, + "step": 1875 + }, + { + "epoch": 1.3493975903614457, + "grad_norm": 3.453961300947311, + "learning_rate": 2.873163178886614e-06, + "loss": 0.0584, + "step": 1876 + }, + { + "epoch": 1.3501168854522567, + "grad_norm": 4.282843496149778, + "learning_rate": 2.873026212202535e-06, + "loss": 0.0776, + "step": 1877 + }, + { + "epoch": 1.3508361805430678, + "grad_norm": 3.671254299140581, + "learning_rate": 2.8728891748738466e-06, + "loss": 0.0674, + "step": 1878 + }, + { + "epoch": 1.3515554756338788, + "grad_norm": 5.896455441705231, + "learning_rate": 2.8727520669076007e-06, + "loss": 0.2183, + "step": 1879 + }, + { + "epoch": 1.3522747707246898, + "grad_norm": 4.721397643763593, + "learning_rate": 2.8726148883108505e-06, + "loss": 0.1232, + "step": 1880 + }, + { + "epoch": 1.3529940658155009, + "grad_norm": 4.928749330687315, + "learning_rate": 2.8724776390906553e-06, + "loss": 0.2478, + "step": 1881 + }, + { + "epoch": 1.353713360906312, + "grad_norm": 6.17180554171681, + "learning_rate": 2.8723403192540763e-06, + "loss": 0.2045, + "step": 1882 + }, + { + "epoch": 1.354432655997123, + "grad_norm": 6.8157629871334935, + "learning_rate": 2.872202928808178e-06, + "loss": 0.0328, + "step": 1883 + }, + { + "epoch": 1.355151951087934, + "grad_norm": 4.703999473831349, + "learning_rate": 2.8720654677600314e-06, + "loss": 0.0672, + "step": 1884 + }, + { + "epoch": 1.3558712461787448, + "grad_norm": 5.759290786253367, + "learning_rate": 2.871927936116707e-06, + "loss": 0.1364, + "step": 1885 + }, + { + "epoch": 1.3565905412695558, + "grad_norm": 4.724110816705259, + "learning_rate": 2.871790333885282e-06, + "loss": 0.0973, + "step": 1886 + }, + { + "epoch": 1.3573098363603668, + "grad_norm": 5.829353646123308, + "learning_rate": 2.8716526610728364e-06, + "loss": 0.21, + "step": 1887 + }, + { + "epoch": 1.3580291314511779, + "grad_norm": 4.067270789819697, + "learning_rate": 2.871514917686454e-06, + "loss": 0.2588, + "step": 1888 + }, + { + "epoch": 1.358748426541989, + "grad_norm": 2.4918752137717415, + "learning_rate": 2.8713771037332207e-06, + "loss": 0.0774, + "step": 1889 + }, + { + "epoch": 1.3594677216327997, + "grad_norm": 3.8950955116659096, + "learning_rate": 2.8712392192202284e-06, + "loss": 0.2138, + "step": 1890 + }, + { + "epoch": 1.3601870167236108, + "grad_norm": 4.90310513012652, + "learning_rate": 2.8711012641545715e-06, + "loss": 0.1013, + "step": 1891 + }, + { + "epoch": 1.3609063118144218, + "grad_norm": 5.761451843950282, + "learning_rate": 2.870963238543347e-06, + "loss": 0.1667, + "step": 1892 + }, + { + "epoch": 1.3616256069052328, + "grad_norm": 6.030882256630764, + "learning_rate": 2.8708251423936573e-06, + "loss": 0.2053, + "step": 1893 + }, + { + "epoch": 1.3623449019960439, + "grad_norm": 6.522936530240776, + "learning_rate": 2.8706869757126077e-06, + "loss": 0.0359, + "step": 1894 + }, + { + "epoch": 1.363064197086855, + "grad_norm": 6.601411931507219, + "learning_rate": 2.870548738507307e-06, + "loss": 0.2296, + "step": 1895 + }, + { + "epoch": 1.363783492177666, + "grad_norm": 5.691795061079199, + "learning_rate": 2.8704104307848684e-06, + "loss": 0.3567, + "step": 1896 + }, + { + "epoch": 1.364502787268477, + "grad_norm": 1.283835894083121, + "learning_rate": 2.870272052552407e-06, + "loss": 0.0189, + "step": 1897 + }, + { + "epoch": 1.365222082359288, + "grad_norm": 4.106284504778463, + "learning_rate": 2.8701336038170428e-06, + "loss": 0.0741, + "step": 1898 + }, + { + "epoch": 1.365941377450099, + "grad_norm": 3.5059526807046923, + "learning_rate": 2.8699950845858995e-06, + "loss": 0.0234, + "step": 1899 + }, + { + "epoch": 1.3666606725409098, + "grad_norm": 3.779700276545603, + "learning_rate": 2.8698564948661045e-06, + "loss": 0.0147, + "step": 1900 + }, + { + "epoch": 1.3673799676317209, + "grad_norm": 6.783809531972101, + "learning_rate": 2.8697178346647882e-06, + "loss": 0.2371, + "step": 1901 + }, + { + "epoch": 1.368099262722532, + "grad_norm": 0.9677415429901468, + "learning_rate": 2.8695791039890843e-06, + "loss": 0.0115, + "step": 1902 + }, + { + "epoch": 1.368818557813343, + "grad_norm": 6.727792863141242, + "learning_rate": 2.869440302846132e-06, + "loss": 0.2896, + "step": 1903 + }, + { + "epoch": 1.369537852904154, + "grad_norm": 4.147112617344063, + "learning_rate": 2.8693014312430718e-06, + "loss": 0.0431, + "step": 1904 + }, + { + "epoch": 1.370257147994965, + "grad_norm": 4.241363176441661, + "learning_rate": 2.869162489187049e-06, + "loss": 0.1397, + "step": 1905 + }, + { + "epoch": 1.3709764430857758, + "grad_norm": 6.687088618579141, + "learning_rate": 2.8690234766852132e-06, + "loss": 0.1491, + "step": 1906 + }, + { + "epoch": 1.3716957381765869, + "grad_norm": 3.7765268032910586, + "learning_rate": 2.868884393744716e-06, + "loss": 0.1477, + "step": 1907 + }, + { + "epoch": 1.3724150332673979, + "grad_norm": 4.139740864851264, + "learning_rate": 2.868745240372713e-06, + "loss": 0.0792, + "step": 1908 + }, + { + "epoch": 1.373134328358209, + "grad_norm": 3.9534063648908515, + "learning_rate": 2.868606016576366e-06, + "loss": 0.0556, + "step": 1909 + }, + { + "epoch": 1.37385362344902, + "grad_norm": 7.8093498646775545, + "learning_rate": 2.8684667223628362e-06, + "loss": 0.2441, + "step": 1910 + }, + { + "epoch": 1.374572918539831, + "grad_norm": 3.716601863584651, + "learning_rate": 2.868327357739291e-06, + "loss": 0.1437, + "step": 1911 + }, + { + "epoch": 1.375292213630642, + "grad_norm": 2.91209494429562, + "learning_rate": 2.8681879227129014e-06, + "loss": 0.0644, + "step": 1912 + }, + { + "epoch": 1.376011508721453, + "grad_norm": 8.512505309921577, + "learning_rate": 2.8680484172908416e-06, + "loss": 0.0947, + "step": 1913 + }, + { + "epoch": 1.376730803812264, + "grad_norm": 4.730720273817467, + "learning_rate": 2.867908841480289e-06, + "loss": 0.1724, + "step": 1914 + }, + { + "epoch": 1.377450098903075, + "grad_norm": 4.575301320873031, + "learning_rate": 2.8677691952884254e-06, + "loss": 0.0307, + "step": 1915 + }, + { + "epoch": 1.378169393993886, + "grad_norm": 2.7056726784039946, + "learning_rate": 2.8676294787224355e-06, + "loss": 0.0763, + "step": 1916 + }, + { + "epoch": 1.378888689084697, + "grad_norm": 5.9917559907685005, + "learning_rate": 2.867489691789508e-06, + "loss": 0.0435, + "step": 1917 + }, + { + "epoch": 1.379607984175508, + "grad_norm": 2.2920247880790816, + "learning_rate": 2.867349834496836e-06, + "loss": 0.0218, + "step": 1918 + }, + { + "epoch": 1.380327279266319, + "grad_norm": 4.274463707796197, + "learning_rate": 2.8672099068516137e-06, + "loss": 0.1012, + "step": 1919 + }, + { + "epoch": 1.38104657435713, + "grad_norm": 3.3508697117175243, + "learning_rate": 2.867069908861042e-06, + "loss": 0.1919, + "step": 1920 + }, + { + "epoch": 1.3817658694479409, + "grad_norm": 9.257193476715038, + "learning_rate": 2.866929840532324e-06, + "loss": 0.2183, + "step": 1921 + }, + { + "epoch": 1.382485164538752, + "grad_norm": 4.9685247566317985, + "learning_rate": 2.8667897018726663e-06, + "loss": 0.128, + "step": 1922 + }, + { + "epoch": 1.383204459629563, + "grad_norm": 0.1546847182092274, + "learning_rate": 2.866649492889279e-06, + "loss": 0.0009, + "step": 1923 + }, + { + "epoch": 1.383923754720374, + "grad_norm": 5.234305306662609, + "learning_rate": 2.8665092135893755e-06, + "loss": 0.0833, + "step": 1924 + }, + { + "epoch": 1.384643049811185, + "grad_norm": 4.213856902876313, + "learning_rate": 2.8663688639801747e-06, + "loss": 0.1818, + "step": 1925 + }, + { + "epoch": 1.385362344901996, + "grad_norm": 5.7795682626638305, + "learning_rate": 2.8662284440688973e-06, + "loss": 0.1123, + "step": 1926 + }, + { + "epoch": 1.386081639992807, + "grad_norm": 3.0411290605332306, + "learning_rate": 2.8660879538627676e-06, + "loss": 0.0811, + "step": 1927 + }, + { + "epoch": 1.3868009350836181, + "grad_norm": 1.4826076423309897, + "learning_rate": 2.8659473933690157e-06, + "loss": 0.0117, + "step": 1928 + }, + { + "epoch": 1.3875202301744292, + "grad_norm": 5.216213811121435, + "learning_rate": 2.8658067625948716e-06, + "loss": 0.1762, + "step": 1929 + }, + { + "epoch": 1.3882395252652402, + "grad_norm": 4.634746753642027, + "learning_rate": 2.865666061547572e-06, + "loss": 0.1135, + "step": 1930 + }, + { + "epoch": 1.388958820356051, + "grad_norm": 4.5587033326342095, + "learning_rate": 2.865525290234356e-06, + "loss": 0.086, + "step": 1931 + }, + { + "epoch": 1.389678115446862, + "grad_norm": 2.290336732356614, + "learning_rate": 2.8653844486624675e-06, + "loss": 0.0473, + "step": 1932 + }, + { + "epoch": 1.390397410537673, + "grad_norm": 7.910976969174364, + "learning_rate": 2.8652435368391522e-06, + "loss": 0.2336, + "step": 1933 + }, + { + "epoch": 1.391116705628484, + "grad_norm": 2.9833598994318775, + "learning_rate": 2.8651025547716598e-06, + "loss": 0.0678, + "step": 1934 + }, + { + "epoch": 1.3918360007192951, + "grad_norm": 3.9513792992593246, + "learning_rate": 2.8649615024672445e-06, + "loss": 0.125, + "step": 1935 + }, + { + "epoch": 1.3925552958101062, + "grad_norm": 3.6054297150420154, + "learning_rate": 2.864820379933164e-06, + "loss": 0.1097, + "step": 1936 + }, + { + "epoch": 1.393274590900917, + "grad_norm": 8.399979399115058, + "learning_rate": 2.8646791871766795e-06, + "loss": 0.1244, + "step": 1937 + }, + { + "epoch": 1.393993885991728, + "grad_norm": 4.452107154972758, + "learning_rate": 2.8645379242050546e-06, + "loss": 0.0801, + "step": 1938 + }, + { + "epoch": 1.394713181082539, + "grad_norm": 2.0151878951527493, + "learning_rate": 2.8643965910255584e-06, + "loss": 0.0397, + "step": 1939 + }, + { + "epoch": 1.39543247617335, + "grad_norm": 5.397892435682733, + "learning_rate": 2.8642551876454624e-06, + "loss": 0.1883, + "step": 1940 + }, + { + "epoch": 1.3961517712641611, + "grad_norm": 4.773125167591133, + "learning_rate": 2.8641137140720425e-06, + "loss": 0.026, + "step": 1941 + }, + { + "epoch": 1.3968710663549722, + "grad_norm": 4.9317297230557875, + "learning_rate": 2.863972170312577e-06, + "loss": 0.1169, + "step": 1942 + }, + { + "epoch": 1.3975903614457832, + "grad_norm": 7.644149189830625, + "learning_rate": 2.863830556374349e-06, + "loss": 0.1355, + "step": 1943 + }, + { + "epoch": 1.3983096565365942, + "grad_norm": 5.867470273592033, + "learning_rate": 2.8636888722646445e-06, + "loss": 0.1344, + "step": 1944 + }, + { + "epoch": 1.3990289516274053, + "grad_norm": 7.036349942610618, + "learning_rate": 2.863547117990754e-06, + "loss": 0.0937, + "step": 1945 + }, + { + "epoch": 1.399748246718216, + "grad_norm": 6.743646316357328, + "learning_rate": 2.863405293559971e-06, + "loss": 0.2231, + "step": 1946 + }, + { + "epoch": 1.400467541809027, + "grad_norm": 6.694266047154306, + "learning_rate": 2.8632633989795912e-06, + "loss": 0.1484, + "step": 1947 + }, + { + "epoch": 1.4011868368998381, + "grad_norm": 5.865500304378709, + "learning_rate": 2.863121434256917e-06, + "loss": 0.1529, + "step": 1948 + }, + { + "epoch": 1.4019061319906492, + "grad_norm": 3.2047204858265337, + "learning_rate": 2.8629793993992527e-06, + "loss": 0.0637, + "step": 1949 + }, + { + "epoch": 1.4026254270814602, + "grad_norm": 4.871879921497448, + "learning_rate": 2.862837294413905e-06, + "loss": 0.1437, + "step": 1950 + }, + { + "epoch": 1.4033447221722712, + "grad_norm": 7.193840915806507, + "learning_rate": 2.8626951193081856e-06, + "loss": 0.1608, + "step": 1951 + }, + { + "epoch": 1.404064017263082, + "grad_norm": 5.911742055638088, + "learning_rate": 2.8625528740894114e-06, + "loss": 0.0722, + "step": 1952 + }, + { + "epoch": 1.404783312353893, + "grad_norm": 8.623124966845783, + "learning_rate": 2.862410558764899e-06, + "loss": 0.3501, + "step": 1953 + }, + { + "epoch": 1.4055026074447041, + "grad_norm": 5.022809816515642, + "learning_rate": 2.8622681733419724e-06, + "loss": 0.1108, + "step": 1954 + }, + { + "epoch": 1.4062219025355152, + "grad_norm": 2.080315819619389, + "learning_rate": 2.862125717827956e-06, + "loss": 0.0042, + "step": 1955 + }, + { + "epoch": 1.4069411976263262, + "grad_norm": 6.969493607240946, + "learning_rate": 2.861983192230181e-06, + "loss": 0.222, + "step": 1956 + }, + { + "epoch": 1.4076604927171372, + "grad_norm": 7.1644543279506445, + "learning_rate": 2.86184059655598e-06, + "loss": 0.0592, + "step": 1957 + }, + { + "epoch": 1.4083797878079483, + "grad_norm": 4.117975356153995, + "learning_rate": 2.861697930812689e-06, + "loss": 0.1629, + "step": 1958 + }, + { + "epoch": 1.4090990828987593, + "grad_norm": 3.9833930394152643, + "learning_rate": 2.8615551950076496e-06, + "loss": 0.1811, + "step": 1959 + }, + { + "epoch": 1.4098183779895703, + "grad_norm": 2.1241573575202835, + "learning_rate": 2.861412389148205e-06, + "loss": 0.0064, + "step": 1960 + }, + { + "epoch": 1.4105376730803814, + "grad_norm": 6.982257172465749, + "learning_rate": 2.8612695132417037e-06, + "loss": 0.088, + "step": 1961 + }, + { + "epoch": 1.4112569681711922, + "grad_norm": 2.871719751941688, + "learning_rate": 2.8611265672954954e-06, + "loss": 0.08, + "step": 1962 + }, + { + "epoch": 1.4119762632620032, + "grad_norm": 4.6275268980567335, + "learning_rate": 2.8609835513169364e-06, + "loss": 0.1588, + "step": 1963 + }, + { + "epoch": 1.4126955583528142, + "grad_norm": 3.7635037556838276, + "learning_rate": 2.8608404653133843e-06, + "loss": 0.0447, + "step": 1964 + }, + { + "epoch": 1.4134148534436253, + "grad_norm": 4.483019355120998, + "learning_rate": 2.8606973092922017e-06, + "loss": 0.1542, + "step": 1965 + }, + { + "epoch": 1.4141341485344363, + "grad_norm": 5.338802657060233, + "learning_rate": 2.8605540832607533e-06, + "loss": 0.1126, + "step": 1966 + }, + { + "epoch": 1.4148534436252471, + "grad_norm": 3.7953944018148422, + "learning_rate": 2.8604107872264094e-06, + "loss": 0.109, + "step": 1967 + }, + { + "epoch": 1.4155727387160582, + "grad_norm": 4.2957269082654665, + "learning_rate": 2.8602674211965424e-06, + "loss": 0.1856, + "step": 1968 + }, + { + "epoch": 1.4162920338068692, + "grad_norm": 8.252263302831576, + "learning_rate": 2.8601239851785287e-06, + "loss": 0.0896, + "step": 1969 + }, + { + "epoch": 1.4170113288976802, + "grad_norm": 1.4726144052907457, + "learning_rate": 2.859980479179748e-06, + "loss": 0.0347, + "step": 1970 + }, + { + "epoch": 1.4177306239884913, + "grad_norm": 4.323203799427797, + "learning_rate": 2.8598369032075848e-06, + "loss": 0.1692, + "step": 1971 + }, + { + "epoch": 1.4184499190793023, + "grad_norm": 2.5610834467759713, + "learning_rate": 2.859693257269426e-06, + "loss": 0.0318, + "step": 1972 + }, + { + "epoch": 1.4191692141701133, + "grad_norm": 5.7596271294171695, + "learning_rate": 2.859549541372661e-06, + "loss": 0.0675, + "step": 1973 + }, + { + "epoch": 1.4198885092609244, + "grad_norm": 6.049257903513594, + "learning_rate": 2.8594057555246864e-06, + "loss": 0.1191, + "step": 1974 + }, + { + "epoch": 1.4206078043517354, + "grad_norm": 3.3476359232453765, + "learning_rate": 2.8592618997328993e-06, + "loss": 0.0815, + "step": 1975 + }, + { + "epoch": 1.4213270994425464, + "grad_norm": 5.203874376363021, + "learning_rate": 2.859117974004701e-06, + "loss": 0.1282, + "step": 1976 + }, + { + "epoch": 1.4220463945333572, + "grad_norm": 6.927045448765321, + "learning_rate": 2.8589739783474977e-06, + "loss": 0.2118, + "step": 1977 + }, + { + "epoch": 1.4227656896241683, + "grad_norm": 6.792837192376061, + "learning_rate": 2.858829912768697e-06, + "loss": 0.1443, + "step": 1978 + }, + { + "epoch": 1.4234849847149793, + "grad_norm": 5.64689845317184, + "learning_rate": 2.858685777275712e-06, + "loss": 0.1448, + "step": 1979 + }, + { + "epoch": 1.4242042798057903, + "grad_norm": 4.357875012857252, + "learning_rate": 2.858541571875959e-06, + "loss": 0.1041, + "step": 1980 + }, + { + "epoch": 1.4249235748966014, + "grad_norm": 4.810267602299373, + "learning_rate": 2.858397296576857e-06, + "loss": 0.0146, + "step": 1981 + }, + { + "epoch": 1.4256428699874124, + "grad_norm": 4.061077942725784, + "learning_rate": 2.8582529513858298e-06, + "loss": 0.0727, + "step": 1982 + }, + { + "epoch": 1.4263621650782232, + "grad_norm": 3.1351744387694906, + "learning_rate": 2.8581085363103036e-06, + "loss": 0.0809, + "step": 1983 + }, + { + "epoch": 1.4270814601690343, + "grad_norm": 3.1452340296216574, + "learning_rate": 2.8579640513577094e-06, + "loss": 0.0736, + "step": 1984 + }, + { + "epoch": 1.4278007552598453, + "grad_norm": 5.15105769766964, + "learning_rate": 2.8578194965354807e-06, + "loss": 0.0563, + "step": 1985 + }, + { + "epoch": 1.4285200503506563, + "grad_norm": 7.113597312031152, + "learning_rate": 2.8576748718510553e-06, + "loss": 0.278, + "step": 1986 + }, + { + "epoch": 1.4292393454414674, + "grad_norm": 6.395007254471104, + "learning_rate": 2.8575301773118744e-06, + "loss": 0.1952, + "step": 1987 + }, + { + "epoch": 1.4299586405322784, + "grad_norm": 6.705746236761664, + "learning_rate": 2.857385412925383e-06, + "loss": 0.2438, + "step": 1988 + }, + { + "epoch": 1.4306779356230894, + "grad_norm": 8.292862561319234, + "learning_rate": 2.8572405786990296e-06, + "loss": 0.2797, + "step": 1989 + }, + { + "epoch": 1.4313972307139005, + "grad_norm": 2.0034687573800594, + "learning_rate": 2.8570956746402653e-06, + "loss": 0.0477, + "step": 1990 + }, + { + "epoch": 1.4321165258047115, + "grad_norm": 5.050066848276019, + "learning_rate": 2.8569507007565463e-06, + "loss": 0.0882, + "step": 1991 + }, + { + "epoch": 1.4328358208955223, + "grad_norm": 2.437126982427539, + "learning_rate": 2.856805657055332e-06, + "loss": 0.0069, + "step": 1992 + }, + { + "epoch": 1.4335551159863333, + "grad_norm": 5.794620779332187, + "learning_rate": 2.8566605435440846e-06, + "loss": 0.0938, + "step": 1993 + }, + { + "epoch": 1.4342744110771444, + "grad_norm": 5.756558951752547, + "learning_rate": 2.856515360230271e-06, + "loss": 0.1184, + "step": 1994 + }, + { + "epoch": 1.4349937061679554, + "grad_norm": 5.573712647649542, + "learning_rate": 2.8563701071213603e-06, + "loss": 0.2207, + "step": 1995 + }, + { + "epoch": 1.4357130012587664, + "grad_norm": 3.857616752790406, + "learning_rate": 2.856224784224827e-06, + "loss": 0.1051, + "step": 1996 + }, + { + "epoch": 1.4364322963495775, + "grad_norm": 4.030817745454483, + "learning_rate": 2.8560793915481476e-06, + "loss": 0.1789, + "step": 1997 + }, + { + "epoch": 1.4371515914403883, + "grad_norm": 3.3241275062078977, + "learning_rate": 2.8559339290988028e-06, + "loss": 0.0269, + "step": 1998 + }, + { + "epoch": 1.4378708865311993, + "grad_norm": 2.2123817156316297, + "learning_rate": 2.855788396884277e-06, + "loss": 0.0075, + "step": 1999 + }, + { + "epoch": 1.4385901816220104, + "grad_norm": 6.6883924643440995, + "learning_rate": 2.8556427949120587e-06, + "loss": 0.2912, + "step": 2000 + }, + { + "epoch": 1.4393094767128214, + "grad_norm": 6.6418425700081425, + "learning_rate": 2.8554971231896387e-06, + "loss": 0.1646, + "step": 2001 + }, + { + "epoch": 1.4400287718036324, + "grad_norm": 4.106760965615277, + "learning_rate": 2.8553513817245117e-06, + "loss": 0.0397, + "step": 2002 + }, + { + "epoch": 1.4407480668944435, + "grad_norm": 6.027272729648282, + "learning_rate": 2.8552055705241777e-06, + "loss": 0.0673, + "step": 2003 + }, + { + "epoch": 1.4414673619852545, + "grad_norm": 5.069987811777948, + "learning_rate": 2.8550596895961373e-06, + "loss": 0.1126, + "step": 2004 + }, + { + "epoch": 1.4421866570760655, + "grad_norm": 4.201847354434991, + "learning_rate": 2.8549137389478977e-06, + "loss": 0.1438, + "step": 2005 + }, + { + "epoch": 1.4429059521668766, + "grad_norm": 3.975616438255497, + "learning_rate": 2.8547677185869673e-06, + "loss": 0.1163, + "step": 2006 + }, + { + "epoch": 1.4436252472576876, + "grad_norm": 3.5032386694636113, + "learning_rate": 2.8546216285208603e-06, + "loss": 0.184, + "step": 2007 + }, + { + "epoch": 1.4443445423484984, + "grad_norm": 8.388021678809483, + "learning_rate": 2.854475468757092e-06, + "loss": 0.2439, + "step": 2008 + }, + { + "epoch": 1.4450638374393094, + "grad_norm": 4.663240467198657, + "learning_rate": 2.854329239303183e-06, + "loss": 0.0992, + "step": 2009 + }, + { + "epoch": 1.4457831325301205, + "grad_norm": 6.709242882617589, + "learning_rate": 2.8541829401666576e-06, + "loss": 0.1276, + "step": 2010 + }, + { + "epoch": 1.4465024276209315, + "grad_norm": 1.0214979926387269, + "learning_rate": 2.8540365713550427e-06, + "loss": 0.0244, + "step": 2011 + }, + { + "epoch": 1.4472217227117425, + "grad_norm": 5.602976734843479, + "learning_rate": 2.8538901328758686e-06, + "loss": 0.2921, + "step": 2012 + }, + { + "epoch": 1.4479410178025536, + "grad_norm": 5.639849342658185, + "learning_rate": 2.853743624736671e-06, + "loss": 0.1666, + "step": 2013 + }, + { + "epoch": 1.4486603128933644, + "grad_norm": 4.030356849308093, + "learning_rate": 2.853597046944988e-06, + "loss": 0.0789, + "step": 2014 + }, + { + "epoch": 1.4493796079841754, + "grad_norm": 1.9121352050018268, + "learning_rate": 2.85345039950836e-06, + "loss": 0.02, + "step": 2015 + }, + { + "epoch": 1.4500989030749865, + "grad_norm": 4.922742816632193, + "learning_rate": 2.853303682434333e-06, + "loss": 0.0197, + "step": 2016 + }, + { + "epoch": 1.4508181981657975, + "grad_norm": 3.6686333502637885, + "learning_rate": 2.8531568957304567e-06, + "loss": 0.1278, + "step": 2017 + }, + { + "epoch": 1.4515374932566085, + "grad_norm": 7.560223797392064, + "learning_rate": 2.8530100394042818e-06, + "loss": 0.365, + "step": 2018 + }, + { + "epoch": 1.4522567883474196, + "grad_norm": 5.954246705072298, + "learning_rate": 2.8528631134633657e-06, + "loss": 0.0735, + "step": 2019 + }, + { + "epoch": 1.4529760834382306, + "grad_norm": 4.8185372937658295, + "learning_rate": 2.8527161179152675e-06, + "loss": 0.0334, + "step": 2020 + }, + { + "epoch": 1.4536953785290416, + "grad_norm": 3.8525377363520157, + "learning_rate": 2.8525690527675503e-06, + "loss": 0.0583, + "step": 2021 + }, + { + "epoch": 1.4544146736198527, + "grad_norm": 2.8681137010864366, + "learning_rate": 2.852421918027781e-06, + "loss": 0.0701, + "step": 2022 + }, + { + "epoch": 1.4551339687106635, + "grad_norm": 1.4605679622233498, + "learning_rate": 2.85227471370353e-06, + "loss": 0.0138, + "step": 2023 + }, + { + "epoch": 1.4558532638014745, + "grad_norm": 2.2306526101979873, + "learning_rate": 2.852127439802371e-06, + "loss": 0.0083, + "step": 2024 + }, + { + "epoch": 1.4565725588922855, + "grad_norm": 6.412544721770911, + "learning_rate": 2.8519800963318817e-06, + "loss": 0.142, + "step": 2025 + }, + { + "epoch": 1.4572918539830966, + "grad_norm": 2.2109584020261424, + "learning_rate": 2.851832683299643e-06, + "loss": 0.0693, + "step": 2026 + }, + { + "epoch": 1.4580111490739076, + "grad_norm": 3.100725321871934, + "learning_rate": 2.85168520071324e-06, + "loss": 0.0323, + "step": 2027 + }, + { + "epoch": 1.4587304441647186, + "grad_norm": 4.2648413371460565, + "learning_rate": 2.8515376485802603e-06, + "loss": 0.2209, + "step": 2028 + }, + { + "epoch": 1.4594497392555295, + "grad_norm": 5.760164744133177, + "learning_rate": 2.8513900269082963e-06, + "loss": 0.0416, + "step": 2029 + }, + { + "epoch": 1.4601690343463405, + "grad_norm": 1.8832845216831633, + "learning_rate": 2.851242335704943e-06, + "loss": 0.0745, + "step": 2030 + }, + { + "epoch": 1.4608883294371515, + "grad_norm": 3.3336648414078622, + "learning_rate": 2.8510945749777997e-06, + "loss": 0.0488, + "step": 2031 + }, + { + "epoch": 1.4616076245279626, + "grad_norm": 2.8156889918014176, + "learning_rate": 2.8509467447344685e-06, + "loss": 0.0635, + "step": 2032 + }, + { + "epoch": 1.4623269196187736, + "grad_norm": 2.7023495191631186, + "learning_rate": 2.8507988449825556e-06, + "loss": 0.0592, + "step": 2033 + }, + { + "epoch": 1.4630462147095846, + "grad_norm": 3.2573707948615764, + "learning_rate": 2.850650875729671e-06, + "loss": 0.03, + "step": 2034 + }, + { + "epoch": 1.4637655098003957, + "grad_norm": 6.956052131891015, + "learning_rate": 2.850502836983428e-06, + "loss": 0.1123, + "step": 2035 + }, + { + "epoch": 1.4644848048912067, + "grad_norm": 3.7428585578782525, + "learning_rate": 2.850354728751443e-06, + "loss": 0.0842, + "step": 2036 + }, + { + "epoch": 1.4652040999820177, + "grad_norm": 3.805828719010263, + "learning_rate": 2.8502065510413373e-06, + "loss": 0.0932, + "step": 2037 + }, + { + "epoch": 1.4659233950728288, + "grad_norm": 4.28252119918998, + "learning_rate": 2.8500583038607338e-06, + "loss": 0.1393, + "step": 2038 + }, + { + "epoch": 1.4666426901636396, + "grad_norm": 7.973858689388834, + "learning_rate": 2.8499099872172613e-06, + "loss": 0.1917, + "step": 2039 + }, + { + "epoch": 1.4673619852544506, + "grad_norm": 8.223178114991434, + "learning_rate": 2.8497616011185496e-06, + "loss": 0.0656, + "step": 2040 + }, + { + "epoch": 1.4680812803452616, + "grad_norm": 5.168299555372078, + "learning_rate": 2.8496131455722342e-06, + "loss": 0.1017, + "step": 2041 + }, + { + "epoch": 1.4688005754360727, + "grad_norm": 3.559538930193632, + "learning_rate": 2.8494646205859537e-06, + "loss": 0.1353, + "step": 2042 + }, + { + "epoch": 1.4695198705268837, + "grad_norm": 5.266724370704678, + "learning_rate": 2.849316026167349e-06, + "loss": 0.0966, + "step": 2043 + }, + { + "epoch": 1.4702391656176945, + "grad_norm": 2.809598015565207, + "learning_rate": 2.849167362324066e-06, + "loss": 0.0221, + "step": 2044 + }, + { + "epoch": 1.4709584607085056, + "grad_norm": 4.1877156425923125, + "learning_rate": 2.8490186290637544e-06, + "loss": 0.0822, + "step": 2045 + }, + { + "epoch": 1.4716777557993166, + "grad_norm": 4.088972091034248, + "learning_rate": 2.8488698263940664e-06, + "loss": 0.1396, + "step": 2046 + }, + { + "epoch": 1.4723970508901276, + "grad_norm": 2.109749846862112, + "learning_rate": 2.8487209543226572e-06, + "loss": 0.0369, + "step": 2047 + }, + { + "epoch": 1.4731163459809387, + "grad_norm": 2.508024449228336, + "learning_rate": 2.8485720128571874e-06, + "loss": 0.0647, + "step": 2048 + }, + { + "epoch": 1.4738356410717497, + "grad_norm": 2.776322273845714, + "learning_rate": 2.8484230020053206e-06, + "loss": 0.0856, + "step": 2049 + }, + { + "epoch": 1.4745549361625607, + "grad_norm": 7.516425647591088, + "learning_rate": 2.848273921774723e-06, + "loss": 0.094, + "step": 2050 + }, + { + "epoch": 1.4752742312533718, + "grad_norm": 2.093822245357266, + "learning_rate": 2.8481247721730657e-06, + "loss": 0.0419, + "step": 2051 + }, + { + "epoch": 1.4759935263441828, + "grad_norm": 6.340528833023023, + "learning_rate": 2.847975553208022e-06, + "loss": 0.2125, + "step": 2052 + }, + { + "epoch": 1.4767128214349938, + "grad_norm": 7.041347015291274, + "learning_rate": 2.8478262648872696e-06, + "loss": 0.2156, + "step": 2053 + }, + { + "epoch": 1.4774321165258046, + "grad_norm": 3.432607818547082, + "learning_rate": 2.8476769072184897e-06, + "loss": 0.0273, + "step": 2054 + }, + { + "epoch": 1.4781514116166157, + "grad_norm": 4.094829800066383, + "learning_rate": 2.8475274802093675e-06, + "loss": 0.1143, + "step": 2055 + }, + { + "epoch": 1.4788707067074267, + "grad_norm": 4.5262466889442585, + "learning_rate": 2.847377983867591e-06, + "loss": 0.1708, + "step": 2056 + }, + { + "epoch": 1.4795900017982377, + "grad_norm": 2.0823664496833936, + "learning_rate": 2.8472284182008523e-06, + "loss": 0.0622, + "step": 2057 + }, + { + "epoch": 1.4803092968890488, + "grad_norm": 4.501304531567463, + "learning_rate": 2.847078783216846e-06, + "loss": 0.1086, + "step": 2058 + }, + { + "epoch": 1.4810285919798598, + "grad_norm": 2.037806446765926, + "learning_rate": 2.8469290789232715e-06, + "loss": 0.0504, + "step": 2059 + }, + { + "epoch": 1.4817478870706706, + "grad_norm": 2.4295263506985254, + "learning_rate": 2.8467793053278316e-06, + "loss": 0.0509, + "step": 2060 + }, + { + "epoch": 1.4824671821614817, + "grad_norm": 4.652872837521685, + "learning_rate": 2.8466294624382323e-06, + "loss": 0.1332, + "step": 2061 + }, + { + "epoch": 1.4831864772522927, + "grad_norm": 6.830314820488075, + "learning_rate": 2.8464795502621833e-06, + "loss": 0.2922, + "step": 2062 + }, + { + "epoch": 1.4839057723431037, + "grad_norm": 4.253866968653982, + "learning_rate": 2.8463295688073976e-06, + "loss": 0.1309, + "step": 2063 + }, + { + "epoch": 1.4846250674339148, + "grad_norm": 5.088782440817938, + "learning_rate": 2.8461795180815926e-06, + "loss": 0.0897, + "step": 2064 + }, + { + "epoch": 1.4853443625247258, + "grad_norm": 3.3649566755197364, + "learning_rate": 2.8460293980924876e-06, + "loss": 0.0984, + "step": 2065 + }, + { + "epoch": 1.4860636576155368, + "grad_norm": 6.0273737748079075, + "learning_rate": 2.8458792088478072e-06, + "loss": 0.1303, + "step": 2066 + }, + { + "epoch": 1.4867829527063479, + "grad_norm": 8.346312372214113, + "learning_rate": 2.8457289503552794e-06, + "loss": 0.3255, + "step": 2067 + }, + { + "epoch": 1.487502247797159, + "grad_norm": 7.722930486457627, + "learning_rate": 2.8455786226226345e-06, + "loss": 0.2197, + "step": 2068 + }, + { + "epoch": 1.48822154288797, + "grad_norm": 4.1457954984655085, + "learning_rate": 2.8454282256576073e-06, + "loss": 0.1395, + "step": 2069 + }, + { + "epoch": 1.4889408379787807, + "grad_norm": 2.992790058022124, + "learning_rate": 2.845277759467936e-06, + "loss": 0.0806, + "step": 2070 + }, + { + "epoch": 1.4896601330695918, + "grad_norm": 3.988386463574146, + "learning_rate": 2.8451272240613626e-06, + "loss": 0.0759, + "step": 2071 + }, + { + "epoch": 1.4903794281604028, + "grad_norm": 4.860492825399819, + "learning_rate": 2.8449766194456324e-06, + "loss": 0.1433, + "step": 2072 + }, + { + "epoch": 1.4910987232512138, + "grad_norm": 3.730774308350756, + "learning_rate": 2.8448259456284936e-06, + "loss": 0.1141, + "step": 2073 + }, + { + "epoch": 1.4918180183420249, + "grad_norm": 3.162546996326205, + "learning_rate": 2.8446752026176997e-06, + "loss": 0.0965, + "step": 2074 + }, + { + "epoch": 1.4925373134328357, + "grad_norm": 3.013932178510901, + "learning_rate": 2.8445243904210052e-06, + "loss": 0.0767, + "step": 2075 + }, + { + "epoch": 1.4932566085236467, + "grad_norm": 8.156451224532342, + "learning_rate": 2.8443735090461715e-06, + "loss": 0.2668, + "step": 2076 + }, + { + "epoch": 1.4939759036144578, + "grad_norm": 7.42329868406929, + "learning_rate": 2.8442225585009604e-06, + "loss": 0.0384, + "step": 2077 + }, + { + "epoch": 1.4946951987052688, + "grad_norm": 6.91169662433428, + "learning_rate": 2.844071538793139e-06, + "loss": 0.0886, + "step": 2078 + }, + { + "epoch": 1.4954144937960798, + "grad_norm": 5.8427580964208286, + "learning_rate": 2.843920449930478e-06, + "loss": 0.3789, + "step": 2079 + }, + { + "epoch": 1.4961337888868909, + "grad_norm": 0.8940149760610763, + "learning_rate": 2.8437692919207497e-06, + "loss": 0.0159, + "step": 2080 + }, + { + "epoch": 1.496853083977702, + "grad_norm": 4.320881760156023, + "learning_rate": 2.843618064771733e-06, + "loss": 0.094, + "step": 2081 + }, + { + "epoch": 1.497572379068513, + "grad_norm": 3.0063002444265416, + "learning_rate": 2.8434667684912085e-06, + "loss": 0.011, + "step": 2082 + }, + { + "epoch": 1.498291674159324, + "grad_norm": 5.59085829591869, + "learning_rate": 2.84331540308696e-06, + "loss": 0.2387, + "step": 2083 + }, + { + "epoch": 1.499010969250135, + "grad_norm": 6.100495860841695, + "learning_rate": 2.843163968566776e-06, + "loss": 0.1665, + "step": 2084 + }, + { + "epoch": 1.4997302643409458, + "grad_norm": 3.0958914749470394, + "learning_rate": 2.8430124649384477e-06, + "loss": 0.0429, + "step": 2085 + }, + { + "epoch": 1.5004495594317568, + "grad_norm": 4.280542763861561, + "learning_rate": 2.842860892209771e-06, + "loss": 0.0997, + "step": 2086 + }, + { + "epoch": 1.5011688545225679, + "grad_norm": 4.056064159836871, + "learning_rate": 2.8427092503885435e-06, + "loss": 0.1334, + "step": 2087 + }, + { + "epoch": 1.501888149613379, + "grad_norm": 6.841318298812241, + "learning_rate": 2.842557539482568e-06, + "loss": 0.1717, + "step": 2088 + }, + { + "epoch": 1.50260744470419, + "grad_norm": 0.3478902733757561, + "learning_rate": 2.842405759499651e-06, + "loss": 0.0014, + "step": 2089 + }, + { + "epoch": 1.5033267397950008, + "grad_norm": 5.5537515948259575, + "learning_rate": 2.8422539104476008e-06, + "loss": 0.0518, + "step": 2090 + }, + { + "epoch": 1.5040460348858118, + "grad_norm": 4.4363030448350385, + "learning_rate": 2.84210199233423e-06, + "loss": 0.0504, + "step": 2091 + }, + { + "epoch": 1.5047653299766228, + "grad_norm": 4.944684416211211, + "learning_rate": 2.841950005167357e-06, + "loss": 0.1242, + "step": 2092 + }, + { + "epoch": 1.5054846250674339, + "grad_norm": 5.592434384718533, + "learning_rate": 2.8417979489548e-06, + "loss": 0.2782, + "step": 2093 + }, + { + "epoch": 1.506203920158245, + "grad_norm": 5.1179840008794475, + "learning_rate": 2.8416458237043824e-06, + "loss": 0.2164, + "step": 2094 + }, + { + "epoch": 1.506923215249056, + "grad_norm": 3.3254603002561804, + "learning_rate": 2.8414936294239327e-06, + "loss": 0.0649, + "step": 2095 + }, + { + "epoch": 1.507642510339867, + "grad_norm": 5.490782884881453, + "learning_rate": 2.8413413661212804e-06, + "loss": 0.1622, + "step": 2096 + }, + { + "epoch": 1.508361805430678, + "grad_norm": 5.602155450846729, + "learning_rate": 2.8411890338042607e-06, + "loss": 0.1257, + "step": 2097 + }, + { + "epoch": 1.509081100521489, + "grad_norm": 4.74006037373041, + "learning_rate": 2.8410366324807106e-06, + "loss": 0.1273, + "step": 2098 + }, + { + "epoch": 1.5098003956123, + "grad_norm": 4.810121949424923, + "learning_rate": 2.8408841621584715e-06, + "loss": 0.3123, + "step": 2099 + }, + { + "epoch": 1.510519690703111, + "grad_norm": 3.081775619299007, + "learning_rate": 2.8407316228453883e-06, + "loss": 0.0718, + "step": 2100 + }, + { + "epoch": 1.511238985793922, + "grad_norm": 2.262127417717525, + "learning_rate": 2.8405790145493096e-06, + "loss": 0.0666, + "step": 2101 + }, + { + "epoch": 1.511958280884733, + "grad_norm": 3.7019262391922627, + "learning_rate": 2.8404263372780874e-06, + "loss": 0.1146, + "step": 2102 + }, + { + "epoch": 1.512677575975544, + "grad_norm": 2.4618931060105673, + "learning_rate": 2.8402735910395774e-06, + "loss": 0.0159, + "step": 2103 + }, + { + "epoch": 1.513396871066355, + "grad_norm": 3.6972846953421024, + "learning_rate": 2.8401207758416376e-06, + "loss": 0.1453, + "step": 2104 + }, + { + "epoch": 1.5141161661571658, + "grad_norm": 2.796573496131828, + "learning_rate": 2.839967891692132e-06, + "loss": 0.0229, + "step": 2105 + }, + { + "epoch": 1.5148354612479769, + "grad_norm": 3.491710770460096, + "learning_rate": 2.839814938598926e-06, + "loss": 0.122, + "step": 2106 + }, + { + "epoch": 1.515554756338788, + "grad_norm": 2.9846426227939897, + "learning_rate": 2.8396619165698897e-06, + "loss": 0.0535, + "step": 2107 + }, + { + "epoch": 1.516274051429599, + "grad_norm": 4.27477578987767, + "learning_rate": 2.8395088256128955e-06, + "loss": 0.0136, + "step": 2108 + }, + { + "epoch": 1.51699334652041, + "grad_norm": 4.233778688584056, + "learning_rate": 2.839355665735821e-06, + "loss": 0.0798, + "step": 2109 + }, + { + "epoch": 1.517712641611221, + "grad_norm": 3.401993069318298, + "learning_rate": 2.8392024369465464e-06, + "loss": 0.1459, + "step": 2110 + }, + { + "epoch": 1.518431936702032, + "grad_norm": 6.029752369915126, + "learning_rate": 2.8390491392529557e-06, + "loss": 0.1322, + "step": 2111 + }, + { + "epoch": 1.519151231792843, + "grad_norm": 4.32490982016153, + "learning_rate": 2.8388957726629362e-06, + "loss": 0.0405, + "step": 2112 + }, + { + "epoch": 1.519870526883654, + "grad_norm": 4.280474133523757, + "learning_rate": 2.838742337184378e-06, + "loss": 0.1767, + "step": 2113 + }, + { + "epoch": 1.5205898219744651, + "grad_norm": 7.766998333039946, + "learning_rate": 2.8385888328251778e-06, + "loss": 0.0744, + "step": 2114 + }, + { + "epoch": 1.5213091170652762, + "grad_norm": 3.926046147946892, + "learning_rate": 2.8384352595932313e-06, + "loss": 0.0517, + "step": 2115 + }, + { + "epoch": 1.5220284121560872, + "grad_norm": 3.2565401292161242, + "learning_rate": 2.8382816174964415e-06, + "loss": 0.0967, + "step": 2116 + }, + { + "epoch": 1.522747707246898, + "grad_norm": 3.4892653674603062, + "learning_rate": 2.8381279065427133e-06, + "loss": 0.0671, + "step": 2117 + }, + { + "epoch": 1.523467002337709, + "grad_norm": 9.527947159507624, + "learning_rate": 2.837974126739955e-06, + "loss": 0.3026, + "step": 2118 + }, + { + "epoch": 1.52418629742852, + "grad_norm": 5.393193564746958, + "learning_rate": 2.8378202780960796e-06, + "loss": 0.1253, + "step": 2119 + }, + { + "epoch": 1.524905592519331, + "grad_norm": 5.206110148094954, + "learning_rate": 2.837666360619002e-06, + "loss": 0.1697, + "step": 2120 + }, + { + "epoch": 1.525624887610142, + "grad_norm": 4.556115272105652, + "learning_rate": 2.837512374316642e-06, + "loss": 0.1253, + "step": 2121 + }, + { + "epoch": 1.526344182700953, + "grad_norm": 7.753247368191384, + "learning_rate": 2.8373583191969223e-06, + "loss": 0.1275, + "step": 2122 + }, + { + "epoch": 1.527063477791764, + "grad_norm": 2.8192905021359613, + "learning_rate": 2.8372041952677698e-06, + "loss": 0.0379, + "step": 2123 + }, + { + "epoch": 1.527782772882575, + "grad_norm": 4.6326478281778485, + "learning_rate": 2.8370500025371138e-06, + "loss": 0.0361, + "step": 2124 + }, + { + "epoch": 1.528502067973386, + "grad_norm": 5.233258207349289, + "learning_rate": 2.8368957410128885e-06, + "loss": 0.0796, + "step": 2125 + }, + { + "epoch": 1.529221363064197, + "grad_norm": 6.6001008617478245, + "learning_rate": 2.83674141070303e-06, + "loss": 0.045, + "step": 2126 + }, + { + "epoch": 1.5299406581550081, + "grad_norm": 4.955776387563965, + "learning_rate": 2.8365870116154796e-06, + "loss": 0.1265, + "step": 2127 + }, + { + "epoch": 1.5306599532458192, + "grad_norm": 4.485594805529013, + "learning_rate": 2.8364325437581805e-06, + "loss": 0.1343, + "step": 2128 + }, + { + "epoch": 1.5313792483366302, + "grad_norm": 2.7781981584041135, + "learning_rate": 2.8362780071390818e-06, + "loss": 0.0713, + "step": 2129 + }, + { + "epoch": 1.5320985434274412, + "grad_norm": 4.578639959764174, + "learning_rate": 2.8361234017661336e-06, + "loss": 0.0697, + "step": 2130 + }, + { + "epoch": 1.5328178385182523, + "grad_norm": 1.5602621036185553, + "learning_rate": 2.835968727647291e-06, + "loss": 0.0046, + "step": 2131 + }, + { + "epoch": 1.533537133609063, + "grad_norm": 5.95512753043447, + "learning_rate": 2.8358139847905116e-06, + "loss": 0.1301, + "step": 2132 + }, + { + "epoch": 1.534256428699874, + "grad_norm": 4.202836573840528, + "learning_rate": 2.8356591732037583e-06, + "loss": 0.2462, + "step": 2133 + }, + { + "epoch": 1.5349757237906851, + "grad_norm": 4.508608078966251, + "learning_rate": 2.835504292894996e-06, + "loss": 0.1211, + "step": 2134 + }, + { + "epoch": 1.5356950188814962, + "grad_norm": 3.145227810248748, + "learning_rate": 2.835349343872193e-06, + "loss": 0.1013, + "step": 2135 + }, + { + "epoch": 1.536414313972307, + "grad_norm": 2.3072490212915695, + "learning_rate": 2.835194326143322e-06, + "loss": 0.0485, + "step": 2136 + }, + { + "epoch": 1.537133609063118, + "grad_norm": 5.600000829856032, + "learning_rate": 2.835039239716359e-06, + "loss": 0.1174, + "step": 2137 + }, + { + "epoch": 1.537852904153929, + "grad_norm": 7.163622283089642, + "learning_rate": 2.8348840845992843e-06, + "loss": 0.2072, + "step": 2138 + }, + { + "epoch": 1.53857219924474, + "grad_norm": 5.641013139238723, + "learning_rate": 2.8347288608000796e-06, + "loss": 0.1043, + "step": 2139 + }, + { + "epoch": 1.5392914943355511, + "grad_norm": 5.000928066496515, + "learning_rate": 2.834573568326732e-06, + "loss": 0.1205, + "step": 2140 + }, + { + "epoch": 1.5400107894263622, + "grad_norm": 5.483672257956804, + "learning_rate": 2.8344182071872316e-06, + "loss": 0.0393, + "step": 2141 + }, + { + "epoch": 1.5407300845171732, + "grad_norm": 8.45431669748343, + "learning_rate": 2.8342627773895717e-06, + "loss": 0.14, + "step": 2142 + }, + { + "epoch": 1.5414493796079842, + "grad_norm": 5.670442439429874, + "learning_rate": 2.83410727894175e-06, + "loss": 0.2208, + "step": 2143 + }, + { + "epoch": 1.5421686746987953, + "grad_norm": 2.172276367616782, + "learning_rate": 2.833951711851767e-06, + "loss": 0.1047, + "step": 2144 + }, + { + "epoch": 1.5428879697896063, + "grad_norm": 4.652407188300584, + "learning_rate": 2.8337960761276263e-06, + "loss": 0.1199, + "step": 2145 + }, + { + "epoch": 1.5436072648804173, + "grad_norm": 3.3518755937829177, + "learning_rate": 2.8336403717773364e-06, + "loss": 0.087, + "step": 2146 + }, + { + "epoch": 1.5443265599712281, + "grad_norm": 5.78383195192292, + "learning_rate": 2.833484598808908e-06, + "loss": 0.3425, + "step": 2147 + }, + { + "epoch": 1.5450458550620392, + "grad_norm": 1.9750756454812182, + "learning_rate": 2.833328757230356e-06, + "loss": 0.0176, + "step": 2148 + }, + { + "epoch": 1.5457651501528502, + "grad_norm": 1.3809507169742352, + "learning_rate": 2.8331728470496998e-06, + "loss": 0.0079, + "step": 2149 + }, + { + "epoch": 1.5464844452436612, + "grad_norm": 5.144868372888545, + "learning_rate": 2.833016868274959e-06, + "loss": 0.0664, + "step": 2150 + }, + { + "epoch": 1.547203740334472, + "grad_norm": 7.6631992327283545, + "learning_rate": 2.832860820914161e-06, + "loss": 0.2146, + "step": 2151 + }, + { + "epoch": 1.547923035425283, + "grad_norm": 3.8504844262217985, + "learning_rate": 2.832704704975334e-06, + "loss": 0.1732, + "step": 2152 + }, + { + "epoch": 1.5486423305160941, + "grad_norm": 2.2987732676595836, + "learning_rate": 2.8325485204665104e-06, + "loss": 0.0248, + "step": 2153 + }, + { + "epoch": 1.5493616256069052, + "grad_norm": 3.1349656365080882, + "learning_rate": 2.8323922673957264e-06, + "loss": 0.0342, + "step": 2154 + }, + { + "epoch": 1.5500809206977162, + "grad_norm": 4.032634076190705, + "learning_rate": 2.832235945771021e-06, + "loss": 0.1517, + "step": 2155 + }, + { + "epoch": 1.5508002157885272, + "grad_norm": 2.9285883897014675, + "learning_rate": 2.832079555600437e-06, + "loss": 0.0608, + "step": 2156 + }, + { + "epoch": 1.5515195108793383, + "grad_norm": 4.342767791337391, + "learning_rate": 2.831923096892022e-06, + "loss": 0.0206, + "step": 2157 + }, + { + "epoch": 1.5522388059701493, + "grad_norm": 4.963240834884954, + "learning_rate": 2.831766569653826e-06, + "loss": 0.0904, + "step": 2158 + }, + { + "epoch": 1.5529581010609603, + "grad_norm": 5.293310723694916, + "learning_rate": 2.8316099738939016e-06, + "loss": 0.1143, + "step": 2159 + }, + { + "epoch": 1.5536773961517714, + "grad_norm": 2.4109550114761547, + "learning_rate": 2.8314533096203066e-06, + "loss": 0.0669, + "step": 2160 + }, + { + "epoch": 1.5543966912425824, + "grad_norm": 5.225078460235985, + "learning_rate": 2.8312965768411018e-06, + "loss": 0.2022, + "step": 2161 + }, + { + "epoch": 1.5551159863333934, + "grad_norm": 8.052550713583928, + "learning_rate": 2.831139775564351e-06, + "loss": 0.3126, + "step": 2162 + }, + { + "epoch": 1.5558352814242042, + "grad_norm": 4.311607885095645, + "learning_rate": 2.8309829057981213e-06, + "loss": 0.1268, + "step": 2163 + }, + { + "epoch": 1.5565545765150153, + "grad_norm": 5.597379773741944, + "learning_rate": 2.8308259675504854e-06, + "loss": 0.1181, + "step": 2164 + }, + { + "epoch": 1.5572738716058263, + "grad_norm": 3.469146886917397, + "learning_rate": 2.830668960829517e-06, + "loss": 0.1413, + "step": 2165 + }, + { + "epoch": 1.5579931666966373, + "grad_norm": 5.873459104704315, + "learning_rate": 2.830511885643295e-06, + "loss": 0.0917, + "step": 2166 + }, + { + "epoch": 1.5587124617874482, + "grad_norm": 4.351247191572507, + "learning_rate": 2.8303547419999005e-06, + "loss": 0.0438, + "step": 2167 + }, + { + "epoch": 1.5594317568782592, + "grad_norm": 4.579715624408742, + "learning_rate": 2.830197529907419e-06, + "loss": 0.0301, + "step": 2168 + }, + { + "epoch": 1.5601510519690702, + "grad_norm": 3.561731954684281, + "learning_rate": 2.83004024937394e-06, + "loss": 0.0966, + "step": 2169 + }, + { + "epoch": 1.5608703470598813, + "grad_norm": 6.0243790476827, + "learning_rate": 2.829882900407555e-06, + "loss": 0.1343, + "step": 2170 + }, + { + "epoch": 1.5615896421506923, + "grad_norm": 3.2786841480857594, + "learning_rate": 2.8297254830163604e-06, + "loss": 0.1283, + "step": 2171 + }, + { + "epoch": 1.5623089372415033, + "grad_norm": 9.864595709499882, + "learning_rate": 2.8295679972084555e-06, + "loss": 0.4016, + "step": 2172 + }, + { + "epoch": 1.5630282323323144, + "grad_norm": 4.863301507240836, + "learning_rate": 2.829410442991943e-06, + "loss": 0.2683, + "step": 2173 + }, + { + "epoch": 1.5637475274231254, + "grad_norm": 10.295005455750127, + "learning_rate": 2.829252820374929e-06, + "loss": 0.3946, + "step": 2174 + }, + { + "epoch": 1.5644668225139364, + "grad_norm": 4.392501607213936, + "learning_rate": 2.8290951293655245e-06, + "loss": 0.1145, + "step": 2175 + }, + { + "epoch": 1.5651861176047475, + "grad_norm": 3.773643525555693, + "learning_rate": 2.8289373699718425e-06, + "loss": 0.0928, + "step": 2176 + }, + { + "epoch": 1.5659054126955585, + "grad_norm": 3.0179265026340807, + "learning_rate": 2.828779542202e-06, + "loss": 0.0537, + "step": 2177 + }, + { + "epoch": 1.5666247077863693, + "grad_norm": 3.5954299563091676, + "learning_rate": 2.8286216460641173e-06, + "loss": 0.142, + "step": 2178 + }, + { + "epoch": 1.5673440028771803, + "grad_norm": 6.858052538285908, + "learning_rate": 2.8284636815663186e-06, + "loss": 0.2409, + "step": 2179 + }, + { + "epoch": 1.5680632979679914, + "grad_norm": 5.430650453224053, + "learning_rate": 2.828305648716731e-06, + "loss": 0.1071, + "step": 2180 + }, + { + "epoch": 1.5687825930588024, + "grad_norm": 3.248055123298909, + "learning_rate": 2.828147547523487e-06, + "loss": 0.079, + "step": 2181 + }, + { + "epoch": 1.5695018881496132, + "grad_norm": 2.0344750643084364, + "learning_rate": 2.8279893779947197e-06, + "loss": 0.0473, + "step": 2182 + }, + { + "epoch": 1.5702211832404243, + "grad_norm": 3.8886855686785635, + "learning_rate": 2.8278311401385675e-06, + "loss": 0.0261, + "step": 2183 + }, + { + "epoch": 1.5709404783312353, + "grad_norm": 4.521234663506471, + "learning_rate": 2.8276728339631722e-06, + "loss": 0.1006, + "step": 2184 + }, + { + "epoch": 1.5716597734220463, + "grad_norm": 5.85412494818436, + "learning_rate": 2.827514459476679e-06, + "loss": 0.1395, + "step": 2185 + }, + { + "epoch": 1.5723790685128574, + "grad_norm": 3.3530034789998604, + "learning_rate": 2.8273560166872367e-06, + "loss": 0.1225, + "step": 2186 + }, + { + "epoch": 1.5730983636036684, + "grad_norm": 5.210246229598176, + "learning_rate": 2.8271975056029968e-06, + "loss": 0.0473, + "step": 2187 + }, + { + "epoch": 1.5738176586944794, + "grad_norm": 5.026905690085372, + "learning_rate": 2.827038926232116e-06, + "loss": 0.054, + "step": 2188 + }, + { + "epoch": 1.5745369537852905, + "grad_norm": 3.5326371526764517, + "learning_rate": 2.8268802785827525e-06, + "loss": 0.0485, + "step": 2189 + }, + { + "epoch": 1.5752562488761015, + "grad_norm": 2.6430518956004003, + "learning_rate": 2.8267215626630697e-06, + "loss": 0.0354, + "step": 2190 + }, + { + "epoch": 1.5759755439669125, + "grad_norm": 4.005273130577424, + "learning_rate": 2.8265627784812333e-06, + "loss": 0.1501, + "step": 2191 + }, + { + "epoch": 1.5766948390577236, + "grad_norm": 5.806301907848369, + "learning_rate": 2.8264039260454135e-06, + "loss": 0.1681, + "step": 2192 + }, + { + "epoch": 1.5774141341485346, + "grad_norm": 5.590764946339853, + "learning_rate": 2.8262450053637827e-06, + "loss": 0.2601, + "step": 2193 + }, + { + "epoch": 1.5781334292393454, + "grad_norm": 4.194100201745286, + "learning_rate": 2.8260860164445183e-06, + "loss": 0.1805, + "step": 2194 + }, + { + "epoch": 1.5788527243301564, + "grad_norm": 2.791708114715783, + "learning_rate": 2.825926959295801e-06, + "loss": 0.0332, + "step": 2195 + }, + { + "epoch": 1.5795720194209675, + "grad_norm": 3.89971520548335, + "learning_rate": 2.8257678339258136e-06, + "loss": 0.0895, + "step": 2196 + }, + { + "epoch": 1.5802913145117785, + "grad_norm": 2.926517474868108, + "learning_rate": 2.8256086403427444e-06, + "loss": 0.0092, + "step": 2197 + }, + { + "epoch": 1.5810106096025893, + "grad_norm": 3.910635926420911, + "learning_rate": 2.825449378554783e-06, + "loss": 0.0686, + "step": 2198 + }, + { + "epoch": 1.5817299046934004, + "grad_norm": 5.32665345577694, + "learning_rate": 2.8252900485701245e-06, + "loss": 0.1869, + "step": 2199 + }, + { + "epoch": 1.5824491997842114, + "grad_norm": 5.924374069458105, + "learning_rate": 2.8251306503969665e-06, + "loss": 0.1249, + "step": 2200 + }, + { + "epoch": 1.5831684948750224, + "grad_norm": 4.296047295252244, + "learning_rate": 2.8249711840435105e-06, + "loss": 0.1586, + "step": 2201 + }, + { + "epoch": 1.5838877899658335, + "grad_norm": 3.996958827891937, + "learning_rate": 2.8248116495179613e-06, + "loss": 0.1272, + "step": 2202 + }, + { + "epoch": 1.5846070850566445, + "grad_norm": 4.970200546125578, + "learning_rate": 2.824652046828527e-06, + "loss": 0.1521, + "step": 2203 + }, + { + "epoch": 1.5853263801474555, + "grad_norm": 4.689615966087016, + "learning_rate": 2.8244923759834194e-06, + "loss": 0.0576, + "step": 2204 + }, + { + "epoch": 1.5860456752382666, + "grad_norm": 6.0204970378752, + "learning_rate": 2.824332636990854e-06, + "loss": 0.0783, + "step": 2205 + }, + { + "epoch": 1.5867649703290776, + "grad_norm": 8.514657860316047, + "learning_rate": 2.82417282985905e-06, + "loss": 0.2059, + "step": 2206 + }, + { + "epoch": 1.5874842654198886, + "grad_norm": 4.1644252119388385, + "learning_rate": 2.8240129545962284e-06, + "loss": 0.1088, + "step": 2207 + }, + { + "epoch": 1.5882035605106997, + "grad_norm": 3.6262486693879663, + "learning_rate": 2.823853011210617e-06, + "loss": 0.1078, + "step": 2208 + }, + { + "epoch": 1.5889228556015105, + "grad_norm": 2.959788960634419, + "learning_rate": 2.823692999710444e-06, + "loss": 0.0832, + "step": 2209 + }, + { + "epoch": 1.5896421506923215, + "grad_norm": 3.762232847064055, + "learning_rate": 2.8235329201039425e-06, + "loss": 0.1442, + "step": 2210 + }, + { + "epoch": 1.5903614457831325, + "grad_norm": 6.115826253694773, + "learning_rate": 2.8233727723993488e-06, + "loss": 0.0253, + "step": 2211 + }, + { + "epoch": 1.5910807408739436, + "grad_norm": 0.34769324865897927, + "learning_rate": 2.823212556604903e-06, + "loss": 0.0012, + "step": 2212 + }, + { + "epoch": 1.5918000359647544, + "grad_norm": 2.547732307064375, + "learning_rate": 2.823052272728848e-06, + "loss": 0.1076, + "step": 2213 + }, + { + "epoch": 1.5925193310555654, + "grad_norm": 3.0890038567971003, + "learning_rate": 2.822891920779431e-06, + "loss": 0.0732, + "step": 2214 + }, + { + "epoch": 1.5932386261463765, + "grad_norm": 7.256156531818067, + "learning_rate": 2.822731500764903e-06, + "loss": 0.3677, + "step": 2215 + }, + { + "epoch": 1.5939579212371875, + "grad_norm": 3.471913647728683, + "learning_rate": 2.822571012693517e-06, + "loss": 0.1305, + "step": 2216 + }, + { + "epoch": 1.5946772163279985, + "grad_norm": 6.163097763007754, + "learning_rate": 2.8224104565735303e-06, + "loss": 0.0803, + "step": 2217 + }, + { + "epoch": 1.5953965114188096, + "grad_norm": 3.3098696649244452, + "learning_rate": 2.8222498324132045e-06, + "loss": 0.0445, + "step": 2218 + }, + { + "epoch": 1.5961158065096206, + "grad_norm": 6.246364002162796, + "learning_rate": 2.8220891402208037e-06, + "loss": 0.1418, + "step": 2219 + }, + { + "epoch": 1.5968351016004316, + "grad_norm": 4.306632523253456, + "learning_rate": 2.821928380004596e-06, + "loss": 0.1117, + "step": 2220 + }, + { + "epoch": 1.5975543966912427, + "grad_norm": 3.4714969927413044, + "learning_rate": 2.8217675517728523e-06, + "loss": 0.0765, + "step": 2221 + }, + { + "epoch": 1.5982736917820537, + "grad_norm": 3.8205470990773076, + "learning_rate": 2.821606655533848e-06, + "loss": 0.1076, + "step": 2222 + }, + { + "epoch": 1.5989929868728647, + "grad_norm": 2.8841160104454735, + "learning_rate": 2.821445691295861e-06, + "loss": 0.1278, + "step": 2223 + }, + { + "epoch": 1.5997122819636758, + "grad_norm": 3.6395173762428796, + "learning_rate": 2.821284659067174e-06, + "loss": 0.1872, + "step": 2224 + }, + { + "epoch": 1.6004315770544866, + "grad_norm": 2.8901669681731494, + "learning_rate": 2.821123558856071e-06, + "loss": 0.0316, + "step": 2225 + }, + { + "epoch": 1.6011508721452976, + "grad_norm": 2.5862905311458317, + "learning_rate": 2.820962390670842e-06, + "loss": 0.1112, + "step": 2226 + }, + { + "epoch": 1.6018701672361086, + "grad_norm": 5.803221851571076, + "learning_rate": 2.820801154519779e-06, + "loss": 0.264, + "step": 2227 + }, + { + "epoch": 1.6025894623269195, + "grad_norm": 3.11843554181106, + "learning_rate": 2.8206398504111787e-06, + "loss": 0.039, + "step": 2228 + }, + { + "epoch": 1.6033087574177305, + "grad_norm": 6.86113282444148, + "learning_rate": 2.820478478353339e-06, + "loss": 0.2357, + "step": 2229 + }, + { + "epoch": 1.6040280525085415, + "grad_norm": 6.941501109239274, + "learning_rate": 2.8203170383545643e-06, + "loss": 0.179, + "step": 2230 + }, + { + "epoch": 1.6047473475993526, + "grad_norm": 2.240716476669486, + "learning_rate": 2.82015553042316e-06, + "loss": 0.0272, + "step": 2231 + }, + { + "epoch": 1.6054666426901636, + "grad_norm": 3.111928094951867, + "learning_rate": 2.819993954567436e-06, + "loss": 0.1592, + "step": 2232 + }, + { + "epoch": 1.6061859377809746, + "grad_norm": 3.0023741082243434, + "learning_rate": 2.8198323107957055e-06, + "loss": 0.0648, + "step": 2233 + }, + { + "epoch": 1.6069052328717857, + "grad_norm": 5.8812597294182005, + "learning_rate": 2.8196705991162862e-06, + "loss": 0.0343, + "step": 2234 + }, + { + "epoch": 1.6076245279625967, + "grad_norm": 2.7439908128283856, + "learning_rate": 2.819508819537498e-06, + "loss": 0.0889, + "step": 2235 + }, + { + "epoch": 1.6083438230534077, + "grad_norm": 2.923020489316437, + "learning_rate": 2.8193469720676648e-06, + "loss": 0.0244, + "step": 2236 + }, + { + "epoch": 1.6090631181442188, + "grad_norm": 3.3689334171131167, + "learning_rate": 2.8191850567151135e-06, + "loss": 0.1249, + "step": 2237 + }, + { + "epoch": 1.6097824132350298, + "grad_norm": 4.561428389612156, + "learning_rate": 2.8190230734881753e-06, + "loss": 0.0173, + "step": 2238 + }, + { + "epoch": 1.6105017083258408, + "grad_norm": 5.528875668548176, + "learning_rate": 2.818861022395185e-06, + "loss": 0.1071, + "step": 2239 + }, + { + "epoch": 1.6112210034166516, + "grad_norm": 5.414652000744565, + "learning_rate": 2.8186989034444794e-06, + "loss": 0.2672, + "step": 2240 + }, + { + "epoch": 1.6119402985074627, + "grad_norm": 0.1915798511413906, + "learning_rate": 2.8185367166444e-06, + "loss": 0.0006, + "step": 2241 + }, + { + "epoch": 1.6126595935982737, + "grad_norm": 1.7896864197806093, + "learning_rate": 2.8183744620032927e-06, + "loss": 0.0489, + "step": 2242 + }, + { + "epoch": 1.6133788886890847, + "grad_norm": 5.360176939414118, + "learning_rate": 2.818212139529505e-06, + "loss": 0.2154, + "step": 2243 + }, + { + "epoch": 1.6140981837798956, + "grad_norm": 3.614753660243484, + "learning_rate": 2.818049749231388e-06, + "loss": 0.0255, + "step": 2244 + }, + { + "epoch": 1.6148174788707066, + "grad_norm": 0.8295565308390117, + "learning_rate": 2.817887291117298e-06, + "loss": 0.0016, + "step": 2245 + }, + { + "epoch": 1.6155367739615176, + "grad_norm": 5.539456900748424, + "learning_rate": 2.8177247651955935e-06, + "loss": 0.1398, + "step": 2246 + }, + { + "epoch": 1.6162560690523287, + "grad_norm": 4.704211833566174, + "learning_rate": 2.817562171474636e-06, + "loss": 0.0343, + "step": 2247 + }, + { + "epoch": 1.6169753641431397, + "grad_norm": 5.250055237749541, + "learning_rate": 2.817399509962793e-06, + "loss": 0.2029, + "step": 2248 + }, + { + "epoch": 1.6176946592339507, + "grad_norm": 3.6006835441705145, + "learning_rate": 2.8172367806684317e-06, + "loss": 0.0214, + "step": 2249 + }, + { + "epoch": 1.6184139543247618, + "grad_norm": 2.011895212746637, + "learning_rate": 2.8170739835999258e-06, + "loss": 0.0094, + "step": 2250 + }, + { + "epoch": 1.6191332494155728, + "grad_norm": 2.8655939483148813, + "learning_rate": 2.8169111187656517e-06, + "loss": 0.0773, + "step": 2251 + }, + { + "epoch": 1.6198525445063838, + "grad_norm": 11.696558726027765, + "learning_rate": 2.8167481861739884e-06, + "loss": 0.131, + "step": 2252 + }, + { + "epoch": 1.6205718395971949, + "grad_norm": 1.549969438440941, + "learning_rate": 2.81658518583332e-06, + "loss": 0.0493, + "step": 2253 + }, + { + "epoch": 1.621291134688006, + "grad_norm": 5.285909841275818, + "learning_rate": 2.816422117752032e-06, + "loss": 0.0803, + "step": 2254 + }, + { + "epoch": 1.6220104297788167, + "grad_norm": 2.9645018350910806, + "learning_rate": 2.8162589819385157e-06, + "loss": 0.057, + "step": 2255 + }, + { + "epoch": 1.6227297248696277, + "grad_norm": 5.523642795654883, + "learning_rate": 2.816095778401164e-06, + "loss": 0.1699, + "step": 2256 + }, + { + "epoch": 1.6234490199604388, + "grad_norm": 4.265287292529384, + "learning_rate": 2.815932507148374e-06, + "loss": 0.1583, + "step": 2257 + }, + { + "epoch": 1.6241683150512498, + "grad_norm": 6.164838547940256, + "learning_rate": 2.8157691681885463e-06, + "loss": 0.1556, + "step": 2258 + }, + { + "epoch": 1.6248876101420606, + "grad_norm": 6.47143839920109, + "learning_rate": 2.8156057615300853e-06, + "loss": 0.2227, + "step": 2259 + }, + { + "epoch": 1.6256069052328717, + "grad_norm": 4.936288979930599, + "learning_rate": 2.815442287181399e-06, + "loss": 0.1467, + "step": 2260 + }, + { + "epoch": 1.6263262003236827, + "grad_norm": 2.452710234812695, + "learning_rate": 2.815278745150897e-06, + "loss": 0.0479, + "step": 2261 + }, + { + "epoch": 1.6270454954144937, + "grad_norm": 5.164236589206694, + "learning_rate": 2.815115135446995e-06, + "loss": 0.0265, + "step": 2262 + }, + { + "epoch": 1.6277647905053048, + "grad_norm": 2.7093230626051636, + "learning_rate": 2.8149514580781105e-06, + "loss": 0.0158, + "step": 2263 + }, + { + "epoch": 1.6284840855961158, + "grad_norm": 4.363892670433591, + "learning_rate": 2.814787713052665e-06, + "loss": 0.1273, + "step": 2264 + }, + { + "epoch": 1.6292033806869268, + "grad_norm": 1.3594032623016834, + "learning_rate": 2.814623900379084e-06, + "loss": 0.0279, + "step": 2265 + }, + { + "epoch": 1.6299226757777379, + "grad_norm": 2.884470539701012, + "learning_rate": 2.814460020065795e-06, + "loss": 0.0159, + "step": 2266 + }, + { + "epoch": 1.630641970868549, + "grad_norm": 4.018579597539334, + "learning_rate": 2.8142960721212316e-06, + "loss": 0.0776, + "step": 2267 + }, + { + "epoch": 1.63136126595936, + "grad_norm": 1.9747994372467361, + "learning_rate": 2.8141320565538275e-06, + "loss": 0.0089, + "step": 2268 + }, + { + "epoch": 1.632080561050171, + "grad_norm": 6.1296844227917715, + "learning_rate": 2.813967973372022e-06, + "loss": 0.0764, + "step": 2269 + }, + { + "epoch": 1.632799856140982, + "grad_norm": 4.987269399869254, + "learning_rate": 2.8138038225842577e-06, + "loss": 0.1961, + "step": 2270 + }, + { + "epoch": 1.6335191512317928, + "grad_norm": 2.2096645390981897, + "learning_rate": 2.813639604198981e-06, + "loss": 0.062, + "step": 2271 + }, + { + "epoch": 1.6342384463226038, + "grad_norm": 4.4752041455864, + "learning_rate": 2.8134753182246398e-06, + "loss": 0.1245, + "step": 2272 + }, + { + "epoch": 1.6349577414134149, + "grad_norm": 5.2397614527439, + "learning_rate": 2.813310964669688e-06, + "loss": 0.1182, + "step": 2273 + }, + { + "epoch": 1.635677036504226, + "grad_norm": 3.070377812357022, + "learning_rate": 2.813146543542582e-06, + "loss": 0.072, + "step": 2274 + }, + { + "epoch": 1.6363963315950367, + "grad_norm": 7.470500337100392, + "learning_rate": 2.8129820548517813e-06, + "loss": 0.1969, + "step": 2275 + }, + { + "epoch": 1.6371156266858478, + "grad_norm": 5.249217968348018, + "learning_rate": 2.8128174986057486e-06, + "loss": 0.1027, + "step": 2276 + }, + { + "epoch": 1.6378349217766588, + "grad_norm": 5.7018996050692605, + "learning_rate": 2.8126528748129506e-06, + "loss": 0.0388, + "step": 2277 + }, + { + "epoch": 1.6385542168674698, + "grad_norm": 1.8024741017662713, + "learning_rate": 2.8124881834818586e-06, + "loss": 0.0147, + "step": 2278 + }, + { + "epoch": 1.6392735119582809, + "grad_norm": 4.54133779935897, + "learning_rate": 2.812323424620946e-06, + "loss": 0.1609, + "step": 2279 + }, + { + "epoch": 1.639992807049092, + "grad_norm": 3.45566076563589, + "learning_rate": 2.8121585982386883e-06, + "loss": 0.0885, + "step": 2280 + }, + { + "epoch": 1.640712102139903, + "grad_norm": 5.113097389723223, + "learning_rate": 2.811993704343568e-06, + "loss": 0.0934, + "step": 2281 + }, + { + "epoch": 1.641431397230714, + "grad_norm": 2.083635302303951, + "learning_rate": 2.8118287429440684e-06, + "loss": 0.0355, + "step": 2282 + }, + { + "epoch": 1.642150692321525, + "grad_norm": 7.046809877712022, + "learning_rate": 2.811663714048677e-06, + "loss": 0.2638, + "step": 2283 + }, + { + "epoch": 1.642869987412336, + "grad_norm": 4.662899348259462, + "learning_rate": 2.811498617665885e-06, + "loss": 0.0872, + "step": 2284 + }, + { + "epoch": 1.643589282503147, + "grad_norm": 6.789058395980352, + "learning_rate": 2.811333453804187e-06, + "loss": 0.1588, + "step": 2285 + }, + { + "epoch": 1.6443085775939579, + "grad_norm": 5.349274469126305, + "learning_rate": 2.8111682224720807e-06, + "loss": 0.1744, + "step": 2286 + }, + { + "epoch": 1.645027872684769, + "grad_norm": 1.6239266166418995, + "learning_rate": 2.8110029236780672e-06, + "loss": 0.0235, + "step": 2287 + }, + { + "epoch": 1.64574716777558, + "grad_norm": 4.6050908785026285, + "learning_rate": 2.8108375574306523e-06, + "loss": 0.0719, + "step": 2288 + }, + { + "epoch": 1.646466462866391, + "grad_norm": 2.215111682739865, + "learning_rate": 2.8106721237383445e-06, + "loss": 0.0406, + "step": 2289 + }, + { + "epoch": 1.6471857579572018, + "grad_norm": 3.0358773674490767, + "learning_rate": 2.810506622609654e-06, + "loss": 0.117, + "step": 2290 + }, + { + "epoch": 1.6479050530480128, + "grad_norm": 1.409113504727841, + "learning_rate": 2.8103410540530984e-06, + "loss": 0.0159, + "step": 2291 + }, + { + "epoch": 1.6486243481388239, + "grad_norm": 3.046594539798307, + "learning_rate": 2.810175418077195e-06, + "loss": 0.0257, + "step": 2292 + }, + { + "epoch": 1.649343643229635, + "grad_norm": 5.131974119578082, + "learning_rate": 2.810009714690466e-06, + "loss": 0.2435, + "step": 2293 + }, + { + "epoch": 1.650062938320446, + "grad_norm": 6.4958482363193895, + "learning_rate": 2.809843943901438e-06, + "loss": 0.2056, + "step": 2294 + }, + { + "epoch": 1.650782233411257, + "grad_norm": 3.564856109982279, + "learning_rate": 2.8096781057186393e-06, + "loss": 0.1636, + "step": 2295 + }, + { + "epoch": 1.651501528502068, + "grad_norm": 0.8193251687915041, + "learning_rate": 2.809512200150603e-06, + "loss": 0.0083, + "step": 2296 + }, + { + "epoch": 1.652220823592879, + "grad_norm": 1.795011701523758, + "learning_rate": 2.8093462272058654e-06, + "loss": 0.0293, + "step": 2297 + }, + { + "epoch": 1.65294011868369, + "grad_norm": 3.3611128272617856, + "learning_rate": 2.809180186892966e-06, + "loss": 0.087, + "step": 2298 + }, + { + "epoch": 1.653659413774501, + "grad_norm": 3.2815134778480903, + "learning_rate": 2.809014079220448e-06, + "loss": 0.1097, + "step": 2299 + }, + { + "epoch": 1.6543787088653121, + "grad_norm": 4.351753010329881, + "learning_rate": 2.808847904196857e-06, + "loss": 0.0863, + "step": 2300 + }, + { + "epoch": 1.6550980039561232, + "grad_norm": 4.096582758129234, + "learning_rate": 2.8086816618307447e-06, + "loss": 0.2683, + "step": 2301 + }, + { + "epoch": 1.655817299046934, + "grad_norm": 6.01764618605431, + "learning_rate": 2.808515352130663e-06, + "loss": 0.1444, + "step": 2302 + }, + { + "epoch": 1.656536594137745, + "grad_norm": 3.543681806265652, + "learning_rate": 2.808348975105169e-06, + "loss": 0.056, + "step": 2303 + }, + { + "epoch": 1.657255889228556, + "grad_norm": 6.168429797203481, + "learning_rate": 2.8081825307628243e-06, + "loss": 0.1253, + "step": 2304 + }, + { + "epoch": 1.6579751843193669, + "grad_norm": 2.1512686325368375, + "learning_rate": 2.8080160191121915e-06, + "loss": 0.0458, + "step": 2305 + }, + { + "epoch": 1.658694479410178, + "grad_norm": 4.939568498325234, + "learning_rate": 2.8078494401618386e-06, + "loss": 0.0663, + "step": 2306 + }, + { + "epoch": 1.659413774500989, + "grad_norm": 4.034034535770205, + "learning_rate": 2.807682793920336e-06, + "loss": 0.1563, + "step": 2307 + }, + { + "epoch": 1.6601330695918, + "grad_norm": 2.9118870979833247, + "learning_rate": 2.807516080396258e-06, + "loss": 0.012, + "step": 2308 + }, + { + "epoch": 1.660852364682611, + "grad_norm": 0.5425354631322455, + "learning_rate": 2.8073492995981823e-06, + "loss": 0.0008, + "step": 2309 + }, + { + "epoch": 1.661571659773422, + "grad_norm": 7.007609927392306, + "learning_rate": 2.8071824515346904e-06, + "loss": 0.2073, + "step": 2310 + }, + { + "epoch": 1.662290954864233, + "grad_norm": 4.882534919897073, + "learning_rate": 2.807015536214367e-06, + "loss": 0.1861, + "step": 2311 + }, + { + "epoch": 1.663010249955044, + "grad_norm": 5.001806483729345, + "learning_rate": 2.8068485536457993e-06, + "loss": 0.1051, + "step": 2312 + }, + { + "epoch": 1.6637295450458551, + "grad_norm": 4.005848121328785, + "learning_rate": 2.8066815038375797e-06, + "loss": 0.1536, + "step": 2313 + }, + { + "epoch": 1.6644488401366662, + "grad_norm": 6.363799871490677, + "learning_rate": 2.8065143867983027e-06, + "loss": 0.1323, + "step": 2314 + }, + { + "epoch": 1.6651681352274772, + "grad_norm": 4.58530015470701, + "learning_rate": 2.806347202536567e-06, + "loss": 0.0796, + "step": 2315 + }, + { + "epoch": 1.6658874303182882, + "grad_norm": 2.350720584461565, + "learning_rate": 2.8061799510609747e-06, + "loss": 0.043, + "step": 2316 + }, + { + "epoch": 1.666606725409099, + "grad_norm": 6.480705472133568, + "learning_rate": 2.806012632380131e-06, + "loss": 0.1831, + "step": 2317 + }, + { + "epoch": 1.66732602049991, + "grad_norm": 2.2798640447567857, + "learning_rate": 2.805845246502645e-06, + "loss": 0.0802, + "step": 2318 + }, + { + "epoch": 1.6680453155907211, + "grad_norm": 5.085479480582992, + "learning_rate": 2.805677793437128e-06, + "loss": 0.1542, + "step": 2319 + }, + { + "epoch": 1.6687646106815321, + "grad_norm": 3.018508368209533, + "learning_rate": 2.8055102731921972e-06, + "loss": 0.0969, + "step": 2320 + }, + { + "epoch": 1.669483905772343, + "grad_norm": 0.34720257861225523, + "learning_rate": 2.8053426857764702e-06, + "loss": 0.0012, + "step": 2321 + }, + { + "epoch": 1.670203200863154, + "grad_norm": 3.736993532295286, + "learning_rate": 2.8051750311985716e-06, + "loss": 0.1521, + "step": 2322 + }, + { + "epoch": 1.670922495953965, + "grad_norm": 6.2985218066676225, + "learning_rate": 2.805007309467126e-06, + "loss": 0.1019, + "step": 2323 + }, + { + "epoch": 1.671641791044776, + "grad_norm": 2.9203803049116113, + "learning_rate": 2.804839520590763e-06, + "loss": 0.0615, + "step": 2324 + }, + { + "epoch": 1.672361086135587, + "grad_norm": 6.425577303323589, + "learning_rate": 2.8046716645781166e-06, + "loss": 0.187, + "step": 2325 + }, + { + "epoch": 1.6730803812263981, + "grad_norm": 1.697400462742743, + "learning_rate": 2.804503741437823e-06, + "loss": 0.0481, + "step": 2326 + }, + { + "epoch": 1.6737996763172092, + "grad_norm": 2.187582140444186, + "learning_rate": 2.8043357511785217e-06, + "loss": 0.0138, + "step": 2327 + }, + { + "epoch": 1.6745189714080202, + "grad_norm": 5.07818170563932, + "learning_rate": 2.804167693808856e-06, + "loss": 0.1954, + "step": 2328 + }, + { + "epoch": 1.6752382664988312, + "grad_norm": 5.310186154045098, + "learning_rate": 2.8039995693374734e-06, + "loss": 0.1073, + "step": 2329 + }, + { + "epoch": 1.6759575615896423, + "grad_norm": 3.3889899030592825, + "learning_rate": 2.803831377773024e-06, + "loss": 0.1176, + "step": 2330 + }, + { + "epoch": 1.6766768566804533, + "grad_norm": 2.49210587642558, + "learning_rate": 2.803663119124161e-06, + "loss": 0.0678, + "step": 2331 + }, + { + "epoch": 1.677396151771264, + "grad_norm": 4.344678116227538, + "learning_rate": 2.8034947933995425e-06, + "loss": 0.1543, + "step": 2332 + }, + { + "epoch": 1.6781154468620751, + "grad_norm": 2.3696161097073363, + "learning_rate": 2.8033264006078283e-06, + "loss": 0.0483, + "step": 2333 + }, + { + "epoch": 1.6788347419528862, + "grad_norm": 3.5993557221384536, + "learning_rate": 2.803157940757683e-06, + "loss": 0.0836, + "step": 2334 + }, + { + "epoch": 1.6795540370436972, + "grad_norm": 7.841194297845523, + "learning_rate": 2.8029894138577735e-06, + "loss": 0.0462, + "step": 2335 + }, + { + "epoch": 1.680273332134508, + "grad_norm": 4.784775815494183, + "learning_rate": 2.802820819916772e-06, + "loss": 0.0759, + "step": 2336 + }, + { + "epoch": 1.680992627225319, + "grad_norm": 4.729570994216389, + "learning_rate": 2.802652158943352e-06, + "loss": 0.1444, + "step": 2337 + }, + { + "epoch": 1.68171192231613, + "grad_norm": 4.215344439412829, + "learning_rate": 2.802483430946192e-06, + "loss": 0.0302, + "step": 2338 + }, + { + "epoch": 1.6824312174069411, + "grad_norm": 6.690240133242299, + "learning_rate": 2.8023146359339725e-06, + "loss": 0.1967, + "step": 2339 + }, + { + "epoch": 1.6831505124977522, + "grad_norm": 3.6220979254508556, + "learning_rate": 2.8021457739153793e-06, + "loss": 0.1321, + "step": 2340 + }, + { + "epoch": 1.6838698075885632, + "grad_norm": 5.424669130094212, + "learning_rate": 2.8019768448991e-06, + "loss": 0.0929, + "step": 2341 + }, + { + "epoch": 1.6845891026793742, + "grad_norm": 1.7144708037615792, + "learning_rate": 2.801807848893827e-06, + "loss": 0.0096, + "step": 2342 + }, + { + "epoch": 1.6853083977701853, + "grad_norm": 4.06530826171108, + "learning_rate": 2.801638785908254e-06, + "loss": 0.0444, + "step": 2343 + }, + { + "epoch": 1.6860276928609963, + "grad_norm": 4.721120526321798, + "learning_rate": 2.801469655951081e-06, + "loss": 0.0998, + "step": 2344 + }, + { + "epoch": 1.6867469879518073, + "grad_norm": 5.8860746771055386, + "learning_rate": 2.8013004590310097e-06, + "loss": 0.078, + "step": 2345 + }, + { + "epoch": 1.6874662830426184, + "grad_norm": 6.064701007299643, + "learning_rate": 2.801131195156745e-06, + "loss": 0.2046, + "step": 2346 + }, + { + "epoch": 1.6881855781334294, + "grad_norm": 6.173058455453307, + "learning_rate": 2.8009618643369964e-06, + "loss": 0.3731, + "step": 2347 + }, + { + "epoch": 1.6889048732242402, + "grad_norm": 1.9379741238649226, + "learning_rate": 2.8007924665804767e-06, + "loss": 0.0068, + "step": 2348 + }, + { + "epoch": 1.6896241683150512, + "grad_norm": 3.7071785250223943, + "learning_rate": 2.8006230018959e-06, + "loss": 0.1862, + "step": 2349 + }, + { + "epoch": 1.6903434634058623, + "grad_norm": 1.7909371751100964, + "learning_rate": 2.8004534702919876e-06, + "loss": 0.0073, + "step": 2350 + }, + { + "epoch": 1.6910627584966733, + "grad_norm": 2.372855211559006, + "learning_rate": 2.8002838717774608e-06, + "loss": 0.0636, + "step": 2351 + }, + { + "epoch": 1.6917820535874841, + "grad_norm": 7.358328824425426, + "learning_rate": 2.8001142063610463e-06, + "loss": 0.4972, + "step": 2352 + }, + { + "epoch": 1.6925013486782952, + "grad_norm": 6.74744800508788, + "learning_rate": 2.7999444740514736e-06, + "loss": 0.2217, + "step": 2353 + }, + { + "epoch": 1.6932206437691062, + "grad_norm": 7.429661107300888, + "learning_rate": 2.799774674857476e-06, + "loss": 0.3916, + "step": 2354 + }, + { + "epoch": 1.6939399388599172, + "grad_norm": 2.3350017624151396, + "learning_rate": 2.7996048087877893e-06, + "loss": 0.0829, + "step": 2355 + }, + { + "epoch": 1.6946592339507283, + "grad_norm": 3.3803922707648595, + "learning_rate": 2.799434875851154e-06, + "loss": 0.1471, + "step": 2356 + }, + { + "epoch": 1.6953785290415393, + "grad_norm": 3.7978222134421586, + "learning_rate": 2.7992648760563134e-06, + "loss": 0.053, + "step": 2357 + }, + { + "epoch": 1.6960978241323503, + "grad_norm": 4.815996112642946, + "learning_rate": 2.7990948094120143e-06, + "loss": 0.1971, + "step": 2358 + }, + { + "epoch": 1.6968171192231614, + "grad_norm": 2.439127198144136, + "learning_rate": 2.7989246759270062e-06, + "loss": 0.0577, + "step": 2359 + }, + { + "epoch": 1.6975364143139724, + "grad_norm": 2.8930296653214094, + "learning_rate": 2.7987544756100436e-06, + "loss": 0.1225, + "step": 2360 + }, + { + "epoch": 1.6982557094047834, + "grad_norm": 5.017556773666212, + "learning_rate": 2.7985842084698836e-06, + "loss": 0.0505, + "step": 2361 + }, + { + "epoch": 1.6989750044955945, + "grad_norm": 4.7436929412514885, + "learning_rate": 2.7984138745152865e-06, + "loss": 0.0374, + "step": 2362 + }, + { + "epoch": 1.6996942995864053, + "grad_norm": 4.041335827226105, + "learning_rate": 2.7982434737550157e-06, + "loss": 0.0388, + "step": 2363 + }, + { + "epoch": 1.7004135946772163, + "grad_norm": 2.9873480818832894, + "learning_rate": 2.7980730061978394e-06, + "loss": 0.143, + "step": 2364 + }, + { + "epoch": 1.7011328897680273, + "grad_norm": 1.9307763350469296, + "learning_rate": 2.797902471852528e-06, + "loss": 0.0321, + "step": 2365 + }, + { + "epoch": 1.7018521848588384, + "grad_norm": 6.812876028175807, + "learning_rate": 2.7977318707278564e-06, + "loss": 0.1261, + "step": 2366 + }, + { + "epoch": 1.7025714799496492, + "grad_norm": 5.680763272515375, + "learning_rate": 2.7975612028326022e-06, + "loss": 0.1719, + "step": 2367 + }, + { + "epoch": 1.7032907750404602, + "grad_norm": 3.264300707776835, + "learning_rate": 2.797390468175546e-06, + "loss": 0.111, + "step": 2368 + }, + { + "epoch": 1.7040100701312713, + "grad_norm": 1.8210095092883605, + "learning_rate": 2.7972196667654726e-06, + "loss": 0.0282, + "step": 2369 + }, + { + "epoch": 1.7047293652220823, + "grad_norm": 4.463694700132819, + "learning_rate": 2.7970487986111705e-06, + "loss": 0.0541, + "step": 2370 + }, + { + "epoch": 1.7054486603128933, + "grad_norm": 1.8725822761990196, + "learning_rate": 2.7968778637214304e-06, + "loss": 0.0048, + "step": 2371 + }, + { + "epoch": 1.7061679554037044, + "grad_norm": 1.9843340567547916, + "learning_rate": 2.796706862105048e-06, + "loss": 0.0092, + "step": 2372 + }, + { + "epoch": 1.7068872504945154, + "grad_norm": 3.875985553398004, + "learning_rate": 2.7965357937708207e-06, + "loss": 0.1433, + "step": 2373 + }, + { + "epoch": 1.7076065455853264, + "grad_norm": 3.7845446624688823, + "learning_rate": 2.7963646587275513e-06, + "loss": 0.1838, + "step": 2374 + }, + { + "epoch": 1.7083258406761375, + "grad_norm": 2.8864845486353286, + "learning_rate": 2.7961934569840443e-06, + "loss": 0.0611, + "step": 2375 + }, + { + "epoch": 1.7090451357669485, + "grad_norm": 3.0131955885034047, + "learning_rate": 2.7960221885491086e-06, + "loss": 0.0423, + "step": 2376 + }, + { + "epoch": 1.7097644308577595, + "grad_norm": 4.7533239840069195, + "learning_rate": 2.795850853431556e-06, + "loss": 0.1347, + "step": 2377 + }, + { + "epoch": 1.7104837259485706, + "grad_norm": 4.092464808013002, + "learning_rate": 2.795679451640203e-06, + "loss": 0.2292, + "step": 2378 + }, + { + "epoch": 1.7112030210393814, + "grad_norm": 1.995013038112466, + "learning_rate": 2.7955079831838667e-06, + "loss": 0.0116, + "step": 2379 + }, + { + "epoch": 1.7119223161301924, + "grad_norm": 6.157404191331068, + "learning_rate": 2.795336448071371e-06, + "loss": 0.1794, + "step": 2380 + }, + { + "epoch": 1.7126416112210034, + "grad_norm": 5.011503741469248, + "learning_rate": 2.795164846311541e-06, + "loss": 0.0262, + "step": 2381 + }, + { + "epoch": 1.7133609063118143, + "grad_norm": 3.534925477162323, + "learning_rate": 2.7949931779132066e-06, + "loss": 0.1281, + "step": 2382 + }, + { + "epoch": 1.7140802014026253, + "grad_norm": 7.017899273204855, + "learning_rate": 2.794821442885199e-06, + "loss": 0.334, + "step": 2383 + }, + { + "epoch": 1.7147994964934363, + "grad_norm": 6.3196514031515285, + "learning_rate": 2.7946496412363555e-06, + "loss": 0.0319, + "step": 2384 + }, + { + "epoch": 1.7155187915842474, + "grad_norm": 1.8931560590820344, + "learning_rate": 2.794477772975516e-06, + "loss": 0.0558, + "step": 2385 + }, + { + "epoch": 1.7162380866750584, + "grad_norm": 4.286760853514821, + "learning_rate": 2.7943058381115215e-06, + "loss": 0.109, + "step": 2386 + }, + { + "epoch": 1.7169573817658694, + "grad_norm": 3.636483978192362, + "learning_rate": 2.7941338366532204e-06, + "loss": 0.0916, + "step": 2387 + }, + { + "epoch": 1.7176766768566805, + "grad_norm": 5.498435541224947, + "learning_rate": 2.7939617686094615e-06, + "loss": 0.0458, + "step": 2388 + }, + { + "epoch": 1.7183959719474915, + "grad_norm": 2.417437411202421, + "learning_rate": 2.793789633989098e-06, + "loss": 0.0068, + "step": 2389 + }, + { + "epoch": 1.7191152670383025, + "grad_norm": 5.543263452435766, + "learning_rate": 2.793617432800986e-06, + "loss": 0.1707, + "step": 2390 + }, + { + "epoch": 1.7198345621291136, + "grad_norm": 4.396643626007494, + "learning_rate": 2.7934451650539874e-06, + "loss": 0.0112, + "step": 2391 + }, + { + "epoch": 1.7205538572199246, + "grad_norm": 4.964203598794387, + "learning_rate": 2.793272830756964e-06, + "loss": 0.1877, + "step": 2392 + }, + { + "epoch": 1.7212731523107356, + "grad_norm": 5.04100974186584, + "learning_rate": 2.7931004299187827e-06, + "loss": 0.0753, + "step": 2393 + }, + { + "epoch": 1.7219924474015464, + "grad_norm": 7.147974844653717, + "learning_rate": 2.792927962548315e-06, + "loss": 0.3401, + "step": 2394 + }, + { + "epoch": 1.7227117424923575, + "grad_norm": 5.050882737588076, + "learning_rate": 2.792755428654434e-06, + "loss": 0.1955, + "step": 2395 + }, + { + "epoch": 1.7234310375831685, + "grad_norm": 3.722476540875743, + "learning_rate": 2.792582828246016e-06, + "loss": 0.0071, + "step": 2396 + }, + { + "epoch": 1.7241503326739795, + "grad_norm": 2.978173772916557, + "learning_rate": 2.792410161331943e-06, + "loss": 0.0212, + "step": 2397 + }, + { + "epoch": 1.7248696277647904, + "grad_norm": 4.731260403746607, + "learning_rate": 2.792237427921099e-06, + "loss": 0.1153, + "step": 2398 + }, + { + "epoch": 1.7255889228556014, + "grad_norm": 7.267363709414449, + "learning_rate": 2.79206462802237e-06, + "loss": 0.2066, + "step": 2399 + }, + { + "epoch": 1.7263082179464124, + "grad_norm": 4.306413977625975, + "learning_rate": 2.791891761644648e-06, + "loss": 0.1158, + "step": 2400 + }, + { + "epoch": 1.7270275130372235, + "grad_norm": 2.7527140148638307, + "learning_rate": 2.7917188287968268e-06, + "loss": 0.0775, + "step": 2401 + }, + { + "epoch": 1.7277468081280345, + "grad_norm": 5.892912500185107, + "learning_rate": 2.791545829487805e-06, + "loss": 0.16, + "step": 2402 + }, + { + "epoch": 1.7284661032188455, + "grad_norm": 8.155263083235745, + "learning_rate": 2.7913727637264824e-06, + "loss": 0.1022, + "step": 2403 + }, + { + "epoch": 1.7291853983096566, + "grad_norm": 2.999315306442573, + "learning_rate": 2.7911996315217643e-06, + "loss": 0.0848, + "step": 2404 + }, + { + "epoch": 1.7299046934004676, + "grad_norm": 3.5327156826921424, + "learning_rate": 2.7910264328825586e-06, + "loss": 0.1383, + "step": 2405 + }, + { + "epoch": 1.7306239884912786, + "grad_norm": 5.699458381892293, + "learning_rate": 2.7908531678177767e-06, + "loss": 0.0878, + "step": 2406 + }, + { + "epoch": 1.7313432835820897, + "grad_norm": 1.7807204861908321, + "learning_rate": 2.790679836336333e-06, + "loss": 0.006, + "step": 2407 + }, + { + "epoch": 1.7320625786729007, + "grad_norm": 2.197058617885282, + "learning_rate": 2.7905064384471462e-06, + "loss": 0.0478, + "step": 2408 + }, + { + "epoch": 1.7327818737637115, + "grad_norm": 3.579967652177571, + "learning_rate": 2.7903329741591378e-06, + "loss": 0.0559, + "step": 2409 + }, + { + "epoch": 1.7335011688545225, + "grad_norm": 2.665072432924307, + "learning_rate": 2.7901594434812324e-06, + "loss": 0.0534, + "step": 2410 + }, + { + "epoch": 1.7342204639453336, + "grad_norm": 1.8579175465103006, + "learning_rate": 2.7899858464223595e-06, + "loss": 0.0495, + "step": 2411 + }, + { + "epoch": 1.7349397590361446, + "grad_norm": 4.540021825871725, + "learning_rate": 2.78981218299145e-06, + "loss": 0.0163, + "step": 2412 + }, + { + "epoch": 1.7356590541269554, + "grad_norm": 2.9008181430926667, + "learning_rate": 2.7896384531974396e-06, + "loss": 0.055, + "step": 2413 + }, + { + "epoch": 1.7363783492177665, + "grad_norm": 4.175275968242346, + "learning_rate": 2.7894646570492666e-06, + "loss": 0.1024, + "step": 2414 + }, + { + "epoch": 1.7370976443085775, + "grad_norm": 6.8995704900251065, + "learning_rate": 2.7892907945558734e-06, + "loss": 0.1833, + "step": 2415 + }, + { + "epoch": 1.7378169393993885, + "grad_norm": 2.6365483682168978, + "learning_rate": 2.7891168657262056e-06, + "loss": 0.0117, + "step": 2416 + }, + { + "epoch": 1.7385362344901996, + "grad_norm": 4.811859968335681, + "learning_rate": 2.7889428705692125e-06, + "loss": 0.1356, + "step": 2417 + }, + { + "epoch": 1.7392555295810106, + "grad_norm": 8.103032461889205, + "learning_rate": 2.7887688090938455e-06, + "loss": 0.2083, + "step": 2418 + }, + { + "epoch": 1.7399748246718216, + "grad_norm": 5.687230353382354, + "learning_rate": 2.788594681309061e-06, + "loss": 0.203, + "step": 2419 + }, + { + "epoch": 1.7406941197626327, + "grad_norm": 6.527378717586143, + "learning_rate": 2.788420487223818e-06, + "loss": 0.2212, + "step": 2420 + }, + { + "epoch": 1.7414134148534437, + "grad_norm": 3.5944677575595545, + "learning_rate": 2.7882462268470796e-06, + "loss": 0.0949, + "step": 2421 + }, + { + "epoch": 1.7421327099442547, + "grad_norm": 3.900858930748474, + "learning_rate": 2.788071900187811e-06, + "loss": 0.1667, + "step": 2422 + }, + { + "epoch": 1.7428520050350658, + "grad_norm": 4.584469738180817, + "learning_rate": 2.7878975072549823e-06, + "loss": 0.1476, + "step": 2423 + }, + { + "epoch": 1.7435713001258768, + "grad_norm": 2.523693751667623, + "learning_rate": 2.787723048057566e-06, + "loss": 0.0772, + "step": 2424 + }, + { + "epoch": 1.7442905952166876, + "grad_norm": 6.124401260014997, + "learning_rate": 2.7875485226045385e-06, + "loss": 0.2145, + "step": 2425 + }, + { + "epoch": 1.7450098903074986, + "grad_norm": 4.7632250768252575, + "learning_rate": 2.787373930904879e-06, + "loss": 0.2092, + "step": 2426 + }, + { + "epoch": 1.7457291853983097, + "grad_norm": 3.851149311424704, + "learning_rate": 2.7871992729675705e-06, + "loss": 0.1175, + "step": 2427 + }, + { + "epoch": 1.7464484804891207, + "grad_norm": 3.1676414628856917, + "learning_rate": 2.7870245488016005e-06, + "loss": 0.0257, + "step": 2428 + }, + { + "epoch": 1.7471677755799315, + "grad_norm": 2.5536292631549937, + "learning_rate": 2.786849758415958e-06, + "loss": 0.0149, + "step": 2429 + }, + { + "epoch": 1.7478870706707426, + "grad_norm": 4.859296970292306, + "learning_rate": 2.786674901819636e-06, + "loss": 0.2743, + "step": 2430 + }, + { + "epoch": 1.7486063657615536, + "grad_norm": 6.884180404294966, + "learning_rate": 2.786499979021632e-06, + "loss": 0.0488, + "step": 2431 + }, + { + "epoch": 1.7493256608523646, + "grad_norm": 5.287736487607871, + "learning_rate": 2.7863249900309456e-06, + "loss": 0.1154, + "step": 2432 + }, + { + "epoch": 1.7500449559431757, + "grad_norm": 6.156031457626568, + "learning_rate": 2.78614993485658e-06, + "loss": 0.256, + "step": 2433 + }, + { + "epoch": 1.7507642510339867, + "grad_norm": 2.9101570063235265, + "learning_rate": 2.785974813507543e-06, + "loss": 0.0436, + "step": 2434 + }, + { + "epoch": 1.7514835461247977, + "grad_norm": 1.0490987630399833, + "learning_rate": 2.7857996259928445e-06, + "loss": 0.006, + "step": 2435 + }, + { + "epoch": 1.7522028412156088, + "grad_norm": 1.6146349516856136, + "learning_rate": 2.7856243723214972e-06, + "loss": 0.0211, + "step": 2436 + }, + { + "epoch": 1.7529221363064198, + "grad_norm": 2.5100020503515794, + "learning_rate": 2.78544905250252e-06, + "loss": 0.0279, + "step": 2437 + }, + { + "epoch": 1.7536414313972308, + "grad_norm": 0.5825906204160238, + "learning_rate": 2.7852736665449325e-06, + "loss": 0.0029, + "step": 2438 + }, + { + "epoch": 1.7543607264880419, + "grad_norm": 2.820026517255655, + "learning_rate": 2.7850982144577585e-06, + "loss": 0.0271, + "step": 2439 + }, + { + "epoch": 1.7550800215788527, + "grad_norm": 1.776986012527359, + "learning_rate": 2.784922696250025e-06, + "loss": 0.004, + "step": 2440 + }, + { + "epoch": 1.7557993166696637, + "grad_norm": 1.6706054156054357, + "learning_rate": 2.7847471119307635e-06, + "loss": 0.0256, + "step": 2441 + }, + { + "epoch": 1.7565186117604747, + "grad_norm": 4.0572610573536885, + "learning_rate": 2.784571461509008e-06, + "loss": 0.0942, + "step": 2442 + }, + { + "epoch": 1.7572379068512858, + "grad_norm": 3.681397781999924, + "learning_rate": 2.7843957449937956e-06, + "loss": 0.0812, + "step": 2443 + }, + { + "epoch": 1.7579572019420966, + "grad_norm": 2.813872836830971, + "learning_rate": 2.7842199623941674e-06, + "loss": 0.1138, + "step": 2444 + }, + { + "epoch": 1.7586764970329076, + "grad_norm": 4.527030593476702, + "learning_rate": 2.7840441137191676e-06, + "loss": 0.13, + "step": 2445 + }, + { + "epoch": 1.7593957921237187, + "grad_norm": 5.193605920645928, + "learning_rate": 2.7838681989778442e-06, + "loss": 0.0815, + "step": 2446 + }, + { + "epoch": 1.7601150872145297, + "grad_norm": 0.8159936995068852, + "learning_rate": 2.7836922181792483e-06, + "loss": 0.0028, + "step": 2447 + }, + { + "epoch": 1.7608343823053407, + "grad_norm": 5.553293228334078, + "learning_rate": 2.7835161713324346e-06, + "loss": 0.155, + "step": 2448 + }, + { + "epoch": 1.7615536773961518, + "grad_norm": 1.8438614964271234, + "learning_rate": 2.7833400584464604e-06, + "loss": 0.0408, + "step": 2449 + }, + { + "epoch": 1.7622729724869628, + "grad_norm": 5.0434768747216046, + "learning_rate": 2.7831638795303873e-06, + "loss": 0.1741, + "step": 2450 + }, + { + "epoch": 1.7629922675777738, + "grad_norm": 3.008907579746472, + "learning_rate": 2.78298763459328e-06, + "loss": 0.1955, + "step": 2451 + }, + { + "epoch": 1.7637115626685849, + "grad_norm": 3.9620030366647105, + "learning_rate": 2.782811323644207e-06, + "loss": 0.0852, + "step": 2452 + }, + { + "epoch": 1.764430857759396, + "grad_norm": 3.316202661520531, + "learning_rate": 2.7826349466922396e-06, + "loss": 0.0862, + "step": 2453 + }, + { + "epoch": 1.765150152850207, + "grad_norm": 2.2031411658279576, + "learning_rate": 2.782458503746452e-06, + "loss": 0.0337, + "step": 2454 + }, + { + "epoch": 1.765869447941018, + "grad_norm": 3.2166319287286638, + "learning_rate": 2.7822819948159237e-06, + "loss": 0.0957, + "step": 2455 + }, + { + "epoch": 1.7665887430318288, + "grad_norm": 3.240275325137214, + "learning_rate": 2.7821054199097357e-06, + "loss": 0.0836, + "step": 2456 + }, + { + "epoch": 1.7673080381226398, + "grad_norm": 5.202324080627204, + "learning_rate": 2.781928779036973e-06, + "loss": 0.1207, + "step": 2457 + }, + { + "epoch": 1.7680273332134508, + "grad_norm": 7.2558041039963115, + "learning_rate": 2.781752072206724e-06, + "loss": 0.1287, + "step": 2458 + }, + { + "epoch": 1.7687466283042617, + "grad_norm": 5.53981910154158, + "learning_rate": 2.781575299428081e-06, + "loss": 0.1373, + "step": 2459 + }, + { + "epoch": 1.7694659233950727, + "grad_norm": 4.104216394763029, + "learning_rate": 2.7813984607101394e-06, + "loss": 0.1158, + "step": 2460 + }, + { + "epoch": 1.7701852184858837, + "grad_norm": 1.4470698433642928, + "learning_rate": 2.7812215560619977e-06, + "loss": 0.0294, + "step": 2461 + }, + { + "epoch": 1.7709045135766948, + "grad_norm": 5.669550095457154, + "learning_rate": 2.781044585492757e-06, + "loss": 0.2916, + "step": 2462 + }, + { + "epoch": 1.7716238086675058, + "grad_norm": 3.1666442293662467, + "learning_rate": 2.7808675490115246e-06, + "loss": 0.0894, + "step": 2463 + }, + { + "epoch": 1.7723431037583168, + "grad_norm": 2.712345202276612, + "learning_rate": 2.780690446627408e-06, + "loss": 0.0557, + "step": 2464 + }, + { + "epoch": 1.7730623988491279, + "grad_norm": 5.505570748437723, + "learning_rate": 2.7805132783495192e-06, + "loss": 0.2554, + "step": 2465 + }, + { + "epoch": 1.773781693939939, + "grad_norm": 5.542460629541906, + "learning_rate": 2.7803360441869747e-06, + "loss": 0.0789, + "step": 2466 + }, + { + "epoch": 1.77450098903075, + "grad_norm": 4.138194696914727, + "learning_rate": 2.780158744148893e-06, + "loss": 0.1718, + "step": 2467 + }, + { + "epoch": 1.775220284121561, + "grad_norm": 3.203864931136845, + "learning_rate": 2.779981378244397e-06, + "loss": 0.1018, + "step": 2468 + }, + { + "epoch": 1.775939579212372, + "grad_norm": 4.099174749866319, + "learning_rate": 2.7798039464826124e-06, + "loss": 0.0531, + "step": 2469 + }, + { + "epoch": 1.776658874303183, + "grad_norm": 6.047535212216035, + "learning_rate": 2.7796264488726674e-06, + "loss": 0.2787, + "step": 2470 + }, + { + "epoch": 1.7773781693939938, + "grad_norm": 5.314497018473563, + "learning_rate": 2.779448885423696e-06, + "loss": 0.0702, + "step": 2471 + }, + { + "epoch": 1.7780974644848049, + "grad_norm": 2.278598676722101, + "learning_rate": 2.779271256144833e-06, + "loss": 0.0968, + "step": 2472 + }, + { + "epoch": 1.778816759575616, + "grad_norm": 0.05148420749332451, + "learning_rate": 2.779093561045218e-06, + "loss": 0.0003, + "step": 2473 + }, + { + "epoch": 1.779536054666427, + "grad_norm": 6.310569639532884, + "learning_rate": 2.778915800133994e-06, + "loss": 0.4003, + "step": 2474 + }, + { + "epoch": 1.7802553497572378, + "grad_norm": 6.283635654690987, + "learning_rate": 2.7787379734203072e-06, + "loss": 0.0724, + "step": 2475 + }, + { + "epoch": 1.7809746448480488, + "grad_norm": 3.9531614800696007, + "learning_rate": 2.778560080913307e-06, + "loss": 0.1231, + "step": 2476 + }, + { + "epoch": 1.7816939399388598, + "grad_norm": 6.43125342617877, + "learning_rate": 2.7783821226221462e-06, + "loss": 0.0852, + "step": 2477 + }, + { + "epoch": 1.7824132350296709, + "grad_norm": 7.028478314159626, + "learning_rate": 2.778204098555981e-06, + "loss": 0.3916, + "step": 2478 + }, + { + "epoch": 1.783132530120482, + "grad_norm": 4.739546740935013, + "learning_rate": 2.778026008723971e-06, + "loss": 0.0915, + "step": 2479 + }, + { + "epoch": 1.783851825211293, + "grad_norm": 3.5839932374655885, + "learning_rate": 2.7778478531352794e-06, + "loss": 0.1126, + "step": 2480 + }, + { + "epoch": 1.784571120302104, + "grad_norm": 2.3778317157586706, + "learning_rate": 2.777669631799073e-06, + "loss": 0.0692, + "step": 2481 + }, + { + "epoch": 1.785290415392915, + "grad_norm": 3.334496721064537, + "learning_rate": 2.7774913447245202e-06, + "loss": 0.1262, + "step": 2482 + }, + { + "epoch": 1.786009710483726, + "grad_norm": 3.2463333086306787, + "learning_rate": 2.777312991920796e-06, + "loss": 0.046, + "step": 2483 + }, + { + "epoch": 1.786729005574537, + "grad_norm": 3.505916217151454, + "learning_rate": 2.777134573397076e-06, + "loss": 0.082, + "step": 2484 + }, + { + "epoch": 1.787448300665348, + "grad_norm": 6.085618892510015, + "learning_rate": 2.77695608916254e-06, + "loss": 0.2499, + "step": 2485 + }, + { + "epoch": 1.788167595756159, + "grad_norm": 3.48603526679261, + "learning_rate": 2.7767775392263714e-06, + "loss": 0.0787, + "step": 2486 + }, + { + "epoch": 1.78888689084697, + "grad_norm": 2.1435328531321587, + "learning_rate": 2.7765989235977573e-06, + "loss": 0.0138, + "step": 2487 + }, + { + "epoch": 1.789606185937781, + "grad_norm": 5.289632312079522, + "learning_rate": 2.7764202422858873e-06, + "loss": 0.2931, + "step": 2488 + }, + { + "epoch": 1.790325481028592, + "grad_norm": 5.343247603244428, + "learning_rate": 2.776241495299956e-06, + "loss": 0.146, + "step": 2489 + }, + { + "epoch": 1.7910447761194028, + "grad_norm": 4.656668544117325, + "learning_rate": 2.7760626826491588e-06, + "loss": 0.1656, + "step": 2490 + }, + { + "epoch": 1.7917640712102139, + "grad_norm": 6.237523832262931, + "learning_rate": 2.7758838043426965e-06, + "loss": 0.0507, + "step": 2491 + }, + { + "epoch": 1.792483366301025, + "grad_norm": 4.491730845376762, + "learning_rate": 2.7757048603897726e-06, + "loss": 0.1342, + "step": 2492 + }, + { + "epoch": 1.793202661391836, + "grad_norm": 6.478613923551439, + "learning_rate": 2.7755258507995943e-06, + "loss": 0.1437, + "step": 2493 + }, + { + "epoch": 1.793921956482647, + "grad_norm": 4.411105366060626, + "learning_rate": 2.7753467755813724e-06, + "loss": 0.0242, + "step": 2494 + }, + { + "epoch": 1.794641251573458, + "grad_norm": 0.8858500767843286, + "learning_rate": 2.7751676347443194e-06, + "loss": 0.0043, + "step": 2495 + }, + { + "epoch": 1.795360546664269, + "grad_norm": 2.6120016270015123, + "learning_rate": 2.7749884282976534e-06, + "loss": 0.0141, + "step": 2496 + }, + { + "epoch": 1.79607984175508, + "grad_norm": 1.1829032619641606, + "learning_rate": 2.7748091562505946e-06, + "loss": 0.0025, + "step": 2497 + }, + { + "epoch": 1.796799136845891, + "grad_norm": 5.9738657468641705, + "learning_rate": 2.7746298186123667e-06, + "loss": 0.1216, + "step": 2498 + }, + { + "epoch": 1.7975184319367021, + "grad_norm": 3.551417411720854, + "learning_rate": 2.774450415392197e-06, + "loss": 0.1287, + "step": 2499 + }, + { + "epoch": 1.7982377270275132, + "grad_norm": 5.570595282951437, + "learning_rate": 2.774270946599317e-06, + "loss": 0.1204, + "step": 2500 + }, + { + "epoch": 1.7989570221183242, + "grad_norm": 6.634960975916419, + "learning_rate": 2.7740914122429596e-06, + "loss": 0.1355, + "step": 2501 + }, + { + "epoch": 1.799676317209135, + "grad_norm": 1.370184468491438, + "learning_rate": 2.773911812332362e-06, + "loss": 0.0271, + "step": 2502 + }, + { + "epoch": 1.800395612299946, + "grad_norm": 3.719098128948436, + "learning_rate": 2.7737321468767657e-06, + "loss": 0.026, + "step": 2503 + }, + { + "epoch": 1.801114907390757, + "grad_norm": 4.413730591800512, + "learning_rate": 2.7735524158854136e-06, + "loss": 0.1071, + "step": 2504 + }, + { + "epoch": 1.8018342024815681, + "grad_norm": 4.423628648379886, + "learning_rate": 2.773372619367555e-06, + "loss": 0.1083, + "step": 2505 + }, + { + "epoch": 1.802553497572379, + "grad_norm": 6.1161384519081965, + "learning_rate": 2.7731927573324395e-06, + "loss": 0.3374, + "step": 2506 + }, + { + "epoch": 1.80327279266319, + "grad_norm": 3.3458772925524523, + "learning_rate": 2.773012829789322e-06, + "loss": 0.1452, + "step": 2507 + }, + { + "epoch": 1.803992087754001, + "grad_norm": 3.042766699375507, + "learning_rate": 2.7728328367474588e-06, + "loss": 0.0402, + "step": 2508 + }, + { + "epoch": 1.804711382844812, + "grad_norm": 4.947010356396303, + "learning_rate": 2.7726527782161126e-06, + "loss": 0.099, + "step": 2509 + }, + { + "epoch": 1.805430677935623, + "grad_norm": 5.7295888327825155, + "learning_rate": 2.7724726542045464e-06, + "loss": 0.1001, + "step": 2510 + }, + { + "epoch": 1.806149973026434, + "grad_norm": 4.448403931657433, + "learning_rate": 2.772292464722029e-06, + "loss": 0.1106, + "step": 2511 + }, + { + "epoch": 1.8068692681172451, + "grad_norm": 3.1343087476722085, + "learning_rate": 2.77211220977783e-06, + "loss": 0.0629, + "step": 2512 + }, + { + "epoch": 1.8075885632080562, + "grad_norm": 4.799081920086391, + "learning_rate": 2.7719318893812254e-06, + "loss": 0.0605, + "step": 2513 + }, + { + "epoch": 1.8083078582988672, + "grad_norm": 4.8838820525245294, + "learning_rate": 2.7717515035414923e-06, + "loss": 0.0532, + "step": 2514 + }, + { + "epoch": 1.8090271533896782, + "grad_norm": 1.9114660253876055, + "learning_rate": 2.7715710522679113e-06, + "loss": 0.0454, + "step": 2515 + }, + { + "epoch": 1.8097464484804893, + "grad_norm": 3.678612753068636, + "learning_rate": 2.7713905355697676e-06, + "loss": 0.108, + "step": 2516 + }, + { + "epoch": 1.8104657435713, + "grad_norm": 3.132743077380941, + "learning_rate": 2.7712099534563494e-06, + "loss": 0.0211, + "step": 2517 + }, + { + "epoch": 1.8111850386621111, + "grad_norm": 7.180254446978263, + "learning_rate": 2.7710293059369473e-06, + "loss": 0.0688, + "step": 2518 + }, + { + "epoch": 1.8119043337529221, + "grad_norm": 3.3402755016222083, + "learning_rate": 2.7708485930208563e-06, + "loss": 0.0518, + "step": 2519 + }, + { + "epoch": 1.8126236288437332, + "grad_norm": 5.8339881130520626, + "learning_rate": 2.770667814717374e-06, + "loss": 0.1752, + "step": 2520 + }, + { + "epoch": 1.813342923934544, + "grad_norm": 3.3025676992552544, + "learning_rate": 2.7704869710358027e-06, + "loss": 0.0311, + "step": 2521 + }, + { + "epoch": 1.814062219025355, + "grad_norm": 0.6364832336183773, + "learning_rate": 2.7703060619854456e-06, + "loss": 0.0015, + "step": 2522 + }, + { + "epoch": 1.814781514116166, + "grad_norm": 6.372236187824131, + "learning_rate": 2.7701250875756123e-06, + "loss": 0.4278, + "step": 2523 + }, + { + "epoch": 1.815500809206977, + "grad_norm": 3.03470470010061, + "learning_rate": 2.769944047815613e-06, + "loss": 0.098, + "step": 2524 + }, + { + "epoch": 1.8162201042977881, + "grad_norm": 2.490473441477687, + "learning_rate": 2.769762942714764e-06, + "loss": 0.0312, + "step": 2525 + }, + { + "epoch": 1.8169393993885992, + "grad_norm": 6.379695479011435, + "learning_rate": 2.769581772282382e-06, + "loss": 0.2005, + "step": 2526 + }, + { + "epoch": 1.8176586944794102, + "grad_norm": 0.5678288856838097, + "learning_rate": 2.769400536527789e-06, + "loss": 0.0043, + "step": 2527 + }, + { + "epoch": 1.8183779895702212, + "grad_norm": 5.136506839394631, + "learning_rate": 2.76921923546031e-06, + "loss": 0.0227, + "step": 2528 + }, + { + "epoch": 1.8190972846610323, + "grad_norm": 4.59435729111555, + "learning_rate": 2.769037869089274e-06, + "loss": 0.1467, + "step": 2529 + }, + { + "epoch": 1.8198165797518433, + "grad_norm": 6.142522539934962, + "learning_rate": 2.7688564374240113e-06, + "loss": 0.2131, + "step": 2530 + }, + { + "epoch": 1.8205358748426543, + "grad_norm": 2.7401917562809595, + "learning_rate": 2.7686749404738578e-06, + "loss": 0.042, + "step": 2531 + }, + { + "epoch": 1.8212551699334654, + "grad_norm": 4.484183712917439, + "learning_rate": 2.7684933782481516e-06, + "loss": 0.1757, + "step": 2532 + }, + { + "epoch": 1.8219744650242762, + "grad_norm": 4.321577310364181, + "learning_rate": 2.768311750756234e-06, + "loss": 0.0577, + "step": 2533 + }, + { + "epoch": 1.8226937601150872, + "grad_norm": 5.6410512813533655, + "learning_rate": 2.7681300580074503e-06, + "loss": 0.1218, + "step": 2534 + }, + { + "epoch": 1.8234130552058982, + "grad_norm": 1.4061629497049661, + "learning_rate": 2.767948300011149e-06, + "loss": 0.0082, + "step": 2535 + }, + { + "epoch": 1.824132350296709, + "grad_norm": 5.666716902630463, + "learning_rate": 2.7677664767766826e-06, + "loss": 0.214, + "step": 2536 + }, + { + "epoch": 1.82485164538752, + "grad_norm": 5.501311365360597, + "learning_rate": 2.7675845883134043e-06, + "loss": 0.1412, + "step": 2537 + }, + { + "epoch": 1.8255709404783311, + "grad_norm": 6.143742811163902, + "learning_rate": 2.7674026346306745e-06, + "loss": 0.2483, + "step": 2538 + }, + { + "epoch": 1.8262902355691422, + "grad_norm": 3.8972748751103556, + "learning_rate": 2.767220615737854e-06, + "loss": 0.1379, + "step": 2539 + }, + { + "epoch": 1.8270095306599532, + "grad_norm": 4.082049239194885, + "learning_rate": 2.7670385316443084e-06, + "loss": 0.2196, + "step": 2540 + }, + { + "epoch": 1.8277288257507642, + "grad_norm": 0.8622557730121388, + "learning_rate": 2.7668563823594063e-06, + "loss": 0.0042, + "step": 2541 + }, + { + "epoch": 1.8284481208415753, + "grad_norm": 5.33404811561849, + "learning_rate": 2.766674167892519e-06, + "loss": 0.1734, + "step": 2542 + }, + { + "epoch": 1.8291674159323863, + "grad_norm": 9.41902966258388, + "learning_rate": 2.7664918882530226e-06, + "loss": 0.0514, + "step": 2543 + }, + { + "epoch": 1.8298867110231973, + "grad_norm": 2.108167626441434, + "learning_rate": 2.766309543450295e-06, + "loss": 0.0567, + "step": 2544 + }, + { + "epoch": 1.8306060061140084, + "grad_norm": 3.797098679091282, + "learning_rate": 2.766127133493719e-06, + "loss": 0.0969, + "step": 2545 + }, + { + "epoch": 1.8313253012048194, + "grad_norm": 5.9798385156995275, + "learning_rate": 2.7659446583926786e-06, + "loss": 0.2701, + "step": 2546 + }, + { + "epoch": 1.8320445962956304, + "grad_norm": 1.7870722112789903, + "learning_rate": 2.7657621181565637e-06, + "loss": 0.029, + "step": 2547 + }, + { + "epoch": 1.8327638913864412, + "grad_norm": 7.096455078234722, + "learning_rate": 2.765579512794766e-06, + "loss": 0.248, + "step": 2548 + }, + { + "epoch": 1.8334831864772523, + "grad_norm": 2.8736836913931745, + "learning_rate": 2.7653968423166806e-06, + "loss": 0.0865, + "step": 2549 + }, + { + "epoch": 1.8342024815680633, + "grad_norm": 2.703632678442727, + "learning_rate": 2.765214106731706e-06, + "loss": 0.1338, + "step": 2550 + }, + { + "epoch": 1.8349217766588743, + "grad_norm": 4.787599189402569, + "learning_rate": 2.765031306049245e-06, + "loss": 0.2232, + "step": 2551 + }, + { + "epoch": 1.8356410717496852, + "grad_norm": 2.9099792337197896, + "learning_rate": 2.7648484402787024e-06, + "loss": 0.0306, + "step": 2552 + }, + { + "epoch": 1.8363603668404962, + "grad_norm": 3.7077688443027346, + "learning_rate": 2.764665509429487e-06, + "loss": 0.158, + "step": 2553 + }, + { + "epoch": 1.8370796619313072, + "grad_norm": 6.2501465658507565, + "learning_rate": 2.764482513511011e-06, + "loss": 0.1424, + "step": 2554 + }, + { + "epoch": 1.8377989570221183, + "grad_norm": 4.593488218805693, + "learning_rate": 2.7642994525326906e-06, + "loss": 0.097, + "step": 2555 + }, + { + "epoch": 1.8385182521129293, + "grad_norm": 4.865436683626331, + "learning_rate": 2.7641163265039436e-06, + "loss": 0.1259, + "step": 2556 + }, + { + "epoch": 1.8392375472037403, + "grad_norm": 3.4111713371569348, + "learning_rate": 2.7639331354341924e-06, + "loss": 0.1456, + "step": 2557 + }, + { + "epoch": 1.8399568422945514, + "grad_norm": 6.077197315162484, + "learning_rate": 2.763749879332863e-06, + "loss": 0.1722, + "step": 2558 + }, + { + "epoch": 1.8406761373853624, + "grad_norm": 4.350936199232469, + "learning_rate": 2.7635665582093833e-06, + "loss": 0.1583, + "step": 2559 + }, + { + "epoch": 1.8413954324761734, + "grad_norm": 5.393039107024693, + "learning_rate": 2.7633831720731862e-06, + "loss": 0.1353, + "step": 2560 + }, + { + "epoch": 1.8421147275669845, + "grad_norm": 3.724004833307669, + "learning_rate": 2.7631997209337076e-06, + "loss": 0.0764, + "step": 2561 + }, + { + "epoch": 1.8428340226577955, + "grad_norm": 3.39537330649432, + "learning_rate": 2.763016204800385e-06, + "loss": 0.1358, + "step": 2562 + }, + { + "epoch": 1.8435533177486063, + "grad_norm": 4.16350961159215, + "learning_rate": 2.7628326236826623e-06, + "loss": 0.1239, + "step": 2563 + }, + { + "epoch": 1.8442726128394173, + "grad_norm": 2.3934462367152713, + "learning_rate": 2.762648977589984e-06, + "loss": 0.0467, + "step": 2564 + }, + { + "epoch": 1.8449919079302284, + "grad_norm": 4.079145847815187, + "learning_rate": 2.7624652665317993e-06, + "loss": 0.0292, + "step": 2565 + }, + { + "epoch": 1.8457112030210394, + "grad_norm": 3.829549074675402, + "learning_rate": 2.76228149051756e-06, + "loss": 0.1173, + "step": 2566 + }, + { + "epoch": 1.8464304981118502, + "grad_norm": 2.8405438174038697, + "learning_rate": 2.762097649556723e-06, + "loss": 0.1238, + "step": 2567 + }, + { + "epoch": 1.8471497932026613, + "grad_norm": 0.8928345250244636, + "learning_rate": 2.761913743658746e-06, + "loss": 0.0017, + "step": 2568 + }, + { + "epoch": 1.8478690882934723, + "grad_norm": 1.351053113380072, + "learning_rate": 2.7617297728330914e-06, + "loss": 0.0224, + "step": 2569 + }, + { + "epoch": 1.8485883833842833, + "grad_norm": 3.9917059487893614, + "learning_rate": 2.7615457370892255e-06, + "loss": 0.1113, + "step": 2570 + }, + { + "epoch": 1.8493076784750944, + "grad_norm": 5.1258846233141595, + "learning_rate": 2.7613616364366164e-06, + "loss": 0.1161, + "step": 2571 + }, + { + "epoch": 1.8500269735659054, + "grad_norm": 5.3875538424950395, + "learning_rate": 2.7611774708847375e-06, + "loss": 0.1207, + "step": 2572 + }, + { + "epoch": 1.8507462686567164, + "grad_norm": 6.870508253130083, + "learning_rate": 2.7609932404430634e-06, + "loss": 0.134, + "step": 2573 + }, + { + "epoch": 1.8514655637475275, + "grad_norm": 3.103572294299757, + "learning_rate": 2.7608089451210737e-06, + "loss": 0.0105, + "step": 2574 + }, + { + "epoch": 1.8521848588383385, + "grad_norm": 4.7453135593811435, + "learning_rate": 2.76062458492825e-06, + "loss": 0.1542, + "step": 2575 + }, + { + "epoch": 1.8529041539291495, + "grad_norm": 2.735687693644433, + "learning_rate": 2.7604401598740787e-06, + "loss": 0.087, + "step": 2576 + }, + { + "epoch": 1.8536234490199606, + "grad_norm": 4.79297985610655, + "learning_rate": 2.7602556699680486e-06, + "loss": 0.105, + "step": 2577 + }, + { + "epoch": 1.8543427441107716, + "grad_norm": 9.7597237804412, + "learning_rate": 2.760071115219652e-06, + "loss": 0.341, + "step": 2578 + }, + { + "epoch": 1.8550620392015824, + "grad_norm": 1.5567371165359478, + "learning_rate": 2.7598864956383844e-06, + "loss": 0.0037, + "step": 2579 + }, + { + "epoch": 1.8557813342923934, + "grad_norm": 7.088457252028718, + "learning_rate": 2.759701811233745e-06, + "loss": 0.1921, + "step": 2580 + }, + { + "epoch": 1.8565006293832045, + "grad_norm": 4.789658381825956, + "learning_rate": 2.7595170620152367e-06, + "loss": 0.0824, + "step": 2581 + }, + { + "epoch": 1.8572199244740155, + "grad_norm": 7.203203258850551, + "learning_rate": 2.7593322479923637e-06, + "loss": 0.3552, + "step": 2582 + }, + { + "epoch": 1.8579392195648263, + "grad_norm": 3.8164995313332657, + "learning_rate": 2.759147369174636e-06, + "loss": 0.0987, + "step": 2583 + }, + { + "epoch": 1.8586585146556374, + "grad_norm": 4.145443626045281, + "learning_rate": 2.758962425571566e-06, + "loss": 0.1554, + "step": 2584 + }, + { + "epoch": 1.8593778097464484, + "grad_norm": 2.859034568344092, + "learning_rate": 2.7587774171926694e-06, + "loss": 0.0499, + "step": 2585 + }, + { + "epoch": 1.8600971048372594, + "grad_norm": 4.499078457035729, + "learning_rate": 2.7585923440474642e-06, + "loss": 0.0215, + "step": 2586 + }, + { + "epoch": 1.8608163999280705, + "grad_norm": 3.5980214138060136, + "learning_rate": 2.758407206145474e-06, + "loss": 0.1175, + "step": 2587 + }, + { + "epoch": 1.8615356950188815, + "grad_norm": 5.321695178615645, + "learning_rate": 2.7582220034962237e-06, + "loss": 0.2354, + "step": 2588 + }, + { + "epoch": 1.8622549901096925, + "grad_norm": 1.6756558997411046, + "learning_rate": 2.758036736109243e-06, + "loss": 0.0204, + "step": 2589 + }, + { + "epoch": 1.8629742852005036, + "grad_norm": 2.113439479447713, + "learning_rate": 2.7578514039940634e-06, + "loss": 0.0151, + "step": 2590 + }, + { + "epoch": 1.8636935802913146, + "grad_norm": 4.0202960158140355, + "learning_rate": 2.7576660071602217e-06, + "loss": 0.16, + "step": 2591 + }, + { + "epoch": 1.8644128753821256, + "grad_norm": 7.580735602536769, + "learning_rate": 2.7574805456172553e-06, + "loss": 0.2725, + "step": 2592 + }, + { + "epoch": 1.8651321704729367, + "grad_norm": 6.020124808693399, + "learning_rate": 2.757295019374708e-06, + "loss": 0.2323, + "step": 2593 + }, + { + "epoch": 1.8658514655637475, + "grad_norm": 1.9967106864927242, + "learning_rate": 2.7571094284421247e-06, + "loss": 0.0081, + "step": 2594 + }, + { + "epoch": 1.8665707606545585, + "grad_norm": 7.655655997374509, + "learning_rate": 2.7569237728290543e-06, + "loss": 0.3506, + "step": 2595 + }, + { + "epoch": 1.8672900557453695, + "grad_norm": 4.758963619368994, + "learning_rate": 2.7567380525450494e-06, + "loss": 0.1633, + "step": 2596 + }, + { + "epoch": 1.8680093508361806, + "grad_norm": 2.1736695132744055, + "learning_rate": 2.7565522675996656e-06, + "loss": 0.0791, + "step": 2597 + }, + { + "epoch": 1.8687286459269914, + "grad_norm": 2.6852785671423516, + "learning_rate": 2.7563664180024623e-06, + "loss": 0.0302, + "step": 2598 + }, + { + "epoch": 1.8694479410178024, + "grad_norm": 4.44922070142439, + "learning_rate": 2.7561805037630006e-06, + "loss": 0.0816, + "step": 2599 + }, + { + "epoch": 1.8701672361086135, + "grad_norm": 5.138668862679516, + "learning_rate": 2.755994524890847e-06, + "loss": 0.0263, + "step": 2600 + }, + { + "epoch": 1.8708865311994245, + "grad_norm": 5.33058579020469, + "learning_rate": 2.75580848139557e-06, + "loss": 0.108, + "step": 2601 + }, + { + "epoch": 1.8716058262902355, + "grad_norm": 4.620534925820611, + "learning_rate": 2.7556223732867428e-06, + "loss": 0.0601, + "step": 2602 + }, + { + "epoch": 1.8723251213810466, + "grad_norm": 1.4127467525521846, + "learning_rate": 2.75543620057394e-06, + "loss": 0.0293, + "step": 2603 + }, + { + "epoch": 1.8730444164718576, + "grad_norm": 5.735407022147661, + "learning_rate": 2.755249963266741e-06, + "loss": 0.1163, + "step": 2604 + }, + { + "epoch": 1.8737637115626686, + "grad_norm": 2.6254787174382495, + "learning_rate": 2.7550636613747277e-06, + "loss": 0.045, + "step": 2605 + }, + { + "epoch": 1.8744830066534797, + "grad_norm": 5.262904717803658, + "learning_rate": 2.7548772949074858e-06, + "loss": 0.1409, + "step": 2606 + }, + { + "epoch": 1.8752023017442907, + "grad_norm": 3.313369242155628, + "learning_rate": 2.7546908638746044e-06, + "loss": 0.0835, + "step": 2607 + }, + { + "epoch": 1.8759215968351017, + "grad_norm": 5.690519549358206, + "learning_rate": 2.754504368285675e-06, + "loss": 0.1302, + "step": 2608 + }, + { + "epoch": 1.8766408919259128, + "grad_norm": 3.8559148294499064, + "learning_rate": 2.7543178081502944e-06, + "loss": 0.2305, + "step": 2609 + }, + { + "epoch": 1.8773601870167236, + "grad_norm": 4.165036960653529, + "learning_rate": 2.75413118347806e-06, + "loss": 0.1524, + "step": 2610 + }, + { + "epoch": 1.8780794821075346, + "grad_norm": 7.982903612558595, + "learning_rate": 2.7539444942785746e-06, + "loss": 0.2369, + "step": 2611 + }, + { + "epoch": 1.8787987771983456, + "grad_norm": 3.932219405413893, + "learning_rate": 2.753757740561444e-06, + "loss": 0.2065, + "step": 2612 + }, + { + "epoch": 1.8795180722891565, + "grad_norm": 3.8268151438795592, + "learning_rate": 2.753570922336277e-06, + "loss": 0.017, + "step": 2613 + }, + { + "epoch": 1.8802373673799675, + "grad_norm": 2.4966641738816078, + "learning_rate": 2.7533840396126856e-06, + "loss": 0.0936, + "step": 2614 + }, + { + "epoch": 1.8809566624707785, + "grad_norm": 2.907683058504994, + "learning_rate": 2.7531970924002847e-06, + "loss": 0.0721, + "step": 2615 + }, + { + "epoch": 1.8816759575615896, + "grad_norm": 5.254665989150888, + "learning_rate": 2.7530100807086933e-06, + "loss": 0.0943, + "step": 2616 + }, + { + "epoch": 1.8823952526524006, + "grad_norm": 3.4759941344314944, + "learning_rate": 2.752823004547534e-06, + "loss": 0.0914, + "step": 2617 + }, + { + "epoch": 1.8831145477432116, + "grad_norm": 6.3961238833266485, + "learning_rate": 2.752635863926432e-06, + "loss": 0.0936, + "step": 2618 + }, + { + "epoch": 1.8838338428340227, + "grad_norm": 3.7628225317620707, + "learning_rate": 2.752448658855015e-06, + "loss": 0.2094, + "step": 2619 + }, + { + "epoch": 1.8845531379248337, + "grad_norm": 1.5933574618098518, + "learning_rate": 2.752261389342917e-06, + "loss": 0.0059, + "step": 2620 + }, + { + "epoch": 1.8852724330156447, + "grad_norm": 5.573897749077213, + "learning_rate": 2.752074055399772e-06, + "loss": 0.1308, + "step": 2621 + }, + { + "epoch": 1.8859917281064558, + "grad_norm": 5.210401869334891, + "learning_rate": 2.7518866570352184e-06, + "loss": 0.1528, + "step": 2622 + }, + { + "epoch": 1.8867110231972668, + "grad_norm": 4.6923379508666025, + "learning_rate": 2.751699194258899e-06, + "loss": 0.1212, + "step": 2623 + }, + { + "epoch": 1.8874303182880778, + "grad_norm": 2.525731022066795, + "learning_rate": 2.7515116670804587e-06, + "loss": 0.1162, + "step": 2624 + }, + { + "epoch": 1.8881496133788886, + "grad_norm": 3.53999169828396, + "learning_rate": 2.7513240755095462e-06, + "loss": 0.0713, + "step": 2625 + }, + { + "epoch": 1.8888689084696997, + "grad_norm": 4.513193890063883, + "learning_rate": 2.751136419555813e-06, + "loss": 0.1254, + "step": 2626 + }, + { + "epoch": 1.8895882035605107, + "grad_norm": 5.751532062687417, + "learning_rate": 2.7509486992289157e-06, + "loss": 0.1817, + "step": 2627 + }, + { + "epoch": 1.8903074986513217, + "grad_norm": 4.685866684721937, + "learning_rate": 2.7507609145385112e-06, + "loss": 0.128, + "step": 2628 + }, + { + "epoch": 1.8910267937421326, + "grad_norm": 2.0287089955374547, + "learning_rate": 2.750573065494262e-06, + "loss": 0.0492, + "step": 2629 + }, + { + "epoch": 1.8917460888329436, + "grad_norm": 5.949195719536943, + "learning_rate": 2.7503851521058333e-06, + "loss": 0.1583, + "step": 2630 + }, + { + "epoch": 1.8924653839237546, + "grad_norm": 3.2664639232048986, + "learning_rate": 2.750197174382894e-06, + "loss": 0.1058, + "step": 2631 + }, + { + "epoch": 1.8931846790145657, + "grad_norm": 5.8958012880129775, + "learning_rate": 2.7500091323351146e-06, + "loss": 0.0533, + "step": 2632 + }, + { + "epoch": 1.8939039741053767, + "grad_norm": 3.0500755610443875, + "learning_rate": 2.7498210259721717e-06, + "loss": 0.0786, + "step": 2633 + }, + { + "epoch": 1.8946232691961877, + "grad_norm": 0.5781434341025135, + "learning_rate": 2.749632855303743e-06, + "loss": 0.001, + "step": 2634 + }, + { + "epoch": 1.8953425642869988, + "grad_norm": 4.320845989212069, + "learning_rate": 2.7494446203395094e-06, + "loss": 0.0582, + "step": 2635 + }, + { + "epoch": 1.8960618593778098, + "grad_norm": 5.099250674385128, + "learning_rate": 2.7492563210891573e-06, + "loss": 0.1537, + "step": 2636 + }, + { + "epoch": 1.8967811544686208, + "grad_norm": 5.657414802029139, + "learning_rate": 2.7490679575623746e-06, + "loss": 0.1388, + "step": 2637 + }, + { + "epoch": 1.8975004495594319, + "grad_norm": 3.6585869069263803, + "learning_rate": 2.748879529768853e-06, + "loss": 0.198, + "step": 2638 + }, + { + "epoch": 1.898219744650243, + "grad_norm": 4.718944659374927, + "learning_rate": 2.748691037718287e-06, + "loss": 0.0789, + "step": 2639 + }, + { + "epoch": 1.8989390397410537, + "grad_norm": 7.6545431678876295, + "learning_rate": 2.748502481420375e-06, + "loss": 0.0659, + "step": 2640 + }, + { + "epoch": 1.8996583348318647, + "grad_norm": 2.9068554596982046, + "learning_rate": 2.7483138608848186e-06, + "loss": 0.0932, + "step": 2641 + }, + { + "epoch": 1.9003776299226758, + "grad_norm": 6.576430349344039, + "learning_rate": 2.748125176121323e-06, + "loss": 0.1231, + "step": 2642 + }, + { + "epoch": 1.9010969250134868, + "grad_norm": 3.473443227864142, + "learning_rate": 2.7479364271395958e-06, + "loss": 0.0188, + "step": 2643 + }, + { + "epoch": 1.9018162201042976, + "grad_norm": 3.166504488870473, + "learning_rate": 2.7477476139493492e-06, + "loss": 0.057, + "step": 2644 + }, + { + "epoch": 1.9025355151951087, + "grad_norm": 3.9595257531034527, + "learning_rate": 2.7475587365602966e-06, + "loss": 0.086, + "step": 2645 + }, + { + "epoch": 1.9032548102859197, + "grad_norm": 10.51806716799473, + "learning_rate": 2.7473697949821576e-06, + "loss": 0.0866, + "step": 2646 + }, + { + "epoch": 1.9039741053767307, + "grad_norm": 0.824083774413045, + "learning_rate": 2.7471807892246533e-06, + "loss": 0.0054, + "step": 2647 + }, + { + "epoch": 1.9046934004675418, + "grad_norm": 8.661154600812727, + "learning_rate": 2.7469917192975075e-06, + "loss": 0.2572, + "step": 2648 + }, + { + "epoch": 1.9054126955583528, + "grad_norm": 3.1450275233360907, + "learning_rate": 2.7468025852104492e-06, + "loss": 0.0199, + "step": 2649 + }, + { + "epoch": 1.9061319906491638, + "grad_norm": 5.35192237307085, + "learning_rate": 2.746613386973209e-06, + "loss": 0.0682, + "step": 2650 + }, + { + "epoch": 1.9068512857399749, + "grad_norm": 4.055016494275361, + "learning_rate": 2.7464241245955213e-06, + "loss": 0.1067, + "step": 2651 + }, + { + "epoch": 1.907570580830786, + "grad_norm": 3.711703605062127, + "learning_rate": 2.7462347980871244e-06, + "loss": 0.1793, + "step": 2652 + }, + { + "epoch": 1.908289875921597, + "grad_norm": 1.4010358917622097, + "learning_rate": 2.74604540745776e-06, + "loss": 0.0201, + "step": 2653 + }, + { + "epoch": 1.909009171012408, + "grad_norm": 2.327665862942012, + "learning_rate": 2.745855952717171e-06, + "loss": 0.0588, + "step": 2654 + }, + { + "epoch": 1.909728466103219, + "grad_norm": 8.620272490399229, + "learning_rate": 2.7456664338751074e-06, + "loss": 0.3683, + "step": 2655 + }, + { + "epoch": 1.9104477611940298, + "grad_norm": 2.514204228254778, + "learning_rate": 2.745476850941318e-06, + "loss": 0.0156, + "step": 2656 + }, + { + "epoch": 1.9111670562848408, + "grad_norm": 3.194223407839609, + "learning_rate": 2.7452872039255593e-06, + "loss": 0.0932, + "step": 2657 + }, + { + "epoch": 1.9118863513756519, + "grad_norm": 3.0134651312825143, + "learning_rate": 2.7450974928375873e-06, + "loss": 0.0702, + "step": 2658 + }, + { + "epoch": 1.912605646466463, + "grad_norm": 5.905898751964238, + "learning_rate": 2.7449077176871636e-06, + "loss": 0.2473, + "step": 2659 + }, + { + "epoch": 1.9133249415572737, + "grad_norm": 3.5573901440104736, + "learning_rate": 2.7447178784840527e-06, + "loss": 0.1384, + "step": 2660 + }, + { + "epoch": 1.9140442366480848, + "grad_norm": 1.955068255783858, + "learning_rate": 2.744527975238022e-06, + "loss": 0.0273, + "step": 2661 + }, + { + "epoch": 1.9147635317388958, + "grad_norm": 5.929418601396656, + "learning_rate": 2.7443380079588418e-06, + "loss": 0.0575, + "step": 2662 + }, + { + "epoch": 1.9154828268297068, + "grad_norm": 4.742786296076306, + "learning_rate": 2.744147976656287e-06, + "loss": 0.1081, + "step": 2663 + }, + { + "epoch": 1.9162021219205179, + "grad_norm": 0.9801979342671292, + "learning_rate": 2.743957881340135e-06, + "loss": 0.0021, + "step": 2664 + }, + { + "epoch": 1.916921417011329, + "grad_norm": 4.0643078107752295, + "learning_rate": 2.743767722020166e-06, + "loss": 0.0646, + "step": 2665 + }, + { + "epoch": 1.91764071210214, + "grad_norm": 4.589616284196056, + "learning_rate": 2.743577498706164e-06, + "loss": 0.0579, + "step": 2666 + }, + { + "epoch": 1.918360007192951, + "grad_norm": 2.877835691001051, + "learning_rate": 2.7433872114079177e-06, + "loss": 0.021, + "step": 2667 + }, + { + "epoch": 1.919079302283762, + "grad_norm": 5.184181287637767, + "learning_rate": 2.743196860135216e-06, + "loss": 0.173, + "step": 2668 + }, + { + "epoch": 1.919798597374573, + "grad_norm": 3.140217264246085, + "learning_rate": 2.7430064448978536e-06, + "loss": 0.0595, + "step": 2669 + }, + { + "epoch": 1.920517892465384, + "grad_norm": 9.49980406971857, + "learning_rate": 2.742815965705627e-06, + "loss": 0.1151, + "step": 2670 + }, + { + "epoch": 1.9212371875561949, + "grad_norm": 7.643895606210381, + "learning_rate": 2.7426254225683383e-06, + "loss": 0.1632, + "step": 2671 + }, + { + "epoch": 1.921956482647006, + "grad_norm": 2.1327788422579403, + "learning_rate": 2.7424348154957895e-06, + "loss": 0.0511, + "step": 2672 + }, + { + "epoch": 1.922675777737817, + "grad_norm": 3.8966712982055776, + "learning_rate": 2.742244144497789e-06, + "loss": 0.0598, + "step": 2673 + }, + { + "epoch": 1.923395072828628, + "grad_norm": 4.010080813656915, + "learning_rate": 2.742053409584146e-06, + "loss": 0.2432, + "step": 2674 + }, + { + "epoch": 1.9241143679194388, + "grad_norm": 3.805906449783639, + "learning_rate": 2.7418626107646747e-06, + "loss": 0.0423, + "step": 2675 + }, + { + "epoch": 1.9248336630102498, + "grad_norm": 8.633212801204584, + "learning_rate": 2.7416717480491922e-06, + "loss": 0.1525, + "step": 2676 + }, + { + "epoch": 1.9255529581010609, + "grad_norm": 4.549930590222354, + "learning_rate": 2.741480821447519e-06, + "loss": 0.0995, + "step": 2677 + }, + { + "epoch": 1.926272253191872, + "grad_norm": 4.193951764747331, + "learning_rate": 2.7412898309694776e-06, + "loss": 0.1787, + "step": 2678 + }, + { + "epoch": 1.926991548282683, + "grad_norm": 3.4876358720109097, + "learning_rate": 2.741098776624896e-06, + "loss": 0.1067, + "step": 2679 + }, + { + "epoch": 1.927710843373494, + "grad_norm": 1.5812705732780352, + "learning_rate": 2.7409076584236032e-06, + "loss": 0.0281, + "step": 2680 + }, + { + "epoch": 1.928430138464305, + "grad_norm": 6.508487758473994, + "learning_rate": 2.740716476375433e-06, + "loss": 0.0684, + "step": 2681 + }, + { + "epoch": 1.929149433555116, + "grad_norm": 7.090071138435471, + "learning_rate": 2.7405252304902225e-06, + "loss": 0.2076, + "step": 2682 + }, + { + "epoch": 1.929868728645927, + "grad_norm": 9.593967799666707, + "learning_rate": 2.7403339207778105e-06, + "loss": 0.3193, + "step": 2683 + }, + { + "epoch": 1.930588023736738, + "grad_norm": 8.763926929831355, + "learning_rate": 2.7401425472480414e-06, + "loss": 0.4113, + "step": 2684 + }, + { + "epoch": 1.9313073188275491, + "grad_norm": 2.8955824580313805, + "learning_rate": 2.739951109910761e-06, + "loss": 0.1063, + "step": 2685 + }, + { + "epoch": 1.9320266139183602, + "grad_norm": 0.44832818791809065, + "learning_rate": 2.73975960877582e-06, + "loss": 0.0024, + "step": 2686 + }, + { + "epoch": 1.932745909009171, + "grad_norm": 4.736195941011606, + "learning_rate": 2.73956804385307e-06, + "loss": 0.232, + "step": 2687 + }, + { + "epoch": 1.933465204099982, + "grad_norm": 0.9406893099653586, + "learning_rate": 2.7393764151523686e-06, + "loss": 0.003, + "step": 2688 + }, + { + "epoch": 1.934184499190793, + "grad_norm": 4.651353938025924, + "learning_rate": 2.739184722683575e-06, + "loss": 0.1075, + "step": 2689 + }, + { + "epoch": 1.9349037942816039, + "grad_norm": 3.26245172926038, + "learning_rate": 2.738992966456552e-06, + "loss": 0.159, + "step": 2690 + }, + { + "epoch": 1.935623089372415, + "grad_norm": 4.901951626246722, + "learning_rate": 2.738801146481166e-06, + "loss": 0.0288, + "step": 2691 + }, + { + "epoch": 1.936342384463226, + "grad_norm": 8.296549801205924, + "learning_rate": 2.738609262767286e-06, + "loss": 0.1843, + "step": 2692 + }, + { + "epoch": 1.937061679554037, + "grad_norm": 3.739199107182429, + "learning_rate": 2.7384173153247856e-06, + "loss": 0.1029, + "step": 2693 + }, + { + "epoch": 1.937780974644848, + "grad_norm": 7.113141510342706, + "learning_rate": 2.7382253041635402e-06, + "loss": 0.0925, + "step": 2694 + }, + { + "epoch": 1.938500269735659, + "grad_norm": 2.8530266938887974, + "learning_rate": 2.738033229293429e-06, + "loss": 0.0777, + "step": 2695 + }, + { + "epoch": 1.93921956482647, + "grad_norm": 3.2066562835307817, + "learning_rate": 2.737841090724335e-06, + "loss": 0.0451, + "step": 2696 + }, + { + "epoch": 1.939938859917281, + "grad_norm": 5.885262123716507, + "learning_rate": 2.737648888466144e-06, + "loss": 0.0783, + "step": 2697 + }, + { + "epoch": 1.9406581550080921, + "grad_norm": 4.174928963202971, + "learning_rate": 2.7374566225287447e-06, + "loss": 0.2098, + "step": 2698 + }, + { + "epoch": 1.9413774500989032, + "grad_norm": 2.413024916685747, + "learning_rate": 2.7372642929220306e-06, + "loss": 0.0104, + "step": 2699 + }, + { + "epoch": 1.9420967451897142, + "grad_norm": 3.432784003167446, + "learning_rate": 2.737071899655896e-06, + "loss": 0.0233, + "step": 2700 + }, + { + "epoch": 1.9428160402805252, + "grad_norm": 1.1351395875892847, + "learning_rate": 2.736879442740241e-06, + "loss": 0.0122, + "step": 2701 + }, + { + "epoch": 1.943535335371336, + "grad_norm": 3.989611278976564, + "learning_rate": 2.736686922184967e-06, + "loss": 0.0803, + "step": 2702 + }, + { + "epoch": 1.944254630462147, + "grad_norm": 0.875615250929831, + "learning_rate": 2.73649433799998e-06, + "loss": 0.0217, + "step": 2703 + }, + { + "epoch": 1.9449739255529581, + "grad_norm": 2.7863064472305306, + "learning_rate": 2.7363016901951894e-06, + "loss": 0.0119, + "step": 2704 + }, + { + "epoch": 1.9456932206437692, + "grad_norm": 4.530716452232525, + "learning_rate": 2.736108978780506e-06, + "loss": 0.1577, + "step": 2705 + }, + { + "epoch": 1.94641251573458, + "grad_norm": 4.632707994491346, + "learning_rate": 2.7359162037658456e-06, + "loss": 0.1425, + "step": 2706 + }, + { + "epoch": 1.947131810825391, + "grad_norm": 2.108005426848352, + "learning_rate": 2.735723365161127e-06, + "loss": 0.0605, + "step": 2707 + }, + { + "epoch": 1.947851105916202, + "grad_norm": 3.829371895094149, + "learning_rate": 2.735530462976272e-06, + "loss": 0.1068, + "step": 2708 + }, + { + "epoch": 1.948570401007013, + "grad_norm": 7.11910574314732, + "learning_rate": 2.735337497221206e-06, + "loss": 0.0367, + "step": 2709 + }, + { + "epoch": 1.949289696097824, + "grad_norm": 0.1332685536008957, + "learning_rate": 2.735144467905857e-06, + "loss": 0.0006, + "step": 2710 + }, + { + "epoch": 1.9500089911886351, + "grad_norm": 3.9452746036807538, + "learning_rate": 2.7349513750401574e-06, + "loss": 0.071, + "step": 2711 + }, + { + "epoch": 1.9507282862794462, + "grad_norm": 1.879051716352009, + "learning_rate": 2.734758218634041e-06, + "loss": 0.0093, + "step": 2712 + }, + { + "epoch": 1.9514475813702572, + "grad_norm": 5.049579027171884, + "learning_rate": 2.734564998697447e-06, + "loss": 0.1577, + "step": 2713 + }, + { + "epoch": 1.9521668764610682, + "grad_norm": 1.6533844928468322, + "learning_rate": 2.7343717152403164e-06, + "loss": 0.0645, + "step": 2714 + }, + { + "epoch": 1.9528861715518793, + "grad_norm": 3.422795443349315, + "learning_rate": 2.7341783682725946e-06, + "loss": 0.099, + "step": 2715 + }, + { + "epoch": 1.9536054666426903, + "grad_norm": 3.5319280093702177, + "learning_rate": 2.733984957804229e-06, + "loss": 0.0884, + "step": 2716 + }, + { + "epoch": 1.9543247617335011, + "grad_norm": 3.324243580628092, + "learning_rate": 2.733791483845171e-06, + "loss": 0.0941, + "step": 2717 + }, + { + "epoch": 1.9550440568243121, + "grad_norm": 3.338605218176604, + "learning_rate": 2.7335979464053755e-06, + "loss": 0.0484, + "step": 2718 + }, + { + "epoch": 1.9557633519151232, + "grad_norm": 3.7113958770767446, + "learning_rate": 2.7334043454948e-06, + "loss": 0.0239, + "step": 2719 + }, + { + "epoch": 1.9564826470059342, + "grad_norm": 3.779995322362371, + "learning_rate": 2.733210681123406e-06, + "loss": 0.038, + "step": 2720 + }, + { + "epoch": 1.957201942096745, + "grad_norm": 5.2104804683943, + "learning_rate": 2.733016953301157e-06, + "loss": 0.1087, + "step": 2721 + }, + { + "epoch": 1.957921237187556, + "grad_norm": 5.8591328096040876, + "learning_rate": 2.732823162038022e-06, + "loss": 0.035, + "step": 2722 + }, + { + "epoch": 1.958640532278367, + "grad_norm": 2.569942583080411, + "learning_rate": 2.7326293073439703e-06, + "loss": 0.0538, + "step": 2723 + }, + { + "epoch": 1.9593598273691781, + "grad_norm": 5.764905833702417, + "learning_rate": 2.732435389228978e-06, + "loss": 0.132, + "step": 2724 + }, + { + "epoch": 1.9600791224599892, + "grad_norm": 2.434867160531595, + "learning_rate": 2.7322414077030207e-06, + "loss": 0.0145, + "step": 2725 + }, + { + "epoch": 1.9607984175508002, + "grad_norm": 3.100274728637934, + "learning_rate": 2.73204736277608e-06, + "loss": 0.044, + "step": 2726 + }, + { + "epoch": 1.9615177126416112, + "grad_norm": 4.48620272563212, + "learning_rate": 2.7318532544581395e-06, + "loss": 0.1496, + "step": 2727 + }, + { + "epoch": 1.9622370077324223, + "grad_norm": 2.800612765615511, + "learning_rate": 2.731659082759187e-06, + "loss": 0.0753, + "step": 2728 + }, + { + "epoch": 1.9629563028232333, + "grad_norm": 2.932305866401869, + "learning_rate": 2.7314648476892126e-06, + "loss": 0.0116, + "step": 2729 + }, + { + "epoch": 1.9636755979140443, + "grad_norm": 6.408723630327014, + "learning_rate": 2.7312705492582097e-06, + "loss": 0.1096, + "step": 2730 + }, + { + "epoch": 1.9643948930048554, + "grad_norm": 4.255124917951107, + "learning_rate": 2.7310761874761758e-06, + "loss": 0.1877, + "step": 2731 + }, + { + "epoch": 1.9651141880956664, + "grad_norm": 3.2247639669575157, + "learning_rate": 2.730881762353111e-06, + "loss": 0.0845, + "step": 2732 + }, + { + "epoch": 1.9658334831864772, + "grad_norm": 3.9895061821278, + "learning_rate": 2.7306872738990186e-06, + "loss": 0.0611, + "step": 2733 + }, + { + "epoch": 1.9665527782772882, + "grad_norm": 3.089948161774254, + "learning_rate": 2.730492722123906e-06, + "loss": 0.1164, + "step": 2734 + }, + { + "epoch": 1.9672720733680993, + "grad_norm": 4.254484432369005, + "learning_rate": 2.7302981070377823e-06, + "loss": 0.0134, + "step": 2735 + }, + { + "epoch": 1.9679913684589103, + "grad_norm": 3.6114213810200524, + "learning_rate": 2.7301034286506616e-06, + "loss": 0.1131, + "step": 2736 + }, + { + "epoch": 1.9687106635497211, + "grad_norm": 2.9160693307549526, + "learning_rate": 2.7299086869725597e-06, + "loss": 0.056, + "step": 2737 + }, + { + "epoch": 1.9694299586405322, + "grad_norm": 3.7748429195859763, + "learning_rate": 2.729713882013497e-06, + "loss": 0.0821, + "step": 2738 + }, + { + "epoch": 1.9701492537313432, + "grad_norm": 2.5399610012448925, + "learning_rate": 2.7295190137834967e-06, + "loss": 0.0503, + "step": 2739 + }, + { + "epoch": 1.9708685488221542, + "grad_norm": 5.202703122086643, + "learning_rate": 2.729324082292585e-06, + "loss": 0.1298, + "step": 2740 + }, + { + "epoch": 1.9715878439129653, + "grad_norm": 2.4220772728369924, + "learning_rate": 2.7291290875507904e-06, + "loss": 0.0243, + "step": 2741 + }, + { + "epoch": 1.9723071390037763, + "grad_norm": 0.6595968108214497, + "learning_rate": 2.728934029568147e-06, + "loss": 0.0025, + "step": 2742 + }, + { + "epoch": 1.9730264340945873, + "grad_norm": 4.125301787334915, + "learning_rate": 2.72873890835469e-06, + "loss": 0.0697, + "step": 2743 + }, + { + "epoch": 1.9737457291853984, + "grad_norm": 2.7893690605429655, + "learning_rate": 2.7285437239204594e-06, + "loss": 0.0174, + "step": 2744 + }, + { + "epoch": 1.9744650242762094, + "grad_norm": 4.842528841922146, + "learning_rate": 2.7283484762754986e-06, + "loss": 0.0658, + "step": 2745 + }, + { + "epoch": 1.9751843193670204, + "grad_norm": 1.624597495959097, + "learning_rate": 2.7281531654298512e-06, + "loss": 0.0072, + "step": 2746 + }, + { + "epoch": 1.9759036144578315, + "grad_norm": 3.007118944458473, + "learning_rate": 2.727957791393568e-06, + "loss": 0.0979, + "step": 2747 + }, + { + "epoch": 1.9766229095486423, + "grad_norm": 7.194295187142598, + "learning_rate": 2.7277623541767004e-06, + "loss": 0.2966, + "step": 2748 + }, + { + "epoch": 1.9773422046394533, + "grad_norm": 3.6850712655862625, + "learning_rate": 2.7275668537893046e-06, + "loss": 0.0186, + "step": 2749 + }, + { + "epoch": 1.9780614997302643, + "grad_norm": 9.269940023444025, + "learning_rate": 2.7273712902414397e-06, + "loss": 0.1964, + "step": 2750 + }, + { + "epoch": 1.9787807948210754, + "grad_norm": 1.9688814697774948, + "learning_rate": 2.727175663543167e-06, + "loss": 0.0509, + "step": 2751 + }, + { + "epoch": 1.9795000899118862, + "grad_norm": 4.955911007828335, + "learning_rate": 2.726979973704552e-06, + "loss": 0.0855, + "step": 2752 + }, + { + "epoch": 1.9802193850026972, + "grad_norm": 5.2816354592967345, + "learning_rate": 2.7267842207356636e-06, + "loss": 0.0273, + "step": 2753 + }, + { + "epoch": 1.9809386800935083, + "grad_norm": 3.956057839838247, + "learning_rate": 2.7265884046465734e-06, + "loss": 0.1666, + "step": 2754 + }, + { + "epoch": 1.9816579751843193, + "grad_norm": 1.6072058821572746, + "learning_rate": 2.726392525447357e-06, + "loss": 0.0213, + "step": 2755 + }, + { + "epoch": 1.9823772702751303, + "grad_norm": 6.4775060702831695, + "learning_rate": 2.7261965831480917e-06, + "loss": 0.1527, + "step": 2756 + }, + { + "epoch": 1.9830965653659414, + "grad_norm": 4.192038209219215, + "learning_rate": 2.72600057775886e-06, + "loss": 0.0996, + "step": 2757 + }, + { + "epoch": 1.9838158604567524, + "grad_norm": 4.860116427408447, + "learning_rate": 2.725804509289746e-06, + "loss": 0.1071, + "step": 2758 + }, + { + "epoch": 1.9845351555475634, + "grad_norm": 2.3987438825732754, + "learning_rate": 2.725608377750839e-06, + "loss": 0.0711, + "step": 2759 + }, + { + "epoch": 1.9852544506383745, + "grad_norm": 1.7101575941302474, + "learning_rate": 2.725412183152229e-06, + "loss": 0.0082, + "step": 2760 + }, + { + "epoch": 1.9859737457291855, + "grad_norm": 2.3136104429305875, + "learning_rate": 2.7252159255040107e-06, + "loss": 0.0102, + "step": 2761 + }, + { + "epoch": 1.9866930408199965, + "grad_norm": 2.989112624623687, + "learning_rate": 2.7250196048162824e-06, + "loss": 0.095, + "step": 2762 + }, + { + "epoch": 1.9874123359108076, + "grad_norm": 4.803505387877456, + "learning_rate": 2.724823221099145e-06, + "loss": 0.1737, + "step": 2763 + }, + { + "epoch": 1.9881316310016184, + "grad_norm": 3.5616184297518543, + "learning_rate": 2.724626774362703e-06, + "loss": 0.1122, + "step": 2764 + }, + { + "epoch": 1.9888509260924294, + "grad_norm": 3.99910606016285, + "learning_rate": 2.7244302646170637e-06, + "loss": 0.0485, + "step": 2765 + }, + { + "epoch": 1.9895702211832405, + "grad_norm": 5.244362464451042, + "learning_rate": 2.7242336918723377e-06, + "loss": 0.0535, + "step": 2766 + }, + { + "epoch": 1.9902895162740515, + "grad_norm": 3.1386017639220856, + "learning_rate": 2.7240370561386392e-06, + "loss": 0.0881, + "step": 2767 + }, + { + "epoch": 1.9910088113648623, + "grad_norm": 3.2222745840776916, + "learning_rate": 2.7238403574260854e-06, + "loss": 0.0747, + "step": 2768 + }, + { + "epoch": 1.9917281064556733, + "grad_norm": 3.691132173757284, + "learning_rate": 2.723643595744797e-06, + "loss": 0.2281, + "step": 2769 + }, + { + "epoch": 1.9924474015464844, + "grad_norm": 3.003509620936156, + "learning_rate": 2.7234467711048974e-06, + "loss": 0.0048, + "step": 2770 + }, + { + "epoch": 1.9931666966372954, + "grad_norm": 3.9822283598442154, + "learning_rate": 2.7232498835165137e-06, + "loss": 0.0562, + "step": 2771 + }, + { + "epoch": 1.9938859917281064, + "grad_norm": 5.5085108815195944, + "learning_rate": 2.723052932989776e-06, + "loss": 0.0914, + "step": 2772 + }, + { + "epoch": 1.9946052868189175, + "grad_norm": 4.905355923487537, + "learning_rate": 2.7228559195348184e-06, + "loss": 0.193, + "step": 2773 + }, + { + "epoch": 1.9953245819097285, + "grad_norm": 2.6997842587859164, + "learning_rate": 2.7226588431617766e-06, + "loss": 0.1181, + "step": 2774 + }, + { + "epoch": 1.9960438770005395, + "grad_norm": 3.955298846710558, + "learning_rate": 2.7224617038807913e-06, + "loss": 0.1458, + "step": 2775 + }, + { + "epoch": 1.9967631720913506, + "grad_norm": 3.2238497763550464, + "learning_rate": 2.7222645017020053e-06, + "loss": 0.1708, + "step": 2776 + }, + { + "epoch": 1.9974824671821616, + "grad_norm": 7.317171371890231, + "learning_rate": 2.722067236635565e-06, + "loss": 0.074, + "step": 2777 + }, + { + "epoch": 1.9982017622729726, + "grad_norm": 4.054485450254262, + "learning_rate": 2.7218699086916204e-06, + "loss": 0.0595, + "step": 2778 + }, + { + "epoch": 1.9989210573637834, + "grad_norm": 2.029654451210026, + "learning_rate": 2.7216725178803233e-06, + "loss": 0.0104, + "step": 2779 + }, + { + "epoch": 1.9996403524545945, + "grad_norm": 4.55743599142024, + "learning_rate": 2.7214750642118314e-06, + "loss": 0.1304, + "step": 2780 + }, + { + "epoch": 2.0003596475454053, + "grad_norm": 0.6247991836602562, + "learning_rate": 2.721277547696303e-06, + "loss": 0.0013, + "step": 2781 + }, + { + "epoch": 2.0010789426362163, + "grad_norm": 2.219010924610205, + "learning_rate": 2.721079968343901e-06, + "loss": 0.0119, + "step": 2782 + }, + { + "epoch": 2.0017982377270274, + "grad_norm": 0.9012996985022022, + "learning_rate": 2.720882326164791e-06, + "loss": 0.0011, + "step": 2783 + }, + { + "epoch": 2.0025175328178384, + "grad_norm": 4.030547548310015, + "learning_rate": 2.7206846211691415e-06, + "loss": 0.0817, + "step": 2784 + }, + { + "epoch": 2.0032368279086494, + "grad_norm": 5.619457071778356, + "learning_rate": 2.7204868533671258e-06, + "loss": 0.0245, + "step": 2785 + }, + { + "epoch": 2.0039561229994605, + "grad_norm": 4.47418191124176, + "learning_rate": 2.7202890227689194e-06, + "loss": 0.0662, + "step": 2786 + }, + { + "epoch": 2.0046754180902715, + "grad_norm": 2.9616917963504137, + "learning_rate": 2.7200911293847007e-06, + "loss": 0.0271, + "step": 2787 + }, + { + "epoch": 2.0053947131810825, + "grad_norm": 2.066912038355341, + "learning_rate": 2.719893173224651e-06, + "loss": 0.0604, + "step": 2788 + }, + { + "epoch": 2.0061140082718936, + "grad_norm": 4.43492987418814, + "learning_rate": 2.719695154298956e-06, + "loss": 0.1186, + "step": 2789 + }, + { + "epoch": 2.0068333033627046, + "grad_norm": 3.814897479505931, + "learning_rate": 2.7194970726178046e-06, + "loss": 0.1258, + "step": 2790 + }, + { + "epoch": 2.0075525984535156, + "grad_norm": 3.0465531715652094, + "learning_rate": 2.7192989281913882e-06, + "loss": 0.0639, + "step": 2791 + }, + { + "epoch": 2.0082718935443267, + "grad_norm": 4.155592607552763, + "learning_rate": 2.7191007210299014e-06, + "loss": 0.018, + "step": 2792 + }, + { + "epoch": 2.0089911886351377, + "grad_norm": 3.6127933444386837, + "learning_rate": 2.718902451143542e-06, + "loss": 0.114, + "step": 2793 + }, + { + "epoch": 2.0097104837259487, + "grad_norm": 3.052211247299576, + "learning_rate": 2.718704118542512e-06, + "loss": 0.1391, + "step": 2794 + }, + { + "epoch": 2.0104297788167598, + "grad_norm": 1.9941958659627657, + "learning_rate": 2.718505723237016e-06, + "loss": 0.0746, + "step": 2795 + }, + { + "epoch": 2.0111490739075704, + "grad_norm": 6.271733071387418, + "learning_rate": 2.7183072652372614e-06, + "loss": 0.241, + "step": 2796 + }, + { + "epoch": 2.0118683689983814, + "grad_norm": 4.5760518934568, + "learning_rate": 2.7181087445534596e-06, + "loss": 0.0366, + "step": 2797 + }, + { + "epoch": 2.0125876640891924, + "grad_norm": 4.454022890322071, + "learning_rate": 2.717910161195824e-06, + "loss": 0.076, + "step": 2798 + }, + { + "epoch": 2.0133069591800035, + "grad_norm": 0.37003568008588267, + "learning_rate": 2.717711515174573e-06, + "loss": 0.0011, + "step": 2799 + }, + { + "epoch": 2.0140262542708145, + "grad_norm": 3.085489732709743, + "learning_rate": 2.717512806499927e-06, + "loss": 0.1056, + "step": 2800 + }, + { + "epoch": 2.0147455493616255, + "grad_norm": 5.079153035692809, + "learning_rate": 2.7173140351821095e-06, + "loss": 0.2155, + "step": 2801 + }, + { + "epoch": 2.0154648444524366, + "grad_norm": 1.619421015483736, + "learning_rate": 2.717115201231348e-06, + "loss": 0.0067, + "step": 2802 + }, + { + "epoch": 2.0161841395432476, + "grad_norm": 3.1639939847466043, + "learning_rate": 2.716916304657873e-06, + "loss": 0.0518, + "step": 2803 + }, + { + "epoch": 2.0169034346340586, + "grad_norm": 4.318803062561228, + "learning_rate": 2.716717345471918e-06, + "loss": 0.0399, + "step": 2804 + }, + { + "epoch": 2.0176227297248697, + "grad_norm": 2.268858526030059, + "learning_rate": 2.7165183236837197e-06, + "loss": 0.0135, + "step": 2805 + }, + { + "epoch": 2.0183420248156807, + "grad_norm": 1.546237046231724, + "learning_rate": 2.7163192393035175e-06, + "loss": 0.0062, + "step": 2806 + }, + { + "epoch": 2.0190613199064917, + "grad_norm": 1.9073589745711095, + "learning_rate": 2.716120092341556e-06, + "loss": 0.0271, + "step": 2807 + }, + { + "epoch": 2.0197806149973028, + "grad_norm": 8.484870302892674, + "learning_rate": 2.7159208828080803e-06, + "loss": 0.1617, + "step": 2808 + }, + { + "epoch": 2.020499910088114, + "grad_norm": 3.8386481781464195, + "learning_rate": 2.7157216107133414e-06, + "loss": 0.0886, + "step": 2809 + }, + { + "epoch": 2.021219205178925, + "grad_norm": 4.467874712096741, + "learning_rate": 2.715522276067591e-06, + "loss": 0.0824, + "step": 2810 + }, + { + "epoch": 2.021938500269736, + "grad_norm": 3.659386606493036, + "learning_rate": 2.7153228788810856e-06, + "loss": 0.0457, + "step": 2811 + }, + { + "epoch": 2.0226577953605465, + "grad_norm": 5.398311847914288, + "learning_rate": 2.715123419164085e-06, + "loss": 0.0655, + "step": 2812 + }, + { + "epoch": 2.0233770904513575, + "grad_norm": 7.430266630975111, + "learning_rate": 2.7149238969268513e-06, + "loss": 0.0211, + "step": 2813 + }, + { + "epoch": 2.0240963855421685, + "grad_norm": 6.480968913700203, + "learning_rate": 2.7147243121796503e-06, + "loss": 0.2189, + "step": 2814 + }, + { + "epoch": 2.0248156806329796, + "grad_norm": 5.836683650324344, + "learning_rate": 2.714524664932751e-06, + "loss": 0.17, + "step": 2815 + }, + { + "epoch": 2.0255349757237906, + "grad_norm": 3.9349897314803925, + "learning_rate": 2.714324955196426e-06, + "loss": 0.1472, + "step": 2816 + }, + { + "epoch": 2.0262542708146016, + "grad_norm": 2.26787486431004, + "learning_rate": 2.71412518298095e-06, + "loss": 0.0173, + "step": 2817 + }, + { + "epoch": 2.0269735659054127, + "grad_norm": 1.8047415051231606, + "learning_rate": 2.713925348296602e-06, + "loss": 0.0304, + "step": 2818 + }, + { + "epoch": 2.0276928609962237, + "grad_norm": 4.461285807801138, + "learning_rate": 2.713725451153664e-06, + "loss": 0.1124, + "step": 2819 + }, + { + "epoch": 2.0284121560870347, + "grad_norm": 2.7188520446413444, + "learning_rate": 2.713525491562421e-06, + "loss": 0.0862, + "step": 2820 + }, + { + "epoch": 2.0291314511778458, + "grad_norm": 1.0976615378646268, + "learning_rate": 2.7133254695331614e-06, + "loss": 0.0028, + "step": 2821 + }, + { + "epoch": 2.029850746268657, + "grad_norm": 4.15485904843733, + "learning_rate": 2.713125385076176e-06, + "loss": 0.1266, + "step": 2822 + }, + { + "epoch": 2.030570041359468, + "grad_norm": 1.5880528038831108, + "learning_rate": 2.71292523820176e-06, + "loss": 0.0241, + "step": 2823 + }, + { + "epoch": 2.031289336450279, + "grad_norm": 4.33536997293647, + "learning_rate": 2.7127250289202115e-06, + "loss": 0.0895, + "step": 2824 + }, + { + "epoch": 2.03200863154109, + "grad_norm": 2.06688137916887, + "learning_rate": 2.7125247572418315e-06, + "loss": 0.0084, + "step": 2825 + }, + { + "epoch": 2.032727926631901, + "grad_norm": 2.2942832983499306, + "learning_rate": 2.712324423176924e-06, + "loss": 0.1034, + "step": 2826 + }, + { + "epoch": 2.0334472217227115, + "grad_norm": 5.819125177481963, + "learning_rate": 2.7121240267357967e-06, + "loss": 0.1182, + "step": 2827 + }, + { + "epoch": 2.0341665168135226, + "grad_norm": 1.4806506937539476, + "learning_rate": 2.7119235679287605e-06, + "loss": 0.0106, + "step": 2828 + }, + { + "epoch": 2.0348858119043336, + "grad_norm": 3.448025879752963, + "learning_rate": 2.711723046766129e-06, + "loss": 0.1303, + "step": 2829 + }, + { + "epoch": 2.0356051069951446, + "grad_norm": 1.3952158591504202, + "learning_rate": 2.71152246325822e-06, + "loss": 0.0207, + "step": 2830 + }, + { + "epoch": 2.0363244020859557, + "grad_norm": 5.444168632996188, + "learning_rate": 2.7113218174153537e-06, + "loss": 0.3073, + "step": 2831 + }, + { + "epoch": 2.0370436971767667, + "grad_norm": 0.14657342091201225, + "learning_rate": 2.711121109247853e-06, + "loss": 0.0005, + "step": 2832 + }, + { + "epoch": 2.0377629922675777, + "grad_norm": 2.1574770295331027, + "learning_rate": 2.710920338766045e-06, + "loss": 0.0504, + "step": 2833 + }, + { + "epoch": 2.0384822873583888, + "grad_norm": 3.115014121264926, + "learning_rate": 2.7107195059802603e-06, + "loss": 0.0424, + "step": 2834 + }, + { + "epoch": 2.0392015824492, + "grad_norm": 3.994109904266133, + "learning_rate": 2.710518610900832e-06, + "loss": 0.0948, + "step": 2835 + }, + { + "epoch": 2.039920877540011, + "grad_norm": 7.230308004929196, + "learning_rate": 2.710317653538095e-06, + "loss": 0.0622, + "step": 2836 + }, + { + "epoch": 2.040640172630822, + "grad_norm": 7.943485615827747, + "learning_rate": 2.710116633902391e-06, + "loss": 0.1657, + "step": 2837 + }, + { + "epoch": 2.041359467721633, + "grad_norm": 2.4278652976648223, + "learning_rate": 2.709915552004062e-06, + "loss": 0.0511, + "step": 2838 + }, + { + "epoch": 2.042078762812444, + "grad_norm": 3.966996139087217, + "learning_rate": 2.709714407853454e-06, + "loss": 0.1563, + "step": 2839 + }, + { + "epoch": 2.042798057903255, + "grad_norm": 4.267452681601579, + "learning_rate": 2.709513201460915e-06, + "loss": 0.0514, + "step": 2840 + }, + { + "epoch": 2.043517352994066, + "grad_norm": 2.0192790206568336, + "learning_rate": 2.7093119328367997e-06, + "loss": 0.0546, + "step": 2841 + }, + { + "epoch": 2.044236648084877, + "grad_norm": 2.912353301984279, + "learning_rate": 2.709110601991462e-06, + "loss": 0.0456, + "step": 2842 + }, + { + "epoch": 2.0449559431756876, + "grad_norm": 4.978384529183961, + "learning_rate": 2.708909208935262e-06, + "loss": 0.1326, + "step": 2843 + }, + { + "epoch": 2.0456752382664987, + "grad_norm": 3.331634584450944, + "learning_rate": 2.70870775367856e-06, + "loss": 0.1528, + "step": 2844 + }, + { + "epoch": 2.0463945333573097, + "grad_norm": 4.449390741492378, + "learning_rate": 2.708506236231723e-06, + "loss": 0.1946, + "step": 2845 + }, + { + "epoch": 2.0471138284481207, + "grad_norm": 5.246598072230538, + "learning_rate": 2.7083046566051186e-06, + "loss": 0.1447, + "step": 2846 + }, + { + "epoch": 2.0478331235389318, + "grad_norm": 2.644773198980077, + "learning_rate": 2.7081030148091177e-06, + "loss": 0.076, + "step": 2847 + }, + { + "epoch": 2.048552418629743, + "grad_norm": 7.170148536840793, + "learning_rate": 2.707901310854097e-06, + "loss": 0.069, + "step": 2848 + }, + { + "epoch": 2.049271713720554, + "grad_norm": 6.145849778851321, + "learning_rate": 2.707699544750433e-06, + "loss": 0.132, + "step": 2849 + }, + { + "epoch": 2.049991008811365, + "grad_norm": 2.543005988948136, + "learning_rate": 2.7074977165085074e-06, + "loss": 0.0803, + "step": 2850 + }, + { + "epoch": 2.050710303902176, + "grad_norm": 1.735885549214208, + "learning_rate": 2.7072958261387044e-06, + "loss": 0.0249, + "step": 2851 + }, + { + "epoch": 2.051429598992987, + "grad_norm": 2.588109804804989, + "learning_rate": 2.707093873651412e-06, + "loss": 0.0765, + "step": 2852 + }, + { + "epoch": 2.052148894083798, + "grad_norm": 2.663709577313606, + "learning_rate": 2.706891859057021e-06, + "loss": 0.0889, + "step": 2853 + }, + { + "epoch": 2.052868189174609, + "grad_norm": 5.4433859508191516, + "learning_rate": 2.706689782365925e-06, + "loss": 0.1637, + "step": 2854 + }, + { + "epoch": 2.05358748426542, + "grad_norm": 6.236158105254232, + "learning_rate": 2.7064876435885215e-06, + "loss": 0.0482, + "step": 2855 + }, + { + "epoch": 2.054306779356231, + "grad_norm": 6.520916861143226, + "learning_rate": 2.7062854427352108e-06, + "loss": 0.0675, + "step": 2856 + }, + { + "epoch": 2.055026074447042, + "grad_norm": 2.9841237110257652, + "learning_rate": 2.7060831798163965e-06, + "loss": 0.0272, + "step": 2857 + }, + { + "epoch": 2.0557453695378527, + "grad_norm": 3.461535408529019, + "learning_rate": 2.7058808548424852e-06, + "loss": 0.1067, + "step": 2858 + }, + { + "epoch": 2.0564646646286637, + "grad_norm": 7.1922968820156745, + "learning_rate": 2.705678467823887e-06, + "loss": 0.16, + "step": 2859 + }, + { + "epoch": 2.0571839597194748, + "grad_norm": 2.802629225891705, + "learning_rate": 2.7054760187710158e-06, + "loss": 0.0514, + "step": 2860 + }, + { + "epoch": 2.057903254810286, + "grad_norm": 2.645477453293368, + "learning_rate": 2.7052735076942865e-06, + "loss": 0.0393, + "step": 2861 + }, + { + "epoch": 2.058622549901097, + "grad_norm": 3.9603260287374034, + "learning_rate": 2.70507093460412e-06, + "loss": 0.0168, + "step": 2862 + }, + { + "epoch": 2.059341844991908, + "grad_norm": 6.664075058381859, + "learning_rate": 2.7048682995109383e-06, + "loss": 0.1383, + "step": 2863 + }, + { + "epoch": 2.060061140082719, + "grad_norm": 4.179636360368768, + "learning_rate": 2.7046656024251674e-06, + "loss": 0.0505, + "step": 2864 + }, + { + "epoch": 2.06078043517353, + "grad_norm": 1.6382320499211176, + "learning_rate": 2.7044628433572367e-06, + "loss": 0.0414, + "step": 2865 + }, + { + "epoch": 2.061499730264341, + "grad_norm": 4.204107340462218, + "learning_rate": 2.704260022317578e-06, + "loss": 0.0476, + "step": 2866 + }, + { + "epoch": 2.062219025355152, + "grad_norm": 2.9783997258046764, + "learning_rate": 2.7040571393166274e-06, + "loss": 0.0469, + "step": 2867 + }, + { + "epoch": 2.062938320445963, + "grad_norm": 6.045178140926529, + "learning_rate": 2.703854194364823e-06, + "loss": 0.1211, + "step": 2868 + }, + { + "epoch": 2.063657615536774, + "grad_norm": 0.20845555049375436, + "learning_rate": 2.7036511874726077e-06, + "loss": 0.0003, + "step": 2869 + }, + { + "epoch": 2.064376910627585, + "grad_norm": 5.056599188113428, + "learning_rate": 2.7034481186504253e-06, + "loss": 0.0934, + "step": 2870 + }, + { + "epoch": 2.065096205718396, + "grad_norm": 5.7650915892466745, + "learning_rate": 2.703244987908725e-06, + "loss": 0.1872, + "step": 2871 + }, + { + "epoch": 2.065815500809207, + "grad_norm": 2.3393772680989757, + "learning_rate": 2.7030417952579574e-06, + "loss": 0.0714, + "step": 2872 + }, + { + "epoch": 2.0665347959000178, + "grad_norm": 3.6000945563863676, + "learning_rate": 2.7028385407085776e-06, + "loss": 0.0805, + "step": 2873 + }, + { + "epoch": 2.067254090990829, + "grad_norm": 5.535896920535353, + "learning_rate": 2.7026352242710434e-06, + "loss": 0.0992, + "step": 2874 + }, + { + "epoch": 2.06797338608164, + "grad_norm": 1.7338699960326092, + "learning_rate": 2.702431845955816e-06, + "loss": 0.012, + "step": 2875 + }, + { + "epoch": 2.068692681172451, + "grad_norm": 6.501831485576869, + "learning_rate": 2.70222840577336e-06, + "loss": 0.2113, + "step": 2876 + }, + { + "epoch": 2.069411976263262, + "grad_norm": 7.530651501944735, + "learning_rate": 2.7020249037341408e-06, + "loss": 0.0757, + "step": 2877 + }, + { + "epoch": 2.070131271354073, + "grad_norm": 5.358410037982092, + "learning_rate": 2.7018213398486306e-06, + "loss": 0.1405, + "step": 2878 + }, + { + "epoch": 2.070850566444884, + "grad_norm": 2.5619082870532446, + "learning_rate": 2.7016177141273025e-06, + "loss": 0.091, + "step": 2879 + }, + { + "epoch": 2.071569861535695, + "grad_norm": 0.05927888327756036, + "learning_rate": 2.7014140265806347e-06, + "loss": 0.0003, + "step": 2880 + }, + { + "epoch": 2.072289156626506, + "grad_norm": 2.657744763327062, + "learning_rate": 2.701210277219105e-06, + "loss": 0.1127, + "step": 2881 + }, + { + "epoch": 2.073008451717317, + "grad_norm": 4.855462734285935, + "learning_rate": 2.7010064660531988e-06, + "loss": 0.1165, + "step": 2882 + }, + { + "epoch": 2.073727746808128, + "grad_norm": 3.299968811696358, + "learning_rate": 2.700802593093401e-06, + "loss": 0.1322, + "step": 2883 + }, + { + "epoch": 2.074447041898939, + "grad_norm": 2.713412676740052, + "learning_rate": 2.700598658350202e-06, + "loss": 0.0076, + "step": 2884 + }, + { + "epoch": 2.07516633698975, + "grad_norm": 4.071712260648528, + "learning_rate": 2.7003946618340945e-06, + "loss": 0.1469, + "step": 2885 + }, + { + "epoch": 2.075885632080561, + "grad_norm": 1.448805845567579, + "learning_rate": 2.7001906035555747e-06, + "loss": 0.0037, + "step": 2886 + }, + { + "epoch": 2.0766049271713722, + "grad_norm": 4.710089331706898, + "learning_rate": 2.6999864835251407e-06, + "loss": 0.0261, + "step": 2887 + }, + { + "epoch": 2.0773242222621833, + "grad_norm": 3.3906585116952757, + "learning_rate": 2.6997823017532965e-06, + "loss": 0.0049, + "step": 2888 + }, + { + "epoch": 2.078043517352994, + "grad_norm": 4.974308262357309, + "learning_rate": 2.699578058250546e-06, + "loss": 0.1814, + "step": 2889 + }, + { + "epoch": 2.078762812443805, + "grad_norm": 1.9709754085523008, + "learning_rate": 2.6993737530273993e-06, + "loss": 0.0159, + "step": 2890 + }, + { + "epoch": 2.079482107534616, + "grad_norm": 3.7408653619798278, + "learning_rate": 2.6991693860943667e-06, + "loss": 0.1133, + "step": 2891 + }, + { + "epoch": 2.080201402625427, + "grad_norm": 2.935382699602907, + "learning_rate": 2.6989649574619647e-06, + "loss": 0.0218, + "step": 2892 + }, + { + "epoch": 2.080920697716238, + "grad_norm": 4.0281844640731705, + "learning_rate": 2.698760467140711e-06, + "loss": 0.0483, + "step": 2893 + }, + { + "epoch": 2.081639992807049, + "grad_norm": 4.064628009452115, + "learning_rate": 2.6985559151411264e-06, + "loss": 0.0795, + "step": 2894 + }, + { + "epoch": 2.08235928789786, + "grad_norm": 2.501698238278197, + "learning_rate": 2.6983513014737363e-06, + "loss": 0.0288, + "step": 2895 + }, + { + "epoch": 2.083078582988671, + "grad_norm": 2.6206935825263833, + "learning_rate": 2.6981466261490685e-06, + "loss": 0.0106, + "step": 2896 + }, + { + "epoch": 2.083797878079482, + "grad_norm": 2.0805132327356044, + "learning_rate": 2.697941889177653e-06, + "loss": 0.0581, + "step": 2897 + }, + { + "epoch": 2.084517173170293, + "grad_norm": 0.7818793583257235, + "learning_rate": 2.6977370905700245e-06, + "loss": 0.0117, + "step": 2898 + }, + { + "epoch": 2.085236468261104, + "grad_norm": 4.311638089214096, + "learning_rate": 2.6975322303367204e-06, + "loss": 0.1123, + "step": 2899 + }, + { + "epoch": 2.0859557633519152, + "grad_norm": 4.139080905419962, + "learning_rate": 2.69732730848828e-06, + "loss": 0.0907, + "step": 2900 + }, + { + "epoch": 2.0866750584427263, + "grad_norm": 2.214659225866576, + "learning_rate": 2.6971223250352484e-06, + "loss": 0.0411, + "step": 2901 + }, + { + "epoch": 2.0873943535335373, + "grad_norm": 2.413139243340381, + "learning_rate": 2.696917279988172e-06, + "loss": 0.0625, + "step": 2902 + }, + { + "epoch": 2.0881136486243483, + "grad_norm": 4.498754842960439, + "learning_rate": 2.6967121733576e-06, + "loss": 0.1444, + "step": 2903 + }, + { + "epoch": 2.088832943715159, + "grad_norm": 1.036526231725358, + "learning_rate": 2.6965070051540857e-06, + "loss": 0.0033, + "step": 2904 + }, + { + "epoch": 2.08955223880597, + "grad_norm": 4.776313354805104, + "learning_rate": 2.696301775388186e-06, + "loss": 0.2467, + "step": 2905 + }, + { + "epoch": 2.090271533896781, + "grad_norm": 2.52415644235071, + "learning_rate": 2.6960964840704594e-06, + "loss": 0.0637, + "step": 2906 + }, + { + "epoch": 2.090990828987592, + "grad_norm": 2.436682772121266, + "learning_rate": 2.6958911312114694e-06, + "loss": 0.0255, + "step": 2907 + }, + { + "epoch": 2.091710124078403, + "grad_norm": 1.4429705263965487, + "learning_rate": 2.6956857168217814e-06, + "loss": 0.0049, + "step": 2908 + }, + { + "epoch": 2.092429419169214, + "grad_norm": 2.859501605564474, + "learning_rate": 2.695480240911964e-06, + "loss": 0.021, + "step": 2909 + }, + { + "epoch": 2.093148714260025, + "grad_norm": 5.1537380153314665, + "learning_rate": 2.6952747034925904e-06, + "loss": 0.2112, + "step": 2910 + }, + { + "epoch": 2.093868009350836, + "grad_norm": 5.855738569645706, + "learning_rate": 2.695069104574234e-06, + "loss": 0.1484, + "step": 2911 + }, + { + "epoch": 2.094587304441647, + "grad_norm": 0.25332351452389795, + "learning_rate": 2.694863444167475e-06, + "loss": 0.0013, + "step": 2912 + }, + { + "epoch": 2.0953065995324582, + "grad_norm": 3.434302414182239, + "learning_rate": 2.6946577222828937e-06, + "loss": 0.0876, + "step": 2913 + }, + { + "epoch": 2.0960258946232693, + "grad_norm": 4.086544533580528, + "learning_rate": 2.6944519389310765e-06, + "loss": 0.1986, + "step": 2914 + }, + { + "epoch": 2.0967451897140803, + "grad_norm": 3.8506262753492093, + "learning_rate": 2.6942460941226094e-06, + "loss": 0.2298, + "step": 2915 + }, + { + "epoch": 2.0974644848048913, + "grad_norm": 3.3132662908177566, + "learning_rate": 2.6940401878680844e-06, + "loss": 0.0827, + "step": 2916 + }, + { + "epoch": 2.0981837798957024, + "grad_norm": 2.4893188167297944, + "learning_rate": 2.693834220178096e-06, + "loss": 0.0741, + "step": 2917 + }, + { + "epoch": 2.0989030749865134, + "grad_norm": 2.5766126043141147, + "learning_rate": 2.693628191063241e-06, + "loss": 0.0988, + "step": 2918 + }, + { + "epoch": 2.0996223700773244, + "grad_norm": 4.93133336756768, + "learning_rate": 2.69342210053412e-06, + "loss": 0.0295, + "step": 2919 + }, + { + "epoch": 2.100341665168135, + "grad_norm": 1.4705158230720918, + "learning_rate": 2.6932159486013372e-06, + "loss": 0.0377, + "step": 2920 + }, + { + "epoch": 2.101060960258946, + "grad_norm": 2.93231780719383, + "learning_rate": 2.693009735275499e-06, + "loss": 0.1103, + "step": 2921 + }, + { + "epoch": 2.101780255349757, + "grad_norm": 1.6751038900492348, + "learning_rate": 2.692803460567216e-06, + "loss": 0.008, + "step": 2922 + }, + { + "epoch": 2.102499550440568, + "grad_norm": 3.541810755257078, + "learning_rate": 2.6925971244871007e-06, + "loss": 0.2007, + "step": 2923 + }, + { + "epoch": 2.103218845531379, + "grad_norm": 6.288832017446734, + "learning_rate": 2.69239072704577e-06, + "loss": 0.2127, + "step": 2924 + }, + { + "epoch": 2.10393814062219, + "grad_norm": 5.576877420818762, + "learning_rate": 2.6921842682538434e-06, + "loss": 0.0836, + "step": 2925 + }, + { + "epoch": 2.1046574357130012, + "grad_norm": 3.670234693775757, + "learning_rate": 2.691977748121943e-06, + "loss": 0.0174, + "step": 2926 + }, + { + "epoch": 2.1053767308038123, + "grad_norm": 5.309282768027773, + "learning_rate": 2.6917711666606955e-06, + "loss": 0.1346, + "step": 2927 + }, + { + "epoch": 2.1060960258946233, + "grad_norm": 4.252061231370029, + "learning_rate": 2.691564523880729e-06, + "loss": 0.0807, + "step": 2928 + }, + { + "epoch": 2.1068153209854343, + "grad_norm": 1.4268997965102455, + "learning_rate": 2.6913578197926765e-06, + "loss": 0.0402, + "step": 2929 + }, + { + "epoch": 2.1075346160762454, + "grad_norm": 5.973370139577655, + "learning_rate": 2.691151054407172e-06, + "loss": 0.1377, + "step": 2930 + }, + { + "epoch": 2.1082539111670564, + "grad_norm": 0.8930848275859191, + "learning_rate": 2.6909442277348553e-06, + "loss": 0.003, + "step": 2931 + }, + { + "epoch": 2.1089732062578674, + "grad_norm": 4.657529562237361, + "learning_rate": 2.6907373397863675e-06, + "loss": 0.1874, + "step": 2932 + }, + { + "epoch": 2.1096925013486785, + "grad_norm": 2.073690725354036, + "learning_rate": 2.690530390572353e-06, + "loss": 0.0582, + "step": 2933 + }, + { + "epoch": 2.1104117964394895, + "grad_norm": 2.996825676239332, + "learning_rate": 2.69032338010346e-06, + "loss": 0.0522, + "step": 2934 + }, + { + "epoch": 2.1111310915303, + "grad_norm": 6.869905188009063, + "learning_rate": 2.6901163083903396e-06, + "loss": 0.1191, + "step": 2935 + }, + { + "epoch": 2.111850386621111, + "grad_norm": 3.2832506329399034, + "learning_rate": 2.6899091754436464e-06, + "loss": 0.057, + "step": 2936 + }, + { + "epoch": 2.112569681711922, + "grad_norm": 1.9787105990508114, + "learning_rate": 2.689701981274037e-06, + "loss": 0.0085, + "step": 2937 + }, + { + "epoch": 2.113288976802733, + "grad_norm": 1.7328058463020288, + "learning_rate": 2.689494725892172e-06, + "loss": 0.0075, + "step": 2938 + }, + { + "epoch": 2.1140082718935442, + "grad_norm": 2.8948611029461935, + "learning_rate": 2.6892874093087153e-06, + "loss": 0.1409, + "step": 2939 + }, + { + "epoch": 2.1147275669843553, + "grad_norm": 2.525295282714169, + "learning_rate": 2.689080031534334e-06, + "loss": 0.0769, + "step": 2940 + }, + { + "epoch": 2.1154468620751663, + "grad_norm": 2.3096384712137965, + "learning_rate": 2.6888725925796974e-06, + "loss": 0.0807, + "step": 2941 + }, + { + "epoch": 2.1161661571659773, + "grad_norm": 3.4804000809233635, + "learning_rate": 2.6886650924554786e-06, + "loss": 0.1043, + "step": 2942 + }, + { + "epoch": 2.1168854522567884, + "grad_norm": 1.476342280066496, + "learning_rate": 2.688457531172355e-06, + "loss": 0.0109, + "step": 2943 + }, + { + "epoch": 2.1176047473475994, + "grad_norm": 2.520978701311248, + "learning_rate": 2.688249908741005e-06, + "loss": 0.0507, + "step": 2944 + }, + { + "epoch": 2.1183240424384104, + "grad_norm": 2.189522237803197, + "learning_rate": 2.6880422251721104e-06, + "loss": 0.0458, + "step": 2945 + }, + { + "epoch": 2.1190433375292215, + "grad_norm": 0.9308178347925022, + "learning_rate": 2.6878344804763587e-06, + "loss": 0.0026, + "step": 2946 + }, + { + "epoch": 2.1197626326200325, + "grad_norm": 6.920025212568776, + "learning_rate": 2.687626674664437e-06, + "loss": 0.1925, + "step": 2947 + }, + { + "epoch": 2.1204819277108435, + "grad_norm": 4.017538052314308, + "learning_rate": 2.687418807747039e-06, + "loss": 0.1184, + "step": 2948 + }, + { + "epoch": 2.1212012228016546, + "grad_norm": 0.6984131718619034, + "learning_rate": 2.6872108797348583e-06, + "loss": 0.0073, + "step": 2949 + }, + { + "epoch": 2.121920517892465, + "grad_norm": 6.485046510514252, + "learning_rate": 2.6870028906385938e-06, + "loss": 0.0652, + "step": 2950 + }, + { + "epoch": 2.122639812983276, + "grad_norm": 3.9208748495925905, + "learning_rate": 2.686794840468947e-06, + "loss": 0.1427, + "step": 2951 + }, + { + "epoch": 2.1233591080740872, + "grad_norm": 1.018764436705655, + "learning_rate": 2.6865867292366223e-06, + "loss": 0.0015, + "step": 2952 + }, + { + "epoch": 2.1240784031648983, + "grad_norm": 4.228581355661939, + "learning_rate": 2.686378556952327e-06, + "loss": 0.1115, + "step": 2953 + }, + { + "epoch": 2.1247976982557093, + "grad_norm": 4.077305115319506, + "learning_rate": 2.686170323626773e-06, + "loss": 0.0834, + "step": 2954 + }, + { + "epoch": 2.1255169933465203, + "grad_norm": 5.037579660156506, + "learning_rate": 2.685962029270673e-06, + "loss": 0.0277, + "step": 2955 + }, + { + "epoch": 2.1262362884373314, + "grad_norm": 6.254952615196554, + "learning_rate": 2.685753673894745e-06, + "loss": 0.2789, + "step": 2956 + }, + { + "epoch": 2.1269555835281424, + "grad_norm": 5.266516029647939, + "learning_rate": 2.6855452575097094e-06, + "loss": 0.0862, + "step": 2957 + }, + { + "epoch": 2.1276748786189534, + "grad_norm": 5.512030801731916, + "learning_rate": 2.6853367801262883e-06, + "loss": 0.0492, + "step": 2958 + }, + { + "epoch": 2.1283941737097645, + "grad_norm": 3.7268551741164258, + "learning_rate": 2.6851282417552093e-06, + "loss": 0.1145, + "step": 2959 + }, + { + "epoch": 2.1291134688005755, + "grad_norm": 3.8324970526459134, + "learning_rate": 2.6849196424072024e-06, + "loss": 0.0984, + "step": 2960 + }, + { + "epoch": 2.1298327638913865, + "grad_norm": 3.495163608649748, + "learning_rate": 2.6847109820929993e-06, + "loss": 0.0758, + "step": 2961 + }, + { + "epoch": 2.1305520589821976, + "grad_norm": 3.259665676266259, + "learning_rate": 2.6845022608233364e-06, + "loss": 0.1327, + "step": 2962 + }, + { + "epoch": 2.1312713540730086, + "grad_norm": 10.163836273929322, + "learning_rate": 2.684293478608953e-06, + "loss": 0.0334, + "step": 2963 + }, + { + "epoch": 2.1319906491638196, + "grad_norm": 2.3310650358435003, + "learning_rate": 2.6840846354605917e-06, + "loss": 0.063, + "step": 2964 + }, + { + "epoch": 2.1327099442546302, + "grad_norm": 5.0634793593735115, + "learning_rate": 2.683875731388996e-06, + "loss": 0.0674, + "step": 2965 + }, + { + "epoch": 2.1334292393454413, + "grad_norm": 6.464794535905222, + "learning_rate": 2.683666766404917e-06, + "loss": 0.0299, + "step": 2966 + }, + { + "epoch": 2.1341485344362523, + "grad_norm": 5.249986535858385, + "learning_rate": 2.6834577405191044e-06, + "loss": 0.1801, + "step": 2967 + }, + { + "epoch": 2.1348678295270633, + "grad_norm": 4.074269595258422, + "learning_rate": 2.683248653742313e-06, + "loss": 0.0764, + "step": 2968 + }, + { + "epoch": 2.1355871246178744, + "grad_norm": 4.282602739228891, + "learning_rate": 2.6830395060853025e-06, + "loss": 0.1766, + "step": 2969 + }, + { + "epoch": 2.1363064197086854, + "grad_norm": 4.918205245332558, + "learning_rate": 2.682830297558832e-06, + "loss": 0.0471, + "step": 2970 + }, + { + "epoch": 2.1370257147994964, + "grad_norm": 4.4657728485742725, + "learning_rate": 2.6826210281736663e-06, + "loss": 0.1888, + "step": 2971 + }, + { + "epoch": 2.1377450098903075, + "grad_norm": 2.846559633436882, + "learning_rate": 2.682411697940573e-06, + "loss": 0.0919, + "step": 2972 + }, + { + "epoch": 2.1384643049811185, + "grad_norm": 5.699903429233865, + "learning_rate": 2.6822023068703217e-06, + "loss": 0.1778, + "step": 2973 + }, + { + "epoch": 2.1391836000719295, + "grad_norm": 7.0843882964182585, + "learning_rate": 2.6819928549736864e-06, + "loss": 0.0849, + "step": 2974 + }, + { + "epoch": 2.1399028951627406, + "grad_norm": 0.6260016601724464, + "learning_rate": 2.681783342261444e-06, + "loss": 0.0018, + "step": 2975 + }, + { + "epoch": 2.1406221902535516, + "grad_norm": 3.889841875582308, + "learning_rate": 2.6815737687443743e-06, + "loss": 0.1055, + "step": 2976 + }, + { + "epoch": 2.1413414853443626, + "grad_norm": 2.6137391227573405, + "learning_rate": 2.68136413443326e-06, + "loss": 0.0577, + "step": 2977 + }, + { + "epoch": 2.1420607804351737, + "grad_norm": 2.612490229406799, + "learning_rate": 2.6811544393388867e-06, + "loss": 0.0384, + "step": 2978 + }, + { + "epoch": 2.1427800755259847, + "grad_norm": 4.29215554513666, + "learning_rate": 2.6809446834720443e-06, + "loss": 0.1313, + "step": 2979 + }, + { + "epoch": 2.1434993706167957, + "grad_norm": 2.17004453287081, + "learning_rate": 2.680734866843525e-06, + "loss": 0.0502, + "step": 2980 + }, + { + "epoch": 2.1442186657076068, + "grad_norm": 3.3945073452656906, + "learning_rate": 2.680524989464124e-06, + "loss": 0.1423, + "step": 2981 + }, + { + "epoch": 2.1449379607984174, + "grad_norm": 5.202709582701237, + "learning_rate": 2.68031505134464e-06, + "loss": 0.2498, + "step": 2982 + }, + { + "epoch": 2.1456572558892284, + "grad_norm": 8.065446627017817, + "learning_rate": 2.6801050524958743e-06, + "loss": 0.0606, + "step": 2983 + }, + { + "epoch": 2.1463765509800394, + "grad_norm": 2.4804218657275574, + "learning_rate": 2.6798949929286323e-06, + "loss": 0.0246, + "step": 2984 + }, + { + "epoch": 2.1470958460708505, + "grad_norm": 10.265412244776828, + "learning_rate": 2.679684872653722e-06, + "loss": 0.1096, + "step": 2985 + }, + { + "epoch": 2.1478151411616615, + "grad_norm": 0.738843136816604, + "learning_rate": 2.6794746916819536e-06, + "loss": 0.0026, + "step": 2986 + }, + { + "epoch": 2.1485344362524725, + "grad_norm": 5.243622357570583, + "learning_rate": 2.6792644500241416e-06, + "loss": 0.1345, + "step": 2987 + }, + { + "epoch": 2.1492537313432836, + "grad_norm": 10.341664595677095, + "learning_rate": 2.679054147691104e-06, + "loss": 0.1395, + "step": 2988 + }, + { + "epoch": 2.1499730264340946, + "grad_norm": 3.1635200937070285, + "learning_rate": 2.6788437846936608e-06, + "loss": 0.1319, + "step": 2989 + }, + { + "epoch": 2.1506923215249056, + "grad_norm": 1.5905810658491417, + "learning_rate": 2.6786333610426353e-06, + "loss": 0.0523, + "step": 2990 + }, + { + "epoch": 2.1514116166157167, + "grad_norm": 4.24920358932737, + "learning_rate": 2.6784228767488542e-06, + "loss": 0.0663, + "step": 2991 + }, + { + "epoch": 2.1521309117065277, + "grad_norm": 0.11900852508119229, + "learning_rate": 2.6782123318231474e-06, + "loss": 0.0004, + "step": 2992 + }, + { + "epoch": 2.1528502067973387, + "grad_norm": 3.999618463647315, + "learning_rate": 2.678001726276348e-06, + "loss": 0.153, + "step": 2993 + }, + { + "epoch": 2.1535695018881498, + "grad_norm": 2.490362701481361, + "learning_rate": 2.677791060119292e-06, + "loss": 0.0341, + "step": 2994 + }, + { + "epoch": 2.154288796978961, + "grad_norm": 1.6611576826425836, + "learning_rate": 2.6775803333628186e-06, + "loss": 0.0213, + "step": 2995 + }, + { + "epoch": 2.155008092069772, + "grad_norm": 5.226548229925101, + "learning_rate": 2.6773695460177697e-06, + "loss": 0.0671, + "step": 2996 + }, + { + "epoch": 2.1557273871605824, + "grad_norm": 5.062231585762362, + "learning_rate": 2.677158698094991e-06, + "loss": 0.2191, + "step": 2997 + }, + { + "epoch": 2.1564466822513935, + "grad_norm": 4.864395184930188, + "learning_rate": 2.67694778960533e-06, + "loss": 0.0969, + "step": 2998 + }, + { + "epoch": 2.1571659773422045, + "grad_norm": 3.5275070482483373, + "learning_rate": 2.67673682055964e-06, + "loss": 0.0154, + "step": 2999 + }, + { + "epoch": 2.1578852724330155, + "grad_norm": 0.8059943215117032, + "learning_rate": 2.6765257909687743e-06, + "loss": 0.0085, + "step": 3000 + }, + { + "epoch": 2.1586045675238266, + "grad_norm": 0.29667075068835635, + "learning_rate": 2.676314700843592e-06, + "loss": 0.0011, + "step": 3001 + }, + { + "epoch": 2.1593238626146376, + "grad_norm": 3.5364574232061226, + "learning_rate": 2.676103550194953e-06, + "loss": 0.0973, + "step": 3002 + }, + { + "epoch": 2.1600431577054486, + "grad_norm": 4.584638837921622, + "learning_rate": 2.6758923390337216e-06, + "loss": 0.0784, + "step": 3003 + }, + { + "epoch": 2.1607624527962597, + "grad_norm": 3.3382879608375604, + "learning_rate": 2.6756810673707655e-06, + "loss": 0.0122, + "step": 3004 + }, + { + "epoch": 2.1614817478870707, + "grad_norm": 0.07690135658984816, + "learning_rate": 2.6754697352169544e-06, + "loss": 0.0003, + "step": 3005 + }, + { + "epoch": 2.1622010429778817, + "grad_norm": 2.276349708864349, + "learning_rate": 2.675258342583162e-06, + "loss": 0.0591, + "step": 3006 + }, + { + "epoch": 2.1629203380686928, + "grad_norm": 2.167555930442845, + "learning_rate": 2.675046889480265e-06, + "loss": 0.0063, + "step": 3007 + }, + { + "epoch": 2.163639633159504, + "grad_norm": 1.8130281675141482, + "learning_rate": 2.6748353759191424e-06, + "loss": 0.0403, + "step": 3008 + }, + { + "epoch": 2.164358928250315, + "grad_norm": 3.878861015105422, + "learning_rate": 2.6746238019106778e-06, + "loss": 0.146, + "step": 3009 + }, + { + "epoch": 2.165078223341126, + "grad_norm": 4.150130225843815, + "learning_rate": 2.6744121674657563e-06, + "loss": 0.0733, + "step": 3010 + }, + { + "epoch": 2.165797518431937, + "grad_norm": 4.496257504023216, + "learning_rate": 2.6742004725952674e-06, + "loss": 0.1194, + "step": 3011 + }, + { + "epoch": 2.1665168135227475, + "grad_norm": 5.020480889325239, + "learning_rate": 2.6739887173101028e-06, + "loss": 0.1292, + "step": 3012 + }, + { + "epoch": 2.1672361086135585, + "grad_norm": 3.675056734361526, + "learning_rate": 2.6737769016211577e-06, + "loss": 0.0726, + "step": 3013 + }, + { + "epoch": 2.1679554037043696, + "grad_norm": 4.212027705118157, + "learning_rate": 2.6735650255393306e-06, + "loss": 0.1671, + "step": 3014 + }, + { + "epoch": 2.1686746987951806, + "grad_norm": 4.114358040075331, + "learning_rate": 2.673353089075523e-06, + "loss": 0.0312, + "step": 3015 + }, + { + "epoch": 2.1693939938859916, + "grad_norm": 4.180423186928452, + "learning_rate": 2.673141092240639e-06, + "loss": 0.0428, + "step": 3016 + }, + { + "epoch": 2.1701132889768027, + "grad_norm": 4.085663632187638, + "learning_rate": 2.6729290350455863e-06, + "loss": 0.1238, + "step": 3017 + }, + { + "epoch": 2.1708325840676137, + "grad_norm": 4.640466564508529, + "learning_rate": 2.672716917501276e-06, + "loss": 0.1065, + "step": 3018 + }, + { + "epoch": 2.1715518791584247, + "grad_norm": 2.5048164408670575, + "learning_rate": 2.6725047396186214e-06, + "loss": 0.0539, + "step": 3019 + }, + { + "epoch": 2.1722711742492358, + "grad_norm": 3.509772459152732, + "learning_rate": 2.6722925014085397e-06, + "loss": 0.0838, + "step": 3020 + }, + { + "epoch": 2.172990469340047, + "grad_norm": 4.819312111793526, + "learning_rate": 2.6720802028819512e-06, + "loss": 0.194, + "step": 3021 + }, + { + "epoch": 2.173709764430858, + "grad_norm": 4.124789055153528, + "learning_rate": 2.6718678440497785e-06, + "loss": 0.0679, + "step": 3022 + }, + { + "epoch": 2.174429059521669, + "grad_norm": 3.0456272847013452, + "learning_rate": 2.671655424922948e-06, + "loss": 0.062, + "step": 3023 + }, + { + "epoch": 2.17514835461248, + "grad_norm": 2.45044484575865, + "learning_rate": 2.6714429455123895e-06, + "loss": 0.1028, + "step": 3024 + }, + { + "epoch": 2.175867649703291, + "grad_norm": 4.288011935264229, + "learning_rate": 2.6712304058290348e-06, + "loss": 0.1486, + "step": 3025 + }, + { + "epoch": 2.176586944794102, + "grad_norm": 5.760993852396619, + "learning_rate": 2.6710178058838196e-06, + "loss": 0.1088, + "step": 3026 + }, + { + "epoch": 2.1773062398849126, + "grad_norm": 7.444748149105423, + "learning_rate": 2.670805145687683e-06, + "loss": 0.1549, + "step": 3027 + }, + { + "epoch": 2.1780255349757236, + "grad_norm": 3.172223258702454, + "learning_rate": 2.670592425251566e-06, + "loss": 0.095, + "step": 3028 + }, + { + "epoch": 2.1787448300665346, + "grad_norm": 4.424579500391891, + "learning_rate": 2.6703796445864138e-06, + "loss": 0.09, + "step": 3029 + }, + { + "epoch": 2.1794641251573457, + "grad_norm": 4.079890933420054, + "learning_rate": 2.6701668037031744e-06, + "loss": 0.1316, + "step": 3030 + }, + { + "epoch": 2.1801834202481567, + "grad_norm": 3.690180223968812, + "learning_rate": 2.6699539026127992e-06, + "loss": 0.0598, + "step": 3031 + }, + { + "epoch": 2.1809027153389677, + "grad_norm": 2.137619410962381, + "learning_rate": 2.669740941326241e-06, + "loss": 0.0059, + "step": 3032 + }, + { + "epoch": 2.1816220104297788, + "grad_norm": 2.269622753530593, + "learning_rate": 2.669527919854459e-06, + "loss": 0.0062, + "step": 3033 + }, + { + "epoch": 2.18234130552059, + "grad_norm": 4.202944797807136, + "learning_rate": 2.669314838208412e-06, + "loss": 0.1886, + "step": 3034 + }, + { + "epoch": 2.183060600611401, + "grad_norm": 4.811381399924753, + "learning_rate": 2.6691016963990637e-06, + "loss": 0.2848, + "step": 3035 + }, + { + "epoch": 2.183779895702212, + "grad_norm": 0.7006711844653877, + "learning_rate": 2.668888494437381e-06, + "loss": 0.0038, + "step": 3036 + }, + { + "epoch": 2.184499190793023, + "grad_norm": 2.08816362262734, + "learning_rate": 2.6686752323343335e-06, + "loss": 0.0536, + "step": 3037 + }, + { + "epoch": 2.185218485883834, + "grad_norm": 3.292789418425388, + "learning_rate": 2.6684619101008935e-06, + "loss": 0.1126, + "step": 3038 + }, + { + "epoch": 2.185937780974645, + "grad_norm": 1.6968916278192299, + "learning_rate": 2.6682485277480373e-06, + "loss": 0.0497, + "step": 3039 + }, + { + "epoch": 2.186657076065456, + "grad_norm": 4.034804436079525, + "learning_rate": 2.6680350852867433e-06, + "loss": 0.096, + "step": 3040 + }, + { + "epoch": 2.187376371156267, + "grad_norm": 4.226381220734043, + "learning_rate": 2.6678215827279937e-06, + "loss": 0.0218, + "step": 3041 + }, + { + "epoch": 2.1880956662470776, + "grad_norm": 4.35835483225906, + "learning_rate": 2.667608020082774e-06, + "loss": 0.0774, + "step": 3042 + }, + { + "epoch": 2.1888149613378887, + "grad_norm": 5.590528732088101, + "learning_rate": 2.667394397362072e-06, + "loss": 0.0916, + "step": 3043 + }, + { + "epoch": 2.1895342564286997, + "grad_norm": 1.4990879729942823, + "learning_rate": 2.6671807145768783e-06, + "loss": 0.0036, + "step": 3044 + }, + { + "epoch": 2.1902535515195107, + "grad_norm": 1.8407408914777634, + "learning_rate": 2.6669669717381886e-06, + "loss": 0.0091, + "step": 3045 + }, + { + "epoch": 2.1909728466103218, + "grad_norm": 4.632695370374956, + "learning_rate": 2.666753168856999e-06, + "loss": 0.089, + "step": 3046 + }, + { + "epoch": 2.191692141701133, + "grad_norm": 3.699219177846792, + "learning_rate": 2.6665393059443117e-06, + "loss": 0.1049, + "step": 3047 + }, + { + "epoch": 2.192411436791944, + "grad_norm": 3.358600634471962, + "learning_rate": 2.6663253830111284e-06, + "loss": 0.0207, + "step": 3048 + }, + { + "epoch": 2.193130731882755, + "grad_norm": 4.742007947914259, + "learning_rate": 2.6661114000684572e-06, + "loss": 0.0329, + "step": 3049 + }, + { + "epoch": 2.193850026973566, + "grad_norm": 5.744341087408164, + "learning_rate": 2.6658973571273076e-06, + "loss": 0.1555, + "step": 3050 + }, + { + "epoch": 2.194569322064377, + "grad_norm": 3.0015122433218164, + "learning_rate": 2.6656832541986917e-06, + "loss": 0.0614, + "step": 3051 + }, + { + "epoch": 2.195288617155188, + "grad_norm": 2.8691462886200503, + "learning_rate": 2.6654690912936264e-06, + "loss": 0.1052, + "step": 3052 + }, + { + "epoch": 2.196007912245999, + "grad_norm": 1.1360029414327333, + "learning_rate": 2.6652548684231305e-06, + "loss": 0.0065, + "step": 3053 + }, + { + "epoch": 2.19672720733681, + "grad_norm": 3.603605197657369, + "learning_rate": 2.665040585598226e-06, + "loss": 0.0831, + "step": 3054 + }, + { + "epoch": 2.197446502427621, + "grad_norm": 2.88806921822241, + "learning_rate": 2.664826242829938e-06, + "loss": 0.0119, + "step": 3055 + }, + { + "epoch": 2.198165797518432, + "grad_norm": 2.1464758097691026, + "learning_rate": 2.6646118401292954e-06, + "loss": 0.0504, + "step": 3056 + }, + { + "epoch": 2.198885092609243, + "grad_norm": 0.8742601408280847, + "learning_rate": 2.6643973775073287e-06, + "loss": 0.0024, + "step": 3057 + }, + { + "epoch": 2.199604387700054, + "grad_norm": 3.1768981905761366, + "learning_rate": 2.664182854975073e-06, + "loss": 0.0191, + "step": 3058 + }, + { + "epoch": 2.2003236827908648, + "grad_norm": 3.2981551100920825, + "learning_rate": 2.6639682725435664e-06, + "loss": 0.0743, + "step": 3059 + }, + { + "epoch": 2.201042977881676, + "grad_norm": 5.2537328892294655, + "learning_rate": 2.6637536302238483e-06, + "loss": 0.2043, + "step": 3060 + }, + { + "epoch": 2.201762272972487, + "grad_norm": 4.145852215231934, + "learning_rate": 2.6635389280269634e-06, + "loss": 0.1833, + "step": 3061 + }, + { + "epoch": 2.202481568063298, + "grad_norm": 6.971997181248326, + "learning_rate": 2.6633241659639577e-06, + "loss": 0.0389, + "step": 3062 + }, + { + "epoch": 2.203200863154109, + "grad_norm": 3.8223544162447047, + "learning_rate": 2.6631093440458816e-06, + "loss": 0.0119, + "step": 3063 + }, + { + "epoch": 2.20392015824492, + "grad_norm": 0.6589239233339064, + "learning_rate": 2.662894462283788e-06, + "loss": 0.0013, + "step": 3064 + }, + { + "epoch": 2.204639453335731, + "grad_norm": 4.302088979137138, + "learning_rate": 2.6626795206887333e-06, + "loss": 0.1031, + "step": 3065 + }, + { + "epoch": 2.205358748426542, + "grad_norm": 3.447668234683055, + "learning_rate": 2.6624645192717757e-06, + "loss": 0.0361, + "step": 3066 + }, + { + "epoch": 2.206078043517353, + "grad_norm": 2.8259633678433076, + "learning_rate": 2.662249458043978e-06, + "loss": 0.0498, + "step": 3067 + }, + { + "epoch": 2.206797338608164, + "grad_norm": 3.1920687745161582, + "learning_rate": 2.6620343370164062e-06, + "loss": 0.0827, + "step": 3068 + }, + { + "epoch": 2.207516633698975, + "grad_norm": 4.923688877228495, + "learning_rate": 2.6618191562001273e-06, + "loss": 0.0839, + "step": 3069 + }, + { + "epoch": 2.208235928789786, + "grad_norm": 3.841778209790039, + "learning_rate": 2.6616039156062134e-06, + "loss": 0.1114, + "step": 3070 + }, + { + "epoch": 2.208955223880597, + "grad_norm": 4.16621976504901, + "learning_rate": 2.6613886152457386e-06, + "loss": 0.1253, + "step": 3071 + }, + { + "epoch": 2.209674518971408, + "grad_norm": 4.419159339269173, + "learning_rate": 2.661173255129781e-06, + "loss": 0.1242, + "step": 3072 + }, + { + "epoch": 2.2103938140622192, + "grad_norm": 7.488103182362443, + "learning_rate": 2.660957835269421e-06, + "loss": 0.0641, + "step": 3073 + }, + { + "epoch": 2.21111310915303, + "grad_norm": 5.214825466260846, + "learning_rate": 2.6607423556757424e-06, + "loss": 0.128, + "step": 3074 + }, + { + "epoch": 2.211832404243841, + "grad_norm": 3.105561969429023, + "learning_rate": 2.6605268163598324e-06, + "loss": 0.0448, + "step": 3075 + }, + { + "epoch": 2.212551699334652, + "grad_norm": 2.0604305855261233, + "learning_rate": 2.6603112173327794e-06, + "loss": 0.0519, + "step": 3076 + }, + { + "epoch": 2.213270994425463, + "grad_norm": 1.6548731989231549, + "learning_rate": 2.660095558605678e-06, + "loss": 0.0163, + "step": 3077 + }, + { + "epoch": 2.213990289516274, + "grad_norm": 3.4541999606719394, + "learning_rate": 2.659879840189624e-06, + "loss": 0.1401, + "step": 3078 + }, + { + "epoch": 2.214709584607085, + "grad_norm": 2.824920314336351, + "learning_rate": 2.6596640620957156e-06, + "loss": 0.0244, + "step": 3079 + }, + { + "epoch": 2.215428879697896, + "grad_norm": 2.7697266329639545, + "learning_rate": 2.6594482243350557e-06, + "loss": 0.0707, + "step": 3080 + }, + { + "epoch": 2.216148174788707, + "grad_norm": 2.3062724635950773, + "learning_rate": 2.6592323269187496e-06, + "loss": 0.0069, + "step": 3081 + }, + { + "epoch": 2.216867469879518, + "grad_norm": 6.030789082073365, + "learning_rate": 2.659016369857905e-06, + "loss": 0.1482, + "step": 3082 + }, + { + "epoch": 2.217586764970329, + "grad_norm": 4.282264507857953, + "learning_rate": 2.6588003531636334e-06, + "loss": 0.0476, + "step": 3083 + }, + { + "epoch": 2.21830606006114, + "grad_norm": 1.346627419906323, + "learning_rate": 2.65858427684705e-06, + "loss": 0.0108, + "step": 3084 + }, + { + "epoch": 2.219025355151951, + "grad_norm": 3.611694539938345, + "learning_rate": 2.658368140919271e-06, + "loss": 0.0321, + "step": 3085 + }, + { + "epoch": 2.2197446502427622, + "grad_norm": 4.206972357074755, + "learning_rate": 2.658151945391418e-06, + "loss": 0.1496, + "step": 3086 + }, + { + "epoch": 2.2204639453335733, + "grad_norm": 4.308843457335962, + "learning_rate": 2.6579356902746145e-06, + "loss": 0.0401, + "step": 3087 + }, + { + "epoch": 2.2211832404243843, + "grad_norm": 3.8243727547483046, + "learning_rate": 2.657719375579987e-06, + "loss": 0.0375, + "step": 3088 + }, + { + "epoch": 2.221902535515195, + "grad_norm": 5.928819220465108, + "learning_rate": 2.6575030013186654e-06, + "loss": 0.0901, + "step": 3089 + }, + { + "epoch": 2.222621830606006, + "grad_norm": 4.55346488043891, + "learning_rate": 2.657286567501782e-06, + "loss": 0.0099, + "step": 3090 + }, + { + "epoch": 2.223341125696817, + "grad_norm": 4.499254796889424, + "learning_rate": 2.6570700741404732e-06, + "loss": 0.2184, + "step": 3091 + }, + { + "epoch": 2.224060420787628, + "grad_norm": 2.6206496319399064, + "learning_rate": 2.6568535212458786e-06, + "loss": 0.108, + "step": 3092 + }, + { + "epoch": 2.224779715878439, + "grad_norm": 4.782337007279096, + "learning_rate": 2.6566369088291387e-06, + "loss": 0.1914, + "step": 3093 + }, + { + "epoch": 2.22549901096925, + "grad_norm": 1.1555587658614224, + "learning_rate": 2.6564202369014e-06, + "loss": 0.0161, + "step": 3094 + }, + { + "epoch": 2.226218306060061, + "grad_norm": 2.9691056961547186, + "learning_rate": 2.65620350547381e-06, + "loss": 0.0614, + "step": 3095 + }, + { + "epoch": 2.226937601150872, + "grad_norm": 0.9616482974102715, + "learning_rate": 2.65598671455752e-06, + "loss": 0.0186, + "step": 3096 + }, + { + "epoch": 2.227656896241683, + "grad_norm": 1.8839980332932773, + "learning_rate": 2.655769864163684e-06, + "loss": 0.0245, + "step": 3097 + }, + { + "epoch": 2.228376191332494, + "grad_norm": 2.4385583247630684, + "learning_rate": 2.6555529543034596e-06, + "loss": 0.0552, + "step": 3098 + }, + { + "epoch": 2.2290954864233052, + "grad_norm": 6.735281310814383, + "learning_rate": 2.6553359849880075e-06, + "loss": 0.1087, + "step": 3099 + }, + { + "epoch": 2.2298147815141163, + "grad_norm": 2.548770476578207, + "learning_rate": 2.6551189562284905e-06, + "loss": 0.0893, + "step": 3100 + }, + { + "epoch": 2.2305340766049273, + "grad_norm": 3.155652151145516, + "learning_rate": 2.6549018680360758e-06, + "loss": 0.0677, + "step": 3101 + }, + { + "epoch": 2.2312533716957383, + "grad_norm": 2.1949546512118174, + "learning_rate": 2.6546847204219323e-06, + "loss": 0.0368, + "step": 3102 + }, + { + "epoch": 2.2319726667865494, + "grad_norm": 4.9440348416183575, + "learning_rate": 2.6544675133972334e-06, + "loss": 0.1895, + "step": 3103 + }, + { + "epoch": 2.23269196187736, + "grad_norm": 5.463020706055418, + "learning_rate": 2.654250246973154e-06, + "loss": 0.1279, + "step": 3104 + }, + { + "epoch": 2.233411256968171, + "grad_norm": 1.436102030721381, + "learning_rate": 2.654032921160873e-06, + "loss": 0.0059, + "step": 3105 + }, + { + "epoch": 2.234130552058982, + "grad_norm": 2.829540367633825, + "learning_rate": 2.6538155359715727e-06, + "loss": 0.1505, + "step": 3106 + }, + { + "epoch": 2.234849847149793, + "grad_norm": 3.928851933506906, + "learning_rate": 2.6535980914164377e-06, + "loss": 0.1376, + "step": 3107 + }, + { + "epoch": 2.235569142240604, + "grad_norm": 2.770296962778117, + "learning_rate": 2.6533805875066556e-06, + "loss": 0.0936, + "step": 3108 + }, + { + "epoch": 2.236288437331415, + "grad_norm": 1.0681266078493197, + "learning_rate": 2.6531630242534174e-06, + "loss": 0.0254, + "step": 3109 + }, + { + "epoch": 2.237007732422226, + "grad_norm": 0.030510104157994884, + "learning_rate": 2.6529454016679175e-06, + "loss": 0.0002, + "step": 3110 + }, + { + "epoch": 2.237727027513037, + "grad_norm": 5.551385749463073, + "learning_rate": 2.6527277197613525e-06, + "loss": 0.3379, + "step": 3111 + }, + { + "epoch": 2.2384463226038482, + "grad_norm": 1.3553843636496001, + "learning_rate": 2.6525099785449234e-06, + "loss": 0.0035, + "step": 3112 + }, + { + "epoch": 2.2391656176946593, + "grad_norm": 5.718011817443443, + "learning_rate": 2.6522921780298323e-06, + "loss": 0.1444, + "step": 3113 + }, + { + "epoch": 2.2398849127854703, + "grad_norm": 3.593441842493296, + "learning_rate": 2.652074318227286e-06, + "loss": 0.0498, + "step": 3114 + }, + { + "epoch": 2.2406042078762813, + "grad_norm": 3.111707617281264, + "learning_rate": 2.6518563991484932e-06, + "loss": 0.012, + "step": 3115 + }, + { + "epoch": 2.2413235029670924, + "grad_norm": 5.802986567880055, + "learning_rate": 2.651638420804667e-06, + "loss": 0.2772, + "step": 3116 + }, + { + "epoch": 2.2420427980579034, + "grad_norm": 4.121308886450126, + "learning_rate": 2.6514203832070225e-06, + "loss": 0.1063, + "step": 3117 + }, + { + "epoch": 2.2427620931487144, + "grad_norm": 5.8871115749295875, + "learning_rate": 2.6512022863667777e-06, + "loss": 0.0834, + "step": 3118 + }, + { + "epoch": 2.243481388239525, + "grad_norm": 0.9379958287580851, + "learning_rate": 2.6509841302951545e-06, + "loss": 0.0026, + "step": 3119 + }, + { + "epoch": 2.244200683330336, + "grad_norm": 1.8625252842730533, + "learning_rate": 2.6507659150033773e-06, + "loss": 0.0103, + "step": 3120 + }, + { + "epoch": 2.244919978421147, + "grad_norm": 4.693076547652265, + "learning_rate": 2.650547640502674e-06, + "loss": 0.0383, + "step": 3121 + }, + { + "epoch": 2.245639273511958, + "grad_norm": 1.9409617699373858, + "learning_rate": 2.650329306804275e-06, + "loss": 0.0322, + "step": 3122 + }, + { + "epoch": 2.246358568602769, + "grad_norm": 5.047323420218156, + "learning_rate": 2.650110913919413e-06, + "loss": 0.1851, + "step": 3123 + }, + { + "epoch": 2.24707786369358, + "grad_norm": 4.152964517058767, + "learning_rate": 2.6498924618593263e-06, + "loss": 0.0232, + "step": 3124 + }, + { + "epoch": 2.2477971587843912, + "grad_norm": 4.181388212459385, + "learning_rate": 2.6496739506352535e-06, + "loss": 0.04, + "step": 3125 + }, + { + "epoch": 2.2485164538752023, + "grad_norm": 5.541747333116619, + "learning_rate": 2.649455380258438e-06, + "loss": 0.1298, + "step": 3126 + }, + { + "epoch": 2.2492357489660133, + "grad_norm": 3.4677824536010395, + "learning_rate": 2.6492367507401248e-06, + "loss": 0.0331, + "step": 3127 + }, + { + "epoch": 2.2499550440568243, + "grad_norm": 4.108935045626365, + "learning_rate": 2.649018062091564e-06, + "loss": 0.2064, + "step": 3128 + }, + { + "epoch": 2.2506743391476354, + "grad_norm": 3.3692509220688476, + "learning_rate": 2.6487993143240064e-06, + "loss": 0.1277, + "step": 3129 + }, + { + "epoch": 2.2513936342384464, + "grad_norm": 3.8016912674631658, + "learning_rate": 2.6485805074487077e-06, + "loss": 0.0555, + "step": 3130 + }, + { + "epoch": 2.2521129293292574, + "grad_norm": 4.840451003937193, + "learning_rate": 2.6483616414769257e-06, + "loss": 0.1553, + "step": 3131 + }, + { + "epoch": 2.2528322244200685, + "grad_norm": 5.281670339443802, + "learning_rate": 2.648142716419921e-06, + "loss": 0.1385, + "step": 3132 + }, + { + "epoch": 2.2535515195108795, + "grad_norm": 5.755838631696636, + "learning_rate": 2.6479237322889583e-06, + "loss": 0.0462, + "step": 3133 + }, + { + "epoch": 2.25427081460169, + "grad_norm": 3.132203960316398, + "learning_rate": 2.6477046890953044e-06, + "loss": 0.1352, + "step": 3134 + }, + { + "epoch": 2.2549901096925016, + "grad_norm": 4.320073845384806, + "learning_rate": 2.6474855868502297e-06, + "loss": 0.0982, + "step": 3135 + }, + { + "epoch": 2.255709404783312, + "grad_norm": 3.920408125103883, + "learning_rate": 2.6472664255650067e-06, + "loss": 0.0715, + "step": 3136 + }, + { + "epoch": 2.256428699874123, + "grad_norm": 1.860282120490665, + "learning_rate": 2.647047205250913e-06, + "loss": 0.0487, + "step": 3137 + }, + { + "epoch": 2.2571479949649342, + "grad_norm": 4.011568537652182, + "learning_rate": 2.646827925919226e-06, + "loss": 0.1121, + "step": 3138 + }, + { + "epoch": 2.2578672900557453, + "grad_norm": 3.5299060011450956, + "learning_rate": 2.64660858758123e-06, + "loss": 0.1042, + "step": 3139 + }, + { + "epoch": 2.2585865851465563, + "grad_norm": 1.8406257536718567, + "learning_rate": 2.6463891902482088e-06, + "loss": 0.0371, + "step": 3140 + }, + { + "epoch": 2.2593058802373673, + "grad_norm": 3.2687682073787903, + "learning_rate": 2.646169733931451e-06, + "loss": 0.0236, + "step": 3141 + }, + { + "epoch": 2.2600251753281784, + "grad_norm": 2.5703984038313465, + "learning_rate": 2.645950218642249e-06, + "loss": 0.0568, + "step": 3142 + }, + { + "epoch": 2.2607444704189894, + "grad_norm": 7.178317791973445, + "learning_rate": 2.6457306443918964e-06, + "loss": 0.1419, + "step": 3143 + }, + { + "epoch": 2.2614637655098004, + "grad_norm": 2.5026856690020227, + "learning_rate": 2.645511011191691e-06, + "loss": 0.0351, + "step": 3144 + }, + { + "epoch": 2.2621830606006115, + "grad_norm": 2.1657039791045896, + "learning_rate": 2.6452913190529326e-06, + "loss": 0.0088, + "step": 3145 + }, + { + "epoch": 2.2629023556914225, + "grad_norm": 2.279016402288655, + "learning_rate": 2.645071567986926e-06, + "loss": 0.0176, + "step": 3146 + }, + { + "epoch": 2.2636216507822335, + "grad_norm": 6.099628841196772, + "learning_rate": 2.6448517580049767e-06, + "loss": 0.2778, + "step": 3147 + }, + { + "epoch": 2.2643409458730446, + "grad_norm": 2.0492876487644014, + "learning_rate": 2.6446318891183953e-06, + "loss": 0.0078, + "step": 3148 + }, + { + "epoch": 2.2650602409638556, + "grad_norm": 1.4951885996037297, + "learning_rate": 2.644411961338493e-06, + "loss": 0.0211, + "step": 3149 + }, + { + "epoch": 2.2657795360546666, + "grad_norm": 0.33223813321767576, + "learning_rate": 2.644191974676587e-06, + "loss": 0.0012, + "step": 3150 + }, + { + "epoch": 2.2664988311454772, + "grad_norm": 6.302396631023723, + "learning_rate": 2.6439719291439953e-06, + "loss": 0.2991, + "step": 3151 + }, + { + "epoch": 2.2672181262362883, + "grad_norm": 3.8098503758096727, + "learning_rate": 2.6437518247520396e-06, + "loss": 0.1413, + "step": 3152 + }, + { + "epoch": 2.2679374213270993, + "grad_norm": 2.9614581694930684, + "learning_rate": 2.6435316615120447e-06, + "loss": 0.0704, + "step": 3153 + }, + { + "epoch": 2.2686567164179103, + "grad_norm": 3.567948743864831, + "learning_rate": 2.6433114394353384e-06, + "loss": 0.0931, + "step": 3154 + }, + { + "epoch": 2.2693760115087214, + "grad_norm": 2.877146070664812, + "learning_rate": 2.6430911585332513e-06, + "loss": 0.0179, + "step": 3155 + }, + { + "epoch": 2.2700953065995324, + "grad_norm": 4.404098845967638, + "learning_rate": 2.6428708188171174e-06, + "loss": 0.0993, + "step": 3156 + }, + { + "epoch": 2.2708146016903434, + "grad_norm": 5.952296052235995, + "learning_rate": 2.642650420298274e-06, + "loss": 0.2137, + "step": 3157 + }, + { + "epoch": 2.2715338967811545, + "grad_norm": 4.68130613818375, + "learning_rate": 2.6424299629880605e-06, + "loss": 0.0838, + "step": 3158 + }, + { + "epoch": 2.2722531918719655, + "grad_norm": 2.9327875125044818, + "learning_rate": 2.6422094468978202e-06, + "loss": 0.1198, + "step": 3159 + }, + { + "epoch": 2.2729724869627765, + "grad_norm": 7.118918166578994, + "learning_rate": 2.6419888720388984e-06, + "loss": 0.4315, + "step": 3160 + }, + { + "epoch": 2.2736917820535876, + "grad_norm": 5.517152386786748, + "learning_rate": 2.6417682384226445e-06, + "loss": 0.3461, + "step": 3161 + }, + { + "epoch": 2.2744110771443986, + "grad_norm": 3.3752909775236026, + "learning_rate": 2.6415475460604103e-06, + "loss": 0.0737, + "step": 3162 + }, + { + "epoch": 2.2751303722352096, + "grad_norm": 5.323040915013503, + "learning_rate": 2.641326794963551e-06, + "loss": 0.0896, + "step": 3163 + }, + { + "epoch": 2.2758496673260207, + "grad_norm": 2.977056064877686, + "learning_rate": 2.6411059851434243e-06, + "loss": 0.0589, + "step": 3164 + }, + { + "epoch": 2.2765689624168317, + "grad_norm": 2.4402915535172323, + "learning_rate": 2.640885116611392e-06, + "loss": 0.0343, + "step": 3165 + }, + { + "epoch": 2.2772882575076423, + "grad_norm": 3.294882762963217, + "learning_rate": 2.640664189378818e-06, + "loss": 0.0895, + "step": 3166 + }, + { + "epoch": 2.2780075525984533, + "grad_norm": 2.5086126833391162, + "learning_rate": 2.640443203457068e-06, + "loss": 0.0462, + "step": 3167 + }, + { + "epoch": 2.2787268476892644, + "grad_norm": 5.712148283429059, + "learning_rate": 2.6402221588575144e-06, + "loss": 0.2806, + "step": 3168 + }, + { + "epoch": 2.2794461427800754, + "grad_norm": 2.2443941728901726, + "learning_rate": 2.6400010555915287e-06, + "loss": 0.0807, + "step": 3169 + }, + { + "epoch": 2.2801654378708864, + "grad_norm": 4.121774184077429, + "learning_rate": 2.639779893670487e-06, + "loss": 0.1267, + "step": 3170 + }, + { + "epoch": 2.2808847329616975, + "grad_norm": 5.768451589776608, + "learning_rate": 2.6395586731057694e-06, + "loss": 0.0703, + "step": 3171 + }, + { + "epoch": 2.2816040280525085, + "grad_norm": 2.685928919688494, + "learning_rate": 2.6393373939087574e-06, + "loss": 0.0883, + "step": 3172 + }, + { + "epoch": 2.2823233231433195, + "grad_norm": 6.696330085918585, + "learning_rate": 2.6391160560908367e-06, + "loss": 0.2338, + "step": 3173 + }, + { + "epoch": 2.2830426182341306, + "grad_norm": 6.611166296543465, + "learning_rate": 2.638894659663395e-06, + "loss": 0.203, + "step": 3174 + }, + { + "epoch": 2.2837619133249416, + "grad_norm": 0.3734308692080473, + "learning_rate": 2.6386732046378243e-06, + "loss": 0.0014, + "step": 3175 + }, + { + "epoch": 2.2844812084157526, + "grad_norm": 3.2628901400469457, + "learning_rate": 2.6384516910255177e-06, + "loss": 0.1223, + "step": 3176 + }, + { + "epoch": 2.2852005035065637, + "grad_norm": 2.0397670262804404, + "learning_rate": 2.6382301188378733e-06, + "loss": 0.0135, + "step": 3177 + }, + { + "epoch": 2.2859197985973747, + "grad_norm": 1.9947481532087334, + "learning_rate": 2.6380084880862913e-06, + "loss": 0.0349, + "step": 3178 + }, + { + "epoch": 2.2866390936881857, + "grad_norm": 3.6195942903952294, + "learning_rate": 2.6377867987821745e-06, + "loss": 0.1501, + "step": 3179 + }, + { + "epoch": 2.2873583887789968, + "grad_norm": 4.563441919501048, + "learning_rate": 2.6375650509369303e-06, + "loss": 0.1943, + "step": 3180 + }, + { + "epoch": 2.2880776838698074, + "grad_norm": 4.6351640395391795, + "learning_rate": 2.637343244561967e-06, + "loss": 0.0193, + "step": 3181 + }, + { + "epoch": 2.2887969789606184, + "grad_norm": 4.14082647467695, + "learning_rate": 2.637121379668697e-06, + "loss": 0.2096, + "step": 3182 + }, + { + "epoch": 2.2895162740514294, + "grad_norm": 2.4403530330068044, + "learning_rate": 2.6368994562685355e-06, + "loss": 0.0921, + "step": 3183 + }, + { + "epoch": 2.2902355691422405, + "grad_norm": 2.8270799803446063, + "learning_rate": 2.6366774743729024e-06, + "loss": 0.089, + "step": 3184 + }, + { + "epoch": 2.2909548642330515, + "grad_norm": 3.633335232082756, + "learning_rate": 2.636455433993217e-06, + "loss": 0.1557, + "step": 3185 + }, + { + "epoch": 2.2916741593238625, + "grad_norm": 4.766040928579625, + "learning_rate": 2.6362333351409045e-06, + "loss": 0.05, + "step": 3186 + }, + { + "epoch": 2.2923934544146736, + "grad_norm": 2.901915924214587, + "learning_rate": 2.6360111778273924e-06, + "loss": 0.0211, + "step": 3187 + }, + { + "epoch": 2.2931127495054846, + "grad_norm": 3.19360889581358, + "learning_rate": 2.6357889620641117e-06, + "loss": 0.154, + "step": 3188 + }, + { + "epoch": 2.2938320445962956, + "grad_norm": 4.927885647196908, + "learning_rate": 2.635566687862494e-06, + "loss": 0.1897, + "step": 3189 + }, + { + "epoch": 2.2945513396871067, + "grad_norm": 3.826894403536538, + "learning_rate": 2.6353443552339772e-06, + "loss": 0.0127, + "step": 3190 + }, + { + "epoch": 2.2952706347779177, + "grad_norm": 4.488003219315735, + "learning_rate": 2.6351219641900005e-06, + "loss": 0.1046, + "step": 3191 + }, + { + "epoch": 2.2959899298687287, + "grad_norm": 2.6056650237929584, + "learning_rate": 2.634899514742006e-06, + "loss": 0.1008, + "step": 3192 + }, + { + "epoch": 2.2967092249595398, + "grad_norm": 6.445082830139398, + "learning_rate": 2.6346770069014393e-06, + "loss": 0.0784, + "step": 3193 + }, + { + "epoch": 2.297428520050351, + "grad_norm": 3.0478827675390434, + "learning_rate": 2.6344544406797484e-06, + "loss": 0.0672, + "step": 3194 + }, + { + "epoch": 2.298147815141162, + "grad_norm": 4.392440685929836, + "learning_rate": 2.6342318160883857e-06, + "loss": 0.0822, + "step": 3195 + }, + { + "epoch": 2.2988671102319724, + "grad_norm": 3.8900908966153644, + "learning_rate": 2.6340091331388042e-06, + "loss": 0.0228, + "step": 3196 + }, + { + "epoch": 2.299586405322784, + "grad_norm": 4.511217985765185, + "learning_rate": 2.6337863918424627e-06, + "loss": 0.1393, + "step": 3197 + }, + { + "epoch": 2.3003057004135945, + "grad_norm": 2.707253810269414, + "learning_rate": 2.6335635922108207e-06, + "loss": 0.057, + "step": 3198 + }, + { + "epoch": 2.3010249955044055, + "grad_norm": 7.328774868536007, + "learning_rate": 2.633340734255342e-06, + "loss": 0.3008, + "step": 3199 + }, + { + "epoch": 2.3017442905952166, + "grad_norm": 4.227621068655837, + "learning_rate": 2.6331178179874934e-06, + "loss": 0.137, + "step": 3200 + }, + { + "epoch": 2.3024635856860276, + "grad_norm": 2.229177161741265, + "learning_rate": 2.6328948434187433e-06, + "loss": 0.047, + "step": 3201 + }, + { + "epoch": 2.3031828807768386, + "grad_norm": 0.9093004970368942, + "learning_rate": 2.632671810560565e-06, + "loss": 0.0095, + "step": 3202 + }, + { + "epoch": 2.3039021758676497, + "grad_norm": 1.7377163845615304, + "learning_rate": 2.632448719424434e-06, + "loss": 0.0449, + "step": 3203 + }, + { + "epoch": 2.3046214709584607, + "grad_norm": 5.742766020731803, + "learning_rate": 2.6322255700218282e-06, + "loss": 0.1254, + "step": 3204 + }, + { + "epoch": 2.3053407660492717, + "grad_norm": 2.3689766354974426, + "learning_rate": 2.6320023623642295e-06, + "loss": 0.0697, + "step": 3205 + }, + { + "epoch": 2.3060600611400828, + "grad_norm": 4.684064289612578, + "learning_rate": 2.631779096463122e-06, + "loss": 0.1798, + "step": 3206 + }, + { + "epoch": 2.306779356230894, + "grad_norm": 3.0546339029182907, + "learning_rate": 2.631555772329993e-06, + "loss": 0.0891, + "step": 3207 + }, + { + "epoch": 2.307498651321705, + "grad_norm": 6.055255042058, + "learning_rate": 2.631332389976333e-06, + "loss": 0.0968, + "step": 3208 + }, + { + "epoch": 2.308217946412516, + "grad_norm": 5.505575914633913, + "learning_rate": 2.6311089494136357e-06, + "loss": 0.0783, + "step": 3209 + }, + { + "epoch": 2.308937241503327, + "grad_norm": 3.6646172010398823, + "learning_rate": 2.6308854506533974e-06, + "loss": 0.1522, + "step": 3210 + }, + { + "epoch": 2.3096565365941375, + "grad_norm": 5.002773504550886, + "learning_rate": 2.6306618937071177e-06, + "loss": 0.2355, + "step": 3211 + }, + { + "epoch": 2.310375831684949, + "grad_norm": 5.986526501043334, + "learning_rate": 2.630438278586298e-06, + "loss": 0.0623, + "step": 3212 + }, + { + "epoch": 2.3110951267757596, + "grad_norm": 5.714904457199305, + "learning_rate": 2.6302146053024456e-06, + "loss": 0.2362, + "step": 3213 + }, + { + "epoch": 2.3118144218665706, + "grad_norm": 4.022492054063511, + "learning_rate": 2.629990873867067e-06, + "loss": 0.1714, + "step": 3214 + }, + { + "epoch": 2.3125337169573816, + "grad_norm": 2.5832028401638722, + "learning_rate": 2.6297670842916748e-06, + "loss": 0.0263, + "step": 3215 + }, + { + "epoch": 2.3132530120481927, + "grad_norm": 2.6269669518033707, + "learning_rate": 2.629543236587782e-06, + "loss": 0.0252, + "step": 3216 + }, + { + "epoch": 2.3139723071390037, + "grad_norm": 4.417787754433668, + "learning_rate": 2.6293193307669074e-06, + "loss": 0.1439, + "step": 3217 + }, + { + "epoch": 2.3146916022298147, + "grad_norm": 4.479064879562671, + "learning_rate": 2.629095366840571e-06, + "loss": 0.1103, + "step": 3218 + }, + { + "epoch": 2.3154108973206258, + "grad_norm": 1.3272260207938729, + "learning_rate": 2.6288713448202953e-06, + "loss": 0.021, + "step": 3219 + }, + { + "epoch": 2.316130192411437, + "grad_norm": 1.7860871088357648, + "learning_rate": 2.628647264717607e-06, + "loss": 0.0331, + "step": 3220 + }, + { + "epoch": 2.316849487502248, + "grad_norm": 3.16881770218607, + "learning_rate": 2.628423126544037e-06, + "loss": 0.0282, + "step": 3221 + }, + { + "epoch": 2.317568782593059, + "grad_norm": 5.569281482560819, + "learning_rate": 2.6281989303111154e-06, + "loss": 0.2545, + "step": 3222 + }, + { + "epoch": 2.31828807768387, + "grad_norm": 3.937085780683137, + "learning_rate": 2.627974676030378e-06, + "loss": 0.0219, + "step": 3223 + }, + { + "epoch": 2.319007372774681, + "grad_norm": 4.681059642585563, + "learning_rate": 2.627750363713364e-06, + "loss": 0.0379, + "step": 3224 + }, + { + "epoch": 2.319726667865492, + "grad_norm": 3.5556408175114065, + "learning_rate": 2.6275259933716137e-06, + "loss": 0.0806, + "step": 3225 + }, + { + "epoch": 2.320445962956303, + "grad_norm": 4.914838878091459, + "learning_rate": 2.627301565016672e-06, + "loss": 0.2153, + "step": 3226 + }, + { + "epoch": 2.321165258047114, + "grad_norm": 4.421490232990445, + "learning_rate": 2.627077078660086e-06, + "loss": 0.1786, + "step": 3227 + }, + { + "epoch": 2.3218845531379246, + "grad_norm": 3.8499002494768932, + "learning_rate": 2.6268525343134053e-06, + "loss": 0.1013, + "step": 3228 + }, + { + "epoch": 2.3226038482287357, + "grad_norm": 3.084380894997996, + "learning_rate": 2.626627931988184e-06, + "loss": 0.1139, + "step": 3229 + }, + { + "epoch": 2.3233231433195467, + "grad_norm": 2.314498156322512, + "learning_rate": 2.626403271695978e-06, + "loss": 0.0636, + "step": 3230 + }, + { + "epoch": 2.3240424384103577, + "grad_norm": 3.651956815139289, + "learning_rate": 2.626178553448346e-06, + "loss": 0.1136, + "step": 3231 + }, + { + "epoch": 2.3247617335011688, + "grad_norm": 4.452654780777023, + "learning_rate": 2.6259537772568503e-06, + "loss": 0.1856, + "step": 3232 + }, + { + "epoch": 2.32548102859198, + "grad_norm": 0.8384514621235604, + "learning_rate": 2.6257289431330564e-06, + "loss": 0.0038, + "step": 3233 + }, + { + "epoch": 2.326200323682791, + "grad_norm": 0.5008624707146139, + "learning_rate": 2.6255040510885324e-06, + "loss": 0.0016, + "step": 3234 + }, + { + "epoch": 2.326919618773602, + "grad_norm": 3.668534297032654, + "learning_rate": 2.625279101134849e-06, + "loss": 0.0634, + "step": 3235 + }, + { + "epoch": 2.327638913864413, + "grad_norm": 7.115826003727372, + "learning_rate": 2.6250540932835806e-06, + "loss": 0.1129, + "step": 3236 + }, + { + "epoch": 2.328358208955224, + "grad_norm": 1.8504405875231962, + "learning_rate": 2.624829027546304e-06, + "loss": 0.0088, + "step": 3237 + }, + { + "epoch": 2.329077504046035, + "grad_norm": 6.654540441922057, + "learning_rate": 2.6246039039345997e-06, + "loss": 0.1282, + "step": 3238 + }, + { + "epoch": 2.329796799136846, + "grad_norm": 3.385123564832717, + "learning_rate": 2.62437872246005e-06, + "loss": 0.1041, + "step": 3239 + }, + { + "epoch": 2.330516094227657, + "grad_norm": 3.133796092188657, + "learning_rate": 2.6241534831342413e-06, + "loss": 0.1467, + "step": 3240 + }, + { + "epoch": 2.331235389318468, + "grad_norm": 3.4958061330003942, + "learning_rate": 2.6239281859687626e-06, + "loss": 0.0235, + "step": 3241 + }, + { + "epoch": 2.331954684409279, + "grad_norm": 4.6336141186220345, + "learning_rate": 2.623702830975206e-06, + "loss": 0.0945, + "step": 3242 + }, + { + "epoch": 2.3326739795000897, + "grad_norm": 3.740908834709217, + "learning_rate": 2.6234774181651655e-06, + "loss": 0.0151, + "step": 3243 + }, + { + "epoch": 2.3333932745909007, + "grad_norm": 3.2491759287648345, + "learning_rate": 2.6232519475502396e-06, + "loss": 0.111, + "step": 3244 + }, + { + "epoch": 2.3341125696817118, + "grad_norm": 2.8293464501125016, + "learning_rate": 2.6230264191420296e-06, + "loss": 0.1011, + "step": 3245 + }, + { + "epoch": 2.334831864772523, + "grad_norm": 0.8665390429828626, + "learning_rate": 2.6228008329521393e-06, + "loss": 0.0055, + "step": 3246 + }, + { + "epoch": 2.335551159863334, + "grad_norm": 4.305871716659971, + "learning_rate": 2.622575188992175e-06, + "loss": 0.1415, + "step": 3247 + }, + { + "epoch": 2.336270454954145, + "grad_norm": 1.7828022797847378, + "learning_rate": 2.6223494872737457e-06, + "loss": 0.0088, + "step": 3248 + }, + { + "epoch": 2.336989750044956, + "grad_norm": 3.8370622995810537, + "learning_rate": 2.622123727808466e-06, + "loss": 0.0508, + "step": 3249 + }, + { + "epoch": 2.337709045135767, + "grad_norm": 3.915717824877781, + "learning_rate": 2.6218979106079504e-06, + "loss": 0.1283, + "step": 3250 + }, + { + "epoch": 2.338428340226578, + "grad_norm": 0.21989529950977305, + "learning_rate": 2.621672035683818e-06, + "loss": 0.0005, + "step": 3251 + }, + { + "epoch": 2.339147635317389, + "grad_norm": 3.632246475177959, + "learning_rate": 2.62144610304769e-06, + "loss": 0.0683, + "step": 3252 + }, + { + "epoch": 2.3398669304082, + "grad_norm": 5.102368009167012, + "learning_rate": 2.621220112711192e-06, + "loss": 0.0857, + "step": 3253 + }, + { + "epoch": 2.340586225499011, + "grad_norm": 7.120913357059576, + "learning_rate": 2.6209940646859498e-06, + "loss": 0.1972, + "step": 3254 + }, + { + "epoch": 2.341305520589822, + "grad_norm": 2.819765217164804, + "learning_rate": 2.620767958983596e-06, + "loss": 0.0719, + "step": 3255 + }, + { + "epoch": 2.342024815680633, + "grad_norm": 4.521278162963943, + "learning_rate": 2.6205417956157627e-06, + "loss": 0.1091, + "step": 3256 + }, + { + "epoch": 2.342744110771444, + "grad_norm": 3.6339082644166805, + "learning_rate": 2.620315574594088e-06, + "loss": 0.1275, + "step": 3257 + }, + { + "epoch": 2.3434634058622548, + "grad_norm": 3.9336499752771803, + "learning_rate": 2.6200892959302094e-06, + "loss": 0.0788, + "step": 3258 + }, + { + "epoch": 2.3441827009530662, + "grad_norm": 3.4050869661139007, + "learning_rate": 2.6198629596357703e-06, + "loss": 0.0965, + "step": 3259 + }, + { + "epoch": 2.344901996043877, + "grad_norm": 2.9182447108112517, + "learning_rate": 2.6196365657224165e-06, + "loss": 0.0143, + "step": 3260 + }, + { + "epoch": 2.345621291134688, + "grad_norm": 4.367304625559985, + "learning_rate": 2.6194101142017958e-06, + "loss": 0.1405, + "step": 3261 + }, + { + "epoch": 2.346340586225499, + "grad_norm": 3.300698634372535, + "learning_rate": 2.6191836050855596e-06, + "loss": 0.0446, + "step": 3262 + }, + { + "epoch": 2.34705988131631, + "grad_norm": 2.126951173850292, + "learning_rate": 2.6189570383853623e-06, + "loss": 0.0523, + "step": 3263 + }, + { + "epoch": 2.347779176407121, + "grad_norm": 2.753151965244601, + "learning_rate": 2.618730414112861e-06, + "loss": 0.037, + "step": 3264 + }, + { + "epoch": 2.348498471497932, + "grad_norm": 3.4713069535858985, + "learning_rate": 2.618503732279716e-06, + "loss": 0.0494, + "step": 3265 + }, + { + "epoch": 2.349217766588743, + "grad_norm": 2.933278348690487, + "learning_rate": 2.618276992897591e-06, + "loss": 0.0883, + "step": 3266 + }, + { + "epoch": 2.349937061679554, + "grad_norm": 1.6307076004087933, + "learning_rate": 2.618050195978151e-06, + "loss": 0.0258, + "step": 3267 + }, + { + "epoch": 2.350656356770365, + "grad_norm": 2.605336933895664, + "learning_rate": 2.617823341533066e-06, + "loss": 0.0554, + "step": 3268 + }, + { + "epoch": 2.351375651861176, + "grad_norm": 0.6912758079659216, + "learning_rate": 2.617596429574008e-06, + "loss": 0.012, + "step": 3269 + }, + { + "epoch": 2.352094946951987, + "grad_norm": 6.249300910941087, + "learning_rate": 2.617369460112651e-06, + "loss": 0.0425, + "step": 3270 + }, + { + "epoch": 2.352814242042798, + "grad_norm": 1.922700386099548, + "learning_rate": 2.617142433160675e-06, + "loss": 0.0418, + "step": 3271 + }, + { + "epoch": 2.3535335371336092, + "grad_norm": 4.465866071095441, + "learning_rate": 2.616915348729759e-06, + "loss": 0.1555, + "step": 3272 + }, + { + "epoch": 2.35425283222442, + "grad_norm": 2.0089191688740478, + "learning_rate": 2.616688206831588e-06, + "loss": 0.0066, + "step": 3273 + }, + { + "epoch": 2.3549721273152313, + "grad_norm": 0.6821764097499252, + "learning_rate": 2.616461007477848e-06, + "loss": 0.0027, + "step": 3274 + }, + { + "epoch": 2.355691422406042, + "grad_norm": 2.3125492349411987, + "learning_rate": 2.6162337506802296e-06, + "loss": 0.0468, + "step": 3275 + }, + { + "epoch": 2.356410717496853, + "grad_norm": 1.7133712405744657, + "learning_rate": 2.616006436450425e-06, + "loss": 0.0164, + "step": 3276 + }, + { + "epoch": 2.357130012587664, + "grad_norm": 3.6468154805373465, + "learning_rate": 2.6157790648001305e-06, + "loss": 0.1108, + "step": 3277 + }, + { + "epoch": 2.357849307678475, + "grad_norm": 2.630860690600742, + "learning_rate": 2.615551635741044e-06, + "loss": 0.0744, + "step": 3278 + }, + { + "epoch": 2.358568602769286, + "grad_norm": 4.205270685035885, + "learning_rate": 2.6153241492848678e-06, + "loss": 0.1699, + "step": 3279 + }, + { + "epoch": 2.359287897860097, + "grad_norm": 5.15763402108436, + "learning_rate": 2.6150966054433066e-06, + "loss": 0.1655, + "step": 3280 + }, + { + "epoch": 2.360007192950908, + "grad_norm": 6.6261376506575855, + "learning_rate": 2.614869004228067e-06, + "loss": 0.1239, + "step": 3281 + }, + { + "epoch": 2.360726488041719, + "grad_norm": 8.706499400256206, + "learning_rate": 2.6146413456508604e-06, + "loss": 0.2258, + "step": 3282 + }, + { + "epoch": 2.36144578313253, + "grad_norm": 5.074909894744741, + "learning_rate": 2.6144136297233993e-06, + "loss": 0.1163, + "step": 3283 + }, + { + "epoch": 2.362165078223341, + "grad_norm": 7.135554117990332, + "learning_rate": 2.6141858564574014e-06, + "loss": 0.1388, + "step": 3284 + }, + { + "epoch": 2.3628843733141522, + "grad_norm": 4.576786876407508, + "learning_rate": 2.6139580258645847e-06, + "loss": 0.1137, + "step": 3285 + }, + { + "epoch": 2.3636036684049633, + "grad_norm": 6.097897237511379, + "learning_rate": 2.6137301379566725e-06, + "loss": 0.1427, + "step": 3286 + }, + { + "epoch": 2.3643229634957743, + "grad_norm": 2.8765828585811084, + "learning_rate": 2.6135021927453894e-06, + "loss": 0.1441, + "step": 3287 + }, + { + "epoch": 2.365042258586585, + "grad_norm": 1.7587624042284395, + "learning_rate": 2.613274190242464e-06, + "loss": 0.0624, + "step": 3288 + }, + { + "epoch": 2.3657615536773964, + "grad_norm": 2.2563846892607775, + "learning_rate": 2.613046130459627e-06, + "loss": 0.0068, + "step": 3289 + }, + { + "epoch": 2.366480848768207, + "grad_norm": 7.365683069539245, + "learning_rate": 2.6128180134086127e-06, + "loss": 0.3169, + "step": 3290 + }, + { + "epoch": 2.367200143859018, + "grad_norm": 7.742375105819887, + "learning_rate": 2.612589839101158e-06, + "loss": 0.0208, + "step": 3291 + }, + { + "epoch": 2.367919438949829, + "grad_norm": 3.44450696687367, + "learning_rate": 2.612361607549003e-06, + "loss": 0.0654, + "step": 3292 + }, + { + "epoch": 2.36863873404064, + "grad_norm": 1.8758272923381216, + "learning_rate": 2.612133318763891e-06, + "loss": 0.0077, + "step": 3293 + }, + { + "epoch": 2.369358029131451, + "grad_norm": 0.31217239936916535, + "learning_rate": 2.6119049727575674e-06, + "loss": 0.001, + "step": 3294 + }, + { + "epoch": 2.370077324222262, + "grad_norm": 5.398066594081044, + "learning_rate": 2.611676569541781e-06, + "loss": 0.2323, + "step": 3295 + }, + { + "epoch": 2.370796619313073, + "grad_norm": 2.733079687555914, + "learning_rate": 2.611448109128283e-06, + "loss": 0.0251, + "step": 3296 + }, + { + "epoch": 2.371515914403884, + "grad_norm": 4.673694763889029, + "learning_rate": 2.6112195915288295e-06, + "loss": 0.1463, + "step": 3297 + }, + { + "epoch": 2.3722352094946952, + "grad_norm": 2.0619641737898617, + "learning_rate": 2.610991016755177e-06, + "loss": 0.07, + "step": 3298 + }, + { + "epoch": 2.3729545045855063, + "grad_norm": 4.895604290524185, + "learning_rate": 2.6107623848190868e-06, + "loss": 0.1705, + "step": 3299 + }, + { + "epoch": 2.3736737996763173, + "grad_norm": 0.02379951344047758, + "learning_rate": 2.6105336957323216e-06, + "loss": 0.0002, + "step": 3300 + }, + { + "epoch": 2.3743930947671283, + "grad_norm": 2.718621997906582, + "learning_rate": 2.6103049495066484e-06, + "loss": 0.0589, + "step": 3301 + }, + { + "epoch": 2.3751123898579394, + "grad_norm": 0.20847892315946664, + "learning_rate": 2.610076146153836e-06, + "loss": 0.0005, + "step": 3302 + }, + { + "epoch": 2.3758316849487504, + "grad_norm": 5.962183935914835, + "learning_rate": 2.609847285685658e-06, + "loss": 0.1788, + "step": 3303 + }, + { + "epoch": 2.3765509800395614, + "grad_norm": 4.2281524682648985, + "learning_rate": 2.6096183681138883e-06, + "loss": 0.0638, + "step": 3304 + }, + { + "epoch": 2.377270275130372, + "grad_norm": 4.477807691422038, + "learning_rate": 2.609389393450306e-06, + "loss": 0.022, + "step": 3305 + }, + { + "epoch": 2.377989570221183, + "grad_norm": 6.325700217130086, + "learning_rate": 2.6091603617066915e-06, + "loss": 0.1787, + "step": 3306 + }, + { + "epoch": 2.378708865311994, + "grad_norm": 6.746682756191801, + "learning_rate": 2.6089312728948294e-06, + "loss": 0.1382, + "step": 3307 + }, + { + "epoch": 2.379428160402805, + "grad_norm": 4.268627621717964, + "learning_rate": 2.6087021270265074e-06, + "loss": 0.1834, + "step": 3308 + }, + { + "epoch": 2.380147455493616, + "grad_norm": 2.559970540991954, + "learning_rate": 2.608472924113514e-06, + "loss": 0.0432, + "step": 3309 + }, + { + "epoch": 2.380866750584427, + "grad_norm": 1.284566262720924, + "learning_rate": 2.608243664167643e-06, + "loss": 0.0153, + "step": 3310 + }, + { + "epoch": 2.3815860456752382, + "grad_norm": 8.963321732078805, + "learning_rate": 2.60801434720069e-06, + "loss": 0.2047, + "step": 3311 + }, + { + "epoch": 2.3823053407660493, + "grad_norm": 1.4640861417792845, + "learning_rate": 2.607784973224454e-06, + "loss": 0.0189, + "step": 3312 + }, + { + "epoch": 2.3830246358568603, + "grad_norm": 5.548997010812094, + "learning_rate": 2.607555542250736e-06, + "loss": 0.0788, + "step": 3313 + }, + { + "epoch": 2.3837439309476713, + "grad_norm": 4.609212106189434, + "learning_rate": 2.607326054291341e-06, + "loss": 0.0742, + "step": 3314 + }, + { + "epoch": 2.3844632260384824, + "grad_norm": 3.3537050038921765, + "learning_rate": 2.607096509358077e-06, + "loss": 0.0666, + "step": 3315 + }, + { + "epoch": 2.3851825211292934, + "grad_norm": 4.735711732501709, + "learning_rate": 2.6068669074627543e-06, + "loss": 0.1777, + "step": 3316 + }, + { + "epoch": 2.3859018162201044, + "grad_norm": 0.6845861451879028, + "learning_rate": 2.606637248617186e-06, + "loss": 0.0023, + "step": 3317 + }, + { + "epoch": 2.3866211113109155, + "grad_norm": 5.202610760805279, + "learning_rate": 2.6064075328331886e-06, + "loss": 0.221, + "step": 3318 + }, + { + "epoch": 2.3873404064017265, + "grad_norm": 1.4539529324271563, + "learning_rate": 2.6061777601225813e-06, + "loss": 0.0039, + "step": 3319 + }, + { + "epoch": 2.388059701492537, + "grad_norm": 1.7351750997833395, + "learning_rate": 2.6059479304971867e-06, + "loss": 0.0039, + "step": 3320 + }, + { + "epoch": 2.388778996583348, + "grad_norm": 5.693676322666301, + "learning_rate": 2.6057180439688293e-06, + "loss": 0.1728, + "step": 3321 + }, + { + "epoch": 2.389498291674159, + "grad_norm": 5.880913972371311, + "learning_rate": 2.605488100549338e-06, + "loss": 0.0984, + "step": 3322 + }, + { + "epoch": 2.39021758676497, + "grad_norm": 5.657369583615648, + "learning_rate": 2.605258100250543e-06, + "loss": 0.2198, + "step": 3323 + }, + { + "epoch": 2.3909368818557812, + "grad_norm": 4.291443872787153, + "learning_rate": 2.6050280430842783e-06, + "loss": 0.0188, + "step": 3324 + }, + { + "epoch": 2.3916561769465923, + "grad_norm": 0.21068462410677244, + "learning_rate": 2.604797929062381e-06, + "loss": 0.0004, + "step": 3325 + }, + { + "epoch": 2.3923754720374033, + "grad_norm": 1.9287491125814664, + "learning_rate": 2.604567758196691e-06, + "loss": 0.036, + "step": 3326 + }, + { + "epoch": 2.3930947671282143, + "grad_norm": 3.5285771990173767, + "learning_rate": 2.604337530499051e-06, + "loss": 0.1178, + "step": 3327 + }, + { + "epoch": 2.3938140622190254, + "grad_norm": 3.8284554381985654, + "learning_rate": 2.6041072459813064e-06, + "loss": 0.0799, + "step": 3328 + }, + { + "epoch": 2.3945333573098364, + "grad_norm": 3.98823087275605, + "learning_rate": 2.6038769046553056e-06, + "loss": 0.1615, + "step": 3329 + }, + { + "epoch": 2.3952526524006474, + "grad_norm": 2.0777798272628756, + "learning_rate": 2.6036465065329005e-06, + "loss": 0.0334, + "step": 3330 + }, + { + "epoch": 2.3959719474914585, + "grad_norm": 2.2956077869831657, + "learning_rate": 2.603416051625945e-06, + "loss": 0.1066, + "step": 3331 + }, + { + "epoch": 2.3966912425822695, + "grad_norm": 3.910657981411721, + "learning_rate": 2.603185539946297e-06, + "loss": 0.1121, + "step": 3332 + }, + { + "epoch": 2.3974105376730805, + "grad_norm": 4.969600182242484, + "learning_rate": 2.602954971505816e-06, + "loss": 0.1736, + "step": 3333 + }, + { + "epoch": 2.3981298327638916, + "grad_norm": 4.619373398717188, + "learning_rate": 2.6027243463163656e-06, + "loss": 0.178, + "step": 3334 + }, + { + "epoch": 2.398849127854702, + "grad_norm": 4.0176905539229, + "learning_rate": 2.6024936643898117e-06, + "loss": 0.1219, + "step": 3335 + }, + { + "epoch": 2.3995684229455136, + "grad_norm": 2.911164907699844, + "learning_rate": 2.6022629257380238e-06, + "loss": 0.1077, + "step": 3336 + }, + { + "epoch": 2.4002877180363242, + "grad_norm": 0.05520278714887086, + "learning_rate": 2.6020321303728733e-06, + "loss": 0.0002, + "step": 3337 + }, + { + "epoch": 2.4010070131271353, + "grad_norm": 3.700502145048883, + "learning_rate": 2.601801278306235e-06, + "loss": 0.1374, + "step": 3338 + }, + { + "epoch": 2.4017263082179463, + "grad_norm": 4.375445460056996, + "learning_rate": 2.6015703695499867e-06, + "loss": 0.0575, + "step": 3339 + }, + { + "epoch": 2.4024456033087573, + "grad_norm": 2.5373263776325343, + "learning_rate": 2.6013394041160094e-06, + "loss": 0.0529, + "step": 3340 + }, + { + "epoch": 2.4031648983995684, + "grad_norm": 3.969533419084874, + "learning_rate": 2.601108382016186e-06, + "loss": 0.0874, + "step": 3341 + }, + { + "epoch": 2.4038841934903794, + "grad_norm": 3.004498988729512, + "learning_rate": 2.600877303262404e-06, + "loss": 0.1078, + "step": 3342 + }, + { + "epoch": 2.4046034885811904, + "grad_norm": 1.196434756477466, + "learning_rate": 2.600646167866552e-06, + "loss": 0.0032, + "step": 3343 + }, + { + "epoch": 2.4053227836720015, + "grad_norm": 2.9129363951437566, + "learning_rate": 2.6004149758405226e-06, + "loss": 0.0223, + "step": 3344 + }, + { + "epoch": 2.4060420787628125, + "grad_norm": 2.1768567043038445, + "learning_rate": 2.60018372719621e-06, + "loss": 0.0426, + "step": 3345 + }, + { + "epoch": 2.4067613738536235, + "grad_norm": 1.3644177442622671, + "learning_rate": 2.5999524219455147e-06, + "loss": 0.0443, + "step": 3346 + }, + { + "epoch": 2.4074806689444346, + "grad_norm": 4.841823896802837, + "learning_rate": 2.5997210601003357e-06, + "loss": 0.0668, + "step": 3347 + }, + { + "epoch": 2.4081999640352456, + "grad_norm": 1.986431198879765, + "learning_rate": 2.599489641672578e-06, + "loss": 0.0057, + "step": 3348 + }, + { + "epoch": 2.4089192591260566, + "grad_norm": 0.10347057383346972, + "learning_rate": 2.599258166674148e-06, + "loss": 0.0008, + "step": 3349 + }, + { + "epoch": 2.4096385542168672, + "grad_norm": 3.195216799238045, + "learning_rate": 2.5990266351169555e-06, + "loss": 0.0861, + "step": 3350 + }, + { + "epoch": 2.4103578493076787, + "grad_norm": 3.543421611520562, + "learning_rate": 2.598795047012913e-06, + "loss": 0.1782, + "step": 3351 + }, + { + "epoch": 2.4110771443984893, + "grad_norm": 2.5107469882176856, + "learning_rate": 2.598563402373937e-06, + "loss": 0.0421, + "step": 3352 + }, + { + "epoch": 2.4117964394893003, + "grad_norm": 1.8087245629594786, + "learning_rate": 2.5983317012119457e-06, + "loss": 0.0421, + "step": 3353 + }, + { + "epoch": 2.4125157345801114, + "grad_norm": 0.8696523989624212, + "learning_rate": 2.59809994353886e-06, + "loss": 0.0074, + "step": 3354 + }, + { + "epoch": 2.4132350296709224, + "grad_norm": 5.185922568234041, + "learning_rate": 2.5978681293666044e-06, + "loss": 0.1517, + "step": 3355 + }, + { + "epoch": 2.4139543247617334, + "grad_norm": 3.288360429377135, + "learning_rate": 2.5976362587071064e-06, + "loss": 0.1172, + "step": 3356 + }, + { + "epoch": 2.4146736198525445, + "grad_norm": 5.316650332879753, + "learning_rate": 2.597404331572296e-06, + "loss": 0.087, + "step": 3357 + }, + { + "epoch": 2.4153929149433555, + "grad_norm": 3.4457847770267724, + "learning_rate": 2.5971723479741066e-06, + "loss": 0.087, + "step": 3358 + }, + { + "epoch": 2.4161122100341665, + "grad_norm": 3.91358756568543, + "learning_rate": 2.5969403079244732e-06, + "loss": 0.0395, + "step": 3359 + }, + { + "epoch": 2.4168315051249776, + "grad_norm": 2.3142443316994696, + "learning_rate": 2.5967082114353363e-06, + "loss": 0.0491, + "step": 3360 + }, + { + "epoch": 2.4175508002157886, + "grad_norm": 3.9008660445679424, + "learning_rate": 2.596476058518636e-06, + "loss": 0.1718, + "step": 3361 + }, + { + "epoch": 2.4182700953065996, + "grad_norm": 3.5717885099215683, + "learning_rate": 2.596243849186318e-06, + "loss": 0.0752, + "step": 3362 + }, + { + "epoch": 2.4189893903974107, + "grad_norm": 9.375595050626663, + "learning_rate": 2.5960115834503293e-06, + "loss": 0.0237, + "step": 3363 + }, + { + "epoch": 2.4197086854882217, + "grad_norm": 2.87364536482396, + "learning_rate": 2.5957792613226207e-06, + "loss": 0.0441, + "step": 3364 + }, + { + "epoch": 2.4204279805790323, + "grad_norm": 3.288896594087049, + "learning_rate": 2.5955468828151455e-06, + "loss": 0.109, + "step": 3365 + }, + { + "epoch": 2.421147275669844, + "grad_norm": 2.423450768463097, + "learning_rate": 2.5953144479398605e-06, + "loss": 0.0359, + "step": 3366 + }, + { + "epoch": 2.4218665707606544, + "grad_norm": 4.743707451739019, + "learning_rate": 2.5950819567087236e-06, + "loss": 0.1116, + "step": 3367 + }, + { + "epoch": 2.4225858658514654, + "grad_norm": 3.2977940980205696, + "learning_rate": 2.5948494091336983e-06, + "loss": 0.0195, + "step": 3368 + }, + { + "epoch": 2.4233051609422764, + "grad_norm": 6.405287817605597, + "learning_rate": 2.594616805226749e-06, + "loss": 0.0533, + "step": 3369 + }, + { + "epoch": 2.4240244560330875, + "grad_norm": 4.082338278555258, + "learning_rate": 2.594384144999843e-06, + "loss": 0.1438, + "step": 3370 + }, + { + "epoch": 2.4247437511238985, + "grad_norm": 2.1280398269173126, + "learning_rate": 2.5941514284649517e-06, + "loss": 0.0128, + "step": 3371 + }, + { + "epoch": 2.4254630462147095, + "grad_norm": 1.7504449581928783, + "learning_rate": 2.593918655634049e-06, + "loss": 0.0314, + "step": 3372 + }, + { + "epoch": 2.4261823413055206, + "grad_norm": 5.407934457696997, + "learning_rate": 2.5936858265191112e-06, + "loss": 0.1324, + "step": 3373 + }, + { + "epoch": 2.4269016363963316, + "grad_norm": 6.1129463466402685, + "learning_rate": 2.5934529411321173e-06, + "loss": 0.0471, + "step": 3374 + }, + { + "epoch": 2.4276209314871426, + "grad_norm": 1.3093890673171242, + "learning_rate": 2.5932199994850503e-06, + "loss": 0.015, + "step": 3375 + }, + { + "epoch": 2.4283402265779537, + "grad_norm": 2.750444920834696, + "learning_rate": 2.5929870015898955e-06, + "loss": 0.0895, + "step": 3376 + }, + { + "epoch": 2.4290595216687647, + "grad_norm": 3.4624809219481567, + "learning_rate": 2.5927539474586407e-06, + "loss": 0.0205, + "step": 3377 + }, + { + "epoch": 2.4297788167595757, + "grad_norm": 6.362491284606588, + "learning_rate": 2.5925208371032764e-06, + "loss": 0.0884, + "step": 3378 + }, + { + "epoch": 2.4304981118503868, + "grad_norm": 8.514175262125423, + "learning_rate": 2.592287670535798e-06, + "loss": 0.2373, + "step": 3379 + }, + { + "epoch": 2.431217406941198, + "grad_norm": 2.3367082217991233, + "learning_rate": 2.592054447768201e-06, + "loss": 0.0362, + "step": 3380 + }, + { + "epoch": 2.431936702032009, + "grad_norm": 7.921066853596047, + "learning_rate": 2.591821168812486e-06, + "loss": 0.0742, + "step": 3381 + }, + { + "epoch": 2.4326559971228194, + "grad_norm": 4.259824305418047, + "learning_rate": 2.5915878336806546e-06, + "loss": 0.0615, + "step": 3382 + }, + { + "epoch": 2.4333752922136305, + "grad_norm": 3.7321459764353127, + "learning_rate": 2.5913544423847135e-06, + "loss": 0.0887, + "step": 3383 + }, + { + "epoch": 2.4340945873044415, + "grad_norm": 3.062307487019876, + "learning_rate": 2.5911209949366704e-06, + "loss": 0.1347, + "step": 3384 + }, + { + "epoch": 2.4348138823952525, + "grad_norm": 1.6440660304177186, + "learning_rate": 2.5908874913485364e-06, + "loss": 0.0516, + "step": 3385 + }, + { + "epoch": 2.4355331774860636, + "grad_norm": 3.0905183111129633, + "learning_rate": 2.5906539316323265e-06, + "loss": 0.0684, + "step": 3386 + }, + { + "epoch": 2.4362524725768746, + "grad_norm": 3.850520551711745, + "learning_rate": 2.590420315800057e-06, + "loss": 0.0264, + "step": 3387 + }, + { + "epoch": 2.4369717676676856, + "grad_norm": 4.065710165088571, + "learning_rate": 2.5901866438637484e-06, + "loss": 0.0985, + "step": 3388 + }, + { + "epoch": 2.4376910627584967, + "grad_norm": 6.097092239218359, + "learning_rate": 2.589952915835423e-06, + "loss": 0.0885, + "step": 3389 + }, + { + "epoch": 2.4384103578493077, + "grad_norm": 2.5285060496203147, + "learning_rate": 2.5897191317271062e-06, + "loss": 0.0565, + "step": 3390 + }, + { + "epoch": 2.4391296529401187, + "grad_norm": 5.327229582029605, + "learning_rate": 2.5894852915508277e-06, + "loss": 0.1358, + "step": 3391 + }, + { + "epoch": 2.4398489480309298, + "grad_norm": 3.789754346722222, + "learning_rate": 2.589251395318618e-06, + "loss": 0.0603, + "step": 3392 + }, + { + "epoch": 2.440568243121741, + "grad_norm": 4.372608287586437, + "learning_rate": 2.5890174430425126e-06, + "loss": 0.0304, + "step": 3393 + }, + { + "epoch": 2.441287538212552, + "grad_norm": 5.267081450191482, + "learning_rate": 2.5887834347345473e-06, + "loss": 0.0488, + "step": 3394 + }, + { + "epoch": 2.442006833303363, + "grad_norm": 5.036738375049488, + "learning_rate": 2.5885493704067635e-06, + "loss": 0.1051, + "step": 3395 + }, + { + "epoch": 2.442726128394174, + "grad_norm": 6.614851470908532, + "learning_rate": 2.588315250071203e-06, + "loss": 0.182, + "step": 3396 + }, + { + "epoch": 2.4434454234849845, + "grad_norm": 3.506279227999809, + "learning_rate": 2.588081073739913e-06, + "loss": 0.0799, + "step": 3397 + }, + { + "epoch": 2.4441647185757955, + "grad_norm": 0.8487522398469651, + "learning_rate": 2.5878468414249416e-06, + "loss": 0.0022, + "step": 3398 + }, + { + "epoch": 2.4448840136666066, + "grad_norm": 4.75268044529781, + "learning_rate": 2.5876125531383405e-06, + "loss": 0.0647, + "step": 3399 + }, + { + "epoch": 2.4456033087574176, + "grad_norm": 1.9494331968068976, + "learning_rate": 2.5873782088921647e-06, + "loss": 0.0365, + "step": 3400 + }, + { + "epoch": 2.4463226038482286, + "grad_norm": 3.5028073114148173, + "learning_rate": 2.5871438086984704e-06, + "loss": 0.1169, + "step": 3401 + }, + { + "epoch": 2.4470418989390397, + "grad_norm": 7.151831254856234, + "learning_rate": 2.5869093525693194e-06, + "loss": 0.1337, + "step": 3402 + }, + { + "epoch": 2.4477611940298507, + "grad_norm": 2.1325537963916337, + "learning_rate": 2.5866748405167743e-06, + "loss": 0.0062, + "step": 3403 + }, + { + "epoch": 2.4484804891206617, + "grad_norm": 3.1658495566572937, + "learning_rate": 2.5864402725529002e-06, + "loss": 0.1524, + "step": 3404 + }, + { + "epoch": 2.4491997842114728, + "grad_norm": 2.449317908627163, + "learning_rate": 2.586205648689768e-06, + "loss": 0.0784, + "step": 3405 + }, + { + "epoch": 2.449919079302284, + "grad_norm": 2.8773663106903467, + "learning_rate": 2.5859709689394477e-06, + "loss": 0.0751, + "step": 3406 + }, + { + "epoch": 2.450638374393095, + "grad_norm": 4.412699307329153, + "learning_rate": 2.5857362333140147e-06, + "loss": 0.0961, + "step": 3407 + }, + { + "epoch": 2.451357669483906, + "grad_norm": 0.9175684358436597, + "learning_rate": 2.585501441825547e-06, + "loss": 0.0016, + "step": 3408 + }, + { + "epoch": 2.452076964574717, + "grad_norm": 3.7052084265086234, + "learning_rate": 2.585266594486124e-06, + "loss": 0.0676, + "step": 3409 + }, + { + "epoch": 2.452796259665528, + "grad_norm": 3.71279159033265, + "learning_rate": 2.5850316913078297e-06, + "loss": 0.02, + "step": 3410 + }, + { + "epoch": 2.453515554756339, + "grad_norm": 2.2857494777406204, + "learning_rate": 2.58479673230275e-06, + "loss": 0.0719, + "step": 3411 + }, + { + "epoch": 2.4542348498471496, + "grad_norm": 4.196155481203113, + "learning_rate": 2.584561717482975e-06, + "loss": 0.1438, + "step": 3412 + }, + { + "epoch": 2.454954144937961, + "grad_norm": 4.694625183649557, + "learning_rate": 2.5843266468605946e-06, + "loss": 0.1425, + "step": 3413 + }, + { + "epoch": 2.4556734400287716, + "grad_norm": 4.401462365748412, + "learning_rate": 2.5840915204477057e-06, + "loss": 0.1802, + "step": 3414 + }, + { + "epoch": 2.4563927351195827, + "grad_norm": 5.637995206520162, + "learning_rate": 2.5838563382564042e-06, + "loss": 0.0715, + "step": 3415 + }, + { + "epoch": 2.4571120302103937, + "grad_norm": 0.2422150793303434, + "learning_rate": 2.583621100298792e-06, + "loss": 0.0009, + "step": 3416 + }, + { + "epoch": 2.4578313253012047, + "grad_norm": 4.403015170691564, + "learning_rate": 2.583385806586972e-06, + "loss": 0.1129, + "step": 3417 + }, + { + "epoch": 2.4585506203920158, + "grad_norm": 1.161711401840374, + "learning_rate": 2.58315045713305e-06, + "loss": 0.0022, + "step": 3418 + }, + { + "epoch": 2.459269915482827, + "grad_norm": 4.644936655558416, + "learning_rate": 2.582915051949136e-06, + "loss": 0.0829, + "step": 3419 + }, + { + "epoch": 2.459989210573638, + "grad_norm": 4.809733004354658, + "learning_rate": 2.5826795910473416e-06, + "loss": 0.0883, + "step": 3420 + }, + { + "epoch": 2.460708505664449, + "grad_norm": 3.253446473659127, + "learning_rate": 2.5824440744397812e-06, + "loss": 0.0664, + "step": 3421 + }, + { + "epoch": 2.46142780075526, + "grad_norm": 7.668360686525356, + "learning_rate": 2.5822085021385735e-06, + "loss": 0.221, + "step": 3422 + }, + { + "epoch": 2.462147095846071, + "grad_norm": 7.039733166596783, + "learning_rate": 2.581972874155838e-06, + "loss": 0.0302, + "step": 3423 + }, + { + "epoch": 2.462866390936882, + "grad_norm": 1.3166476076117712, + "learning_rate": 2.5817371905036994e-06, + "loss": 0.0042, + "step": 3424 + }, + { + "epoch": 2.463585686027693, + "grad_norm": 9.798752074651636, + "learning_rate": 2.5815014511942827e-06, + "loss": 0.1244, + "step": 3425 + }, + { + "epoch": 2.464304981118504, + "grad_norm": 7.067742758794416, + "learning_rate": 2.5812656562397185e-06, + "loss": 0.1939, + "step": 3426 + }, + { + "epoch": 2.4650242762093146, + "grad_norm": 5.559221270898104, + "learning_rate": 2.581029805652138e-06, + "loss": 0.2307, + "step": 3427 + }, + { + "epoch": 2.465743571300126, + "grad_norm": 3.596524679933892, + "learning_rate": 2.580793899443676e-06, + "loss": 0.1125, + "step": 3428 + }, + { + "epoch": 2.4664628663909367, + "grad_norm": 5.028566565099954, + "learning_rate": 2.5805579376264716e-06, + "loss": 0.1277, + "step": 3429 + }, + { + "epoch": 2.4671821614817477, + "grad_norm": 4.961126121193562, + "learning_rate": 2.5803219202126633e-06, + "loss": 0.2435, + "step": 3430 + }, + { + "epoch": 2.4679014565725588, + "grad_norm": 4.2849133828681385, + "learning_rate": 2.5800858472143964e-06, + "loss": 0.1018, + "step": 3431 + }, + { + "epoch": 2.46862075166337, + "grad_norm": 4.244787306340676, + "learning_rate": 2.579849718643816e-06, + "loss": 0.1088, + "step": 3432 + }, + { + "epoch": 2.469340046754181, + "grad_norm": 1.6101625533924997, + "learning_rate": 2.579613534513073e-06, + "loss": 0.0209, + "step": 3433 + }, + { + "epoch": 2.470059341844992, + "grad_norm": 1.5374963902010577, + "learning_rate": 2.5793772948343175e-06, + "loss": 0.0025, + "step": 3434 + }, + { + "epoch": 2.470778636935803, + "grad_norm": 4.73898260240712, + "learning_rate": 2.5791409996197056e-06, + "loss": 0.0694, + "step": 3435 + }, + { + "epoch": 2.471497932026614, + "grad_norm": 3.4592227554468855, + "learning_rate": 2.578904648881395e-06, + "loss": 0.0908, + "step": 3436 + }, + { + "epoch": 2.472217227117425, + "grad_norm": 2.872259587967067, + "learning_rate": 2.5786682426315463e-06, + "loss": 0.0576, + "step": 3437 + }, + { + "epoch": 2.472936522208236, + "grad_norm": 3.8494832459425736, + "learning_rate": 2.578431780882323e-06, + "loss": 0.0843, + "step": 3438 + }, + { + "epoch": 2.473655817299047, + "grad_norm": 7.9149608087806, + "learning_rate": 2.5781952636458914e-06, + "loss": 0.1545, + "step": 3439 + }, + { + "epoch": 2.474375112389858, + "grad_norm": 3.767159575672662, + "learning_rate": 2.577958690934421e-06, + "loss": 0.1623, + "step": 3440 + }, + { + "epoch": 2.475094407480669, + "grad_norm": 2.568868154022677, + "learning_rate": 2.577722062760083e-06, + "loss": 0.0649, + "step": 3441 + }, + { + "epoch": 2.4758137025714797, + "grad_norm": 2.89387852458236, + "learning_rate": 2.577485379135054e-06, + "loss": 0.0783, + "step": 3442 + }, + { + "epoch": 2.476532997662291, + "grad_norm": 2.8730866278137426, + "learning_rate": 2.57724864007151e-06, + "loss": 0.0619, + "step": 3443 + }, + { + "epoch": 2.4772522927531018, + "grad_norm": 7.879335102715933, + "learning_rate": 2.5770118455816323e-06, + "loss": 0.0585, + "step": 3444 + }, + { + "epoch": 2.477971587843913, + "grad_norm": 3.8749099118810038, + "learning_rate": 2.576774995677605e-06, + "loss": 0.0965, + "step": 3445 + }, + { + "epoch": 2.478690882934724, + "grad_norm": 4.033621101909721, + "learning_rate": 2.5765380903716134e-06, + "loss": 0.1124, + "step": 3446 + }, + { + "epoch": 2.479410178025535, + "grad_norm": 4.765948368994288, + "learning_rate": 2.576301129675848e-06, + "loss": 0.1809, + "step": 3447 + }, + { + "epoch": 2.480129473116346, + "grad_norm": 3.689093766643328, + "learning_rate": 2.5760641136024996e-06, + "loss": 0.1232, + "step": 3448 + }, + { + "epoch": 2.480848768207157, + "grad_norm": 3.3345309188009136, + "learning_rate": 2.5758270421637637e-06, + "loss": 0.0344, + "step": 3449 + }, + { + "epoch": 2.481568063297968, + "grad_norm": 5.93760728144874, + "learning_rate": 2.575589915371838e-06, + "loss": 0.1048, + "step": 3450 + }, + { + "epoch": 2.482287358388779, + "grad_norm": 4.8695160974900435, + "learning_rate": 2.575352733238923e-06, + "loss": 0.2506, + "step": 3451 + }, + { + "epoch": 2.48300665347959, + "grad_norm": 1.622523380034065, + "learning_rate": 2.575115495777222e-06, + "loss": 0.0275, + "step": 3452 + }, + { + "epoch": 2.483725948570401, + "grad_norm": 0.08692547003772613, + "learning_rate": 2.5748782029989416e-06, + "loss": 0.0003, + "step": 3453 + }, + { + "epoch": 2.484445243661212, + "grad_norm": 7.743798592859323, + "learning_rate": 2.5746408549162907e-06, + "loss": 0.2036, + "step": 3454 + }, + { + "epoch": 2.485164538752023, + "grad_norm": 2.971061650563225, + "learning_rate": 2.574403451541481e-06, + "loss": 0.0563, + "step": 3455 + }, + { + "epoch": 2.485883833842834, + "grad_norm": 2.948099268300286, + "learning_rate": 2.5741659928867285e-06, + "loss": 0.0156, + "step": 3456 + }, + { + "epoch": 2.486603128933645, + "grad_norm": 0.48934136727654826, + "learning_rate": 2.5739284789642494e-06, + "loss": 0.0016, + "step": 3457 + }, + { + "epoch": 2.4873224240244562, + "grad_norm": 0.2350650483100412, + "learning_rate": 2.5736909097862647e-06, + "loss": 0.0007, + "step": 3458 + }, + { + "epoch": 2.488041719115267, + "grad_norm": 8.544466973534586, + "learning_rate": 2.573453285364998e-06, + "loss": 0.3311, + "step": 3459 + }, + { + "epoch": 2.488761014206078, + "grad_norm": 1.0104081536986096, + "learning_rate": 2.5732156057126757e-06, + "loss": 0.0119, + "step": 3460 + }, + { + "epoch": 2.489480309296889, + "grad_norm": 3.692569602083198, + "learning_rate": 2.572977870841526e-06, + "loss": 0.1327, + "step": 3461 + }, + { + "epoch": 2.4901996043877, + "grad_norm": 5.525408530482962, + "learning_rate": 2.572740080763782e-06, + "loss": 0.2189, + "step": 3462 + }, + { + "epoch": 2.490918899478511, + "grad_norm": 2.956617739630439, + "learning_rate": 2.5725022354916773e-06, + "loss": 0.0761, + "step": 3463 + }, + { + "epoch": 2.491638194569322, + "grad_norm": 4.020900199064354, + "learning_rate": 2.57226433503745e-06, + "loss": 0.1782, + "step": 3464 + }, + { + "epoch": 2.492357489660133, + "grad_norm": 2.5759569122013466, + "learning_rate": 2.5720263794133404e-06, + "loss": 0.054, + "step": 3465 + }, + { + "epoch": 2.493076784750944, + "grad_norm": 2.789070835502873, + "learning_rate": 2.5717883686315916e-06, + "loss": 0.0527, + "step": 3466 + }, + { + "epoch": 2.493796079841755, + "grad_norm": 3.625469161888554, + "learning_rate": 2.5715503027044494e-06, + "loss": 0.0585, + "step": 3467 + }, + { + "epoch": 2.494515374932566, + "grad_norm": 1.34999479950769, + "learning_rate": 2.5713121816441634e-06, + "loss": 0.0042, + "step": 3468 + }, + { + "epoch": 2.495234670023377, + "grad_norm": 3.7664437012124705, + "learning_rate": 2.571074005462985e-06, + "loss": 0.1147, + "step": 3469 + }, + { + "epoch": 2.495953965114188, + "grad_norm": 2.311293013704647, + "learning_rate": 2.5708357741731687e-06, + "loss": 0.0706, + "step": 3470 + }, + { + "epoch": 2.4966732602049992, + "grad_norm": 2.36359780466598, + "learning_rate": 2.5705974877869723e-06, + "loss": 0.0592, + "step": 3471 + }, + { + "epoch": 2.4973925552958103, + "grad_norm": 4.831511177319388, + "learning_rate": 2.570359146316656e-06, + "loss": 0.1229, + "step": 3472 + }, + { + "epoch": 2.4981118503866213, + "grad_norm": 4.685474798416406, + "learning_rate": 2.570120749774482e-06, + "loss": 0.1808, + "step": 3473 + }, + { + "epoch": 2.498831145477432, + "grad_norm": 2.964450738747615, + "learning_rate": 2.5698822981727173e-06, + "loss": 0.1165, + "step": 3474 + }, + { + "epoch": 2.499550440568243, + "grad_norm": 1.8199873753847282, + "learning_rate": 2.5696437915236303e-06, + "loss": 0.0557, + "step": 3475 + }, + { + "epoch": 2.500269735659054, + "grad_norm": 3.637782794830298, + "learning_rate": 2.5694052298394924e-06, + "loss": 0.0701, + "step": 3476 + }, + { + "epoch": 2.500989030749865, + "grad_norm": 5.478374569787154, + "learning_rate": 2.569166613132578e-06, + "loss": 0.0733, + "step": 3477 + }, + { + "epoch": 2.501708325840676, + "grad_norm": 2.090721205101083, + "learning_rate": 2.5689279414151644e-06, + "loss": 0.0494, + "step": 3478 + }, + { + "epoch": 2.502427620931487, + "grad_norm": 2.428853089569013, + "learning_rate": 2.568689214699532e-06, + "loss": 0.0583, + "step": 3479 + }, + { + "epoch": 2.503146916022298, + "grad_norm": 5.760256976194934, + "learning_rate": 2.5684504329979635e-06, + "loss": 0.0966, + "step": 3480 + }, + { + "epoch": 2.503866211113109, + "grad_norm": 5.131107741427995, + "learning_rate": 2.5682115963227445e-06, + "loss": 0.0426, + "step": 3481 + }, + { + "epoch": 2.50458550620392, + "grad_norm": 2.831223196978667, + "learning_rate": 2.567972704686164e-06, + "loss": 0.0337, + "step": 3482 + }, + { + "epoch": 2.505304801294731, + "grad_norm": 3.4787747516993868, + "learning_rate": 2.5677337581005126e-06, + "loss": 0.1199, + "step": 3483 + }, + { + "epoch": 2.5060240963855422, + "grad_norm": 2.7148804193958016, + "learning_rate": 2.5674947565780856e-06, + "loss": 0.091, + "step": 3484 + }, + { + "epoch": 2.5067433914763533, + "grad_norm": 4.723520996312077, + "learning_rate": 2.567255700131179e-06, + "loss": 0.2543, + "step": 3485 + }, + { + "epoch": 2.5074626865671643, + "grad_norm": 4.047217775691797, + "learning_rate": 2.567016588772093e-06, + "loss": 0.0724, + "step": 3486 + }, + { + "epoch": 2.5081819816579753, + "grad_norm": 5.086265238121583, + "learning_rate": 2.56677742251313e-06, + "loss": 0.1988, + "step": 3487 + }, + { + "epoch": 2.5089012767487864, + "grad_norm": 1.4415772066784551, + "learning_rate": 2.566538201366597e-06, + "loss": 0.048, + "step": 3488 + }, + { + "epoch": 2.509620571839597, + "grad_norm": 3.5788341161110004, + "learning_rate": 2.566298925344801e-06, + "loss": 0.0518, + "step": 3489 + }, + { + "epoch": 2.5103398669304084, + "grad_norm": 2.4690328057732307, + "learning_rate": 2.5660595944600533e-06, + "loss": 0.106, + "step": 3490 + }, + { + "epoch": 2.511059162021219, + "grad_norm": 2.7122239813943496, + "learning_rate": 2.565820208724668e-06, + "loss": 0.0858, + "step": 3491 + }, + { + "epoch": 2.51177845711203, + "grad_norm": 1.432113821623591, + "learning_rate": 2.565580768150962e-06, + "loss": 0.0092, + "step": 3492 + }, + { + "epoch": 2.512497752202841, + "grad_norm": 5.811480143422191, + "learning_rate": 2.565341272751255e-06, + "loss": 0.2093, + "step": 3493 + }, + { + "epoch": 2.513217047293652, + "grad_norm": 1.8972828138963855, + "learning_rate": 2.5651017225378692e-06, + "loss": 0.0043, + "step": 3494 + }, + { + "epoch": 2.513936342384463, + "grad_norm": 8.60955408438331, + "learning_rate": 2.5648621175231304e-06, + "loss": 0.1093, + "step": 3495 + }, + { + "epoch": 2.514655637475274, + "grad_norm": 5.175124785751448, + "learning_rate": 2.564622457719366e-06, + "loss": 0.1054, + "step": 3496 + }, + { + "epoch": 2.5153749325660852, + "grad_norm": 4.668449988014636, + "learning_rate": 2.564382743138907e-06, + "loss": 0.24, + "step": 3497 + }, + { + "epoch": 2.5160942276568963, + "grad_norm": 5.170177622199164, + "learning_rate": 2.564142973794088e-06, + "loss": 0.2063, + "step": 3498 + }, + { + "epoch": 2.5168135227477073, + "grad_norm": 2.9388821711037716, + "learning_rate": 2.563903149697245e-06, + "loss": 0.0764, + "step": 3499 + }, + { + "epoch": 2.5175328178385183, + "grad_norm": 2.9821945593432058, + "learning_rate": 2.563663270860717e-06, + "loss": 0.0143, + "step": 3500 + }, + { + "epoch": 2.5182521129293294, + "grad_norm": 2.98017025394871, + "learning_rate": 2.5634233372968467e-06, + "loss": 0.1017, + "step": 3501 + }, + { + "epoch": 2.5189714080201404, + "grad_norm": 2.1879599259132387, + "learning_rate": 2.5631833490179786e-06, + "loss": 0.0238, + "step": 3502 + }, + { + "epoch": 2.5196907031109514, + "grad_norm": 2.193676767115623, + "learning_rate": 2.562943306036461e-06, + "loss": 0.034, + "step": 3503 + }, + { + "epoch": 2.520409998201762, + "grad_norm": 2.531547326602505, + "learning_rate": 2.562703208364644e-06, + "loss": 0.0473, + "step": 3504 + }, + { + "epoch": 2.5211292932925735, + "grad_norm": 4.256083829310361, + "learning_rate": 2.5624630560148826e-06, + "loss": 0.0634, + "step": 3505 + }, + { + "epoch": 2.521848588383384, + "grad_norm": 3.824612074522197, + "learning_rate": 2.5622228489995307e-06, + "loss": 0.1111, + "step": 3506 + }, + { + "epoch": 2.522567883474195, + "grad_norm": 3.566015620133227, + "learning_rate": 2.561982587330949e-06, + "loss": 0.1917, + "step": 3507 + }, + { + "epoch": 2.523287178565006, + "grad_norm": 4.52746314624229, + "learning_rate": 2.5617422710214986e-06, + "loss": 0.054, + "step": 3508 + }, + { + "epoch": 2.524006473655817, + "grad_norm": 3.674946033888536, + "learning_rate": 2.561501900083545e-06, + "loss": 0.0318, + "step": 3509 + }, + { + "epoch": 2.5247257687466282, + "grad_norm": 2.9951051769960966, + "learning_rate": 2.561261474529455e-06, + "loss": 0.0351, + "step": 3510 + }, + { + "epoch": 2.5254450638374393, + "grad_norm": 3.301985966737383, + "learning_rate": 2.561020994371599e-06, + "loss": 0.0582, + "step": 3511 + }, + { + "epoch": 2.5261643589282503, + "grad_norm": 2.4413480614112237, + "learning_rate": 2.5607804596223503e-06, + "loss": 0.044, + "step": 3512 + }, + { + "epoch": 2.5268836540190613, + "grad_norm": 2.1213398118727937, + "learning_rate": 2.5605398702940855e-06, + "loss": 0.0701, + "step": 3513 + }, + { + "epoch": 2.5276029491098724, + "grad_norm": 1.4305166774404456, + "learning_rate": 2.560299226399182e-06, + "loss": 0.0021, + "step": 3514 + }, + { + "epoch": 2.5283222442006834, + "grad_norm": 2.2873225424740293, + "learning_rate": 2.560058527950022e-06, + "loss": 0.0548, + "step": 3515 + }, + { + "epoch": 2.5290415392914944, + "grad_norm": 4.070064242254187, + "learning_rate": 2.55981777495899e-06, + "loss": 0.0789, + "step": 3516 + }, + { + "epoch": 2.5297608343823055, + "grad_norm": 0.3847651824752757, + "learning_rate": 2.5595769674384725e-06, + "loss": 0.0015, + "step": 3517 + }, + { + "epoch": 2.5304801294731165, + "grad_norm": 5.730510961371827, + "learning_rate": 2.559336105400861e-06, + "loss": 0.2363, + "step": 3518 + }, + { + "epoch": 2.531199424563927, + "grad_norm": 8.84048825219756, + "learning_rate": 2.5590951888585468e-06, + "loss": 0.1629, + "step": 3519 + }, + { + "epoch": 2.5319187196547386, + "grad_norm": 0.7057526693714911, + "learning_rate": 2.558854217823926e-06, + "loss": 0.0015, + "step": 3520 + }, + { + "epoch": 2.532638014745549, + "grad_norm": 2.8979907444052038, + "learning_rate": 2.5586131923093965e-06, + "loss": 0.0778, + "step": 3521 + }, + { + "epoch": 2.53335730983636, + "grad_norm": 5.293105278693575, + "learning_rate": 2.5583721123273604e-06, + "loss": 0.0422, + "step": 3522 + }, + { + "epoch": 2.5340766049271712, + "grad_norm": 3.8822705868467535, + "learning_rate": 2.558130977890221e-06, + "loss": 0.0692, + "step": 3523 + }, + { + "epoch": 2.5347959000179823, + "grad_norm": 4.027518531992648, + "learning_rate": 2.5578897890103857e-06, + "loss": 0.2075, + "step": 3524 + }, + { + "epoch": 2.5355151951087933, + "grad_norm": 5.611138962024131, + "learning_rate": 2.557648545700263e-06, + "loss": 0.1925, + "step": 3525 + }, + { + "epoch": 2.5362344901996043, + "grad_norm": 5.283192825036582, + "learning_rate": 2.5574072479722663e-06, + "loss": 0.1808, + "step": 3526 + }, + { + "epoch": 2.5369537852904154, + "grad_norm": 3.2139419240967677, + "learning_rate": 2.5571658958388103e-06, + "loss": 0.086, + "step": 3527 + }, + { + "epoch": 2.5376730803812264, + "grad_norm": 1.60526938420732, + "learning_rate": 2.5569244893123136e-06, + "loss": 0.04, + "step": 3528 + }, + { + "epoch": 2.5383923754720374, + "grad_norm": 3.818629682478443, + "learning_rate": 2.5566830284051963e-06, + "loss": 0.0781, + "step": 3529 + }, + { + "epoch": 2.5391116705628485, + "grad_norm": 12.924028996405493, + "learning_rate": 2.5564415131298823e-06, + "loss": 0.207, + "step": 3530 + }, + { + "epoch": 2.5398309656536595, + "grad_norm": 1.4297078160043157, + "learning_rate": 2.556199943498798e-06, + "loss": 0.027, + "step": 3531 + }, + { + "epoch": 2.5405502607444705, + "grad_norm": 3.0880962668288987, + "learning_rate": 2.5559583195243725e-06, + "loss": 0.1166, + "step": 3532 + }, + { + "epoch": 2.5412695558352816, + "grad_norm": 7.185202767540575, + "learning_rate": 2.5557166412190373e-06, + "loss": 0.1743, + "step": 3533 + }, + { + "epoch": 2.541988850926092, + "grad_norm": 3.743089599588881, + "learning_rate": 2.5554749085952286e-06, + "loss": 0.1209, + "step": 3534 + }, + { + "epoch": 2.5427081460169036, + "grad_norm": 5.855815763094652, + "learning_rate": 2.555233121665382e-06, + "loss": 0.1971, + "step": 3535 + }, + { + "epoch": 2.5434274411077142, + "grad_norm": 0.8108980233085022, + "learning_rate": 2.5549912804419397e-06, + "loss": 0.0077, + "step": 3536 + }, + { + "epoch": 2.5441467361985257, + "grad_norm": 6.867518261027968, + "learning_rate": 2.5547493849373435e-06, + "loss": 0.021, + "step": 3537 + }, + { + "epoch": 2.5448660312893363, + "grad_norm": 2.1523704520124536, + "learning_rate": 2.55450743516404e-06, + "loss": 0.0782, + "step": 3538 + }, + { + "epoch": 2.5455853263801473, + "grad_norm": 4.555569683926216, + "learning_rate": 2.554265431134478e-06, + "loss": 0.1947, + "step": 3539 + }, + { + "epoch": 2.5463046214709584, + "grad_norm": 3.025986828250512, + "learning_rate": 2.5540233728611086e-06, + "loss": 0.0392, + "step": 3540 + }, + { + "epoch": 2.5470239165617694, + "grad_norm": 4.721736906888111, + "learning_rate": 2.553781260356387e-06, + "loss": 0.0864, + "step": 3541 + }, + { + "epoch": 2.5477432116525804, + "grad_norm": 2.898133230771604, + "learning_rate": 2.553539093632769e-06, + "loss": 0.0931, + "step": 3542 + }, + { + "epoch": 2.5484625067433915, + "grad_norm": 4.790008065763163, + "learning_rate": 2.5532968727027153e-06, + "loss": 0.206, + "step": 3543 + }, + { + "epoch": 2.5491818018342025, + "grad_norm": 1.969083972083599, + "learning_rate": 2.5530545975786888e-06, + "loss": 0.007, + "step": 3544 + }, + { + "epoch": 2.5499010969250135, + "grad_norm": 2.1324070330305482, + "learning_rate": 2.5528122682731545e-06, + "loss": 0.0083, + "step": 3545 + }, + { + "epoch": 2.5506203920158246, + "grad_norm": 1.9699174578661331, + "learning_rate": 2.552569884798581e-06, + "loss": 0.037, + "step": 3546 + }, + { + "epoch": 2.5513396871066356, + "grad_norm": 4.775038969293058, + "learning_rate": 2.5523274471674388e-06, + "loss": 0.2031, + "step": 3547 + }, + { + "epoch": 2.5520589821974466, + "grad_norm": 1.5812838241336613, + "learning_rate": 2.5520849553922023e-06, + "loss": 0.0329, + "step": 3548 + }, + { + "epoch": 2.5527782772882572, + "grad_norm": 6.37634727568618, + "learning_rate": 2.5518424094853485e-06, + "loss": 0.1274, + "step": 3549 + }, + { + "epoch": 2.5534975723790687, + "grad_norm": 6.306381648339605, + "learning_rate": 2.5515998094593556e-06, + "loss": 0.2527, + "step": 3550 + }, + { + "epoch": 2.5542168674698793, + "grad_norm": 2.061940813255708, + "learning_rate": 2.551357155326707e-06, + "loss": 0.0251, + "step": 3551 + }, + { + "epoch": 2.554936162560691, + "grad_norm": 3.3143704393288136, + "learning_rate": 2.551114447099887e-06, + "loss": 0.0125, + "step": 3552 + }, + { + "epoch": 2.5556554576515014, + "grad_norm": 2.5576226238416546, + "learning_rate": 2.550871684791383e-06, + "loss": 0.0625, + "step": 3553 + }, + { + "epoch": 2.5563747527423124, + "grad_norm": 2.324275910227882, + "learning_rate": 2.5506288684136865e-06, + "loss": 0.0491, + "step": 3554 + }, + { + "epoch": 2.5570940478331234, + "grad_norm": 2.1751372224957803, + "learning_rate": 2.550385997979291e-06, + "loss": 0.0637, + "step": 3555 + }, + { + "epoch": 2.5578133429239345, + "grad_norm": 5.3770796886050425, + "learning_rate": 2.5501430735006907e-06, + "loss": 0.1733, + "step": 3556 + }, + { + "epoch": 2.5585326380147455, + "grad_norm": 6.289112856011835, + "learning_rate": 2.5499000949903866e-06, + "loss": 0.1315, + "step": 3557 + }, + { + "epoch": 2.5592519331055565, + "grad_norm": 3.4819626801396617, + "learning_rate": 2.5496570624608793e-06, + "loss": 0.0344, + "step": 3558 + }, + { + "epoch": 2.5599712281963676, + "grad_norm": 2.673207582988735, + "learning_rate": 2.5494139759246738e-06, + "loss": 0.0975, + "step": 3559 + }, + { + "epoch": 2.5606905232871786, + "grad_norm": 4.6567109007560905, + "learning_rate": 2.549170835394277e-06, + "loss": 0.165, + "step": 3560 + }, + { + "epoch": 2.5614098183779896, + "grad_norm": 2.3764473821705425, + "learning_rate": 2.548927640882199e-06, + "loss": 0.0521, + "step": 3561 + }, + { + "epoch": 2.5621291134688007, + "grad_norm": 3.401520100863981, + "learning_rate": 2.5486843924009528e-06, + "loss": 0.1057, + "step": 3562 + }, + { + "epoch": 2.5628484085596117, + "grad_norm": 3.580039724768081, + "learning_rate": 2.548441089963053e-06, + "loss": 0.0448, + "step": 3563 + }, + { + "epoch": 2.5635677036504227, + "grad_norm": 5.1496752627595095, + "learning_rate": 2.548197733581019e-06, + "loss": 0.1372, + "step": 3564 + }, + { + "epoch": 2.564286998741234, + "grad_norm": 3.0066521239283763, + "learning_rate": 2.5479543232673715e-06, + "loss": 0.0194, + "step": 3565 + }, + { + "epoch": 2.5650062938320444, + "grad_norm": 2.6907702309665376, + "learning_rate": 2.5477108590346346e-06, + "loss": 0.0163, + "step": 3566 + }, + { + "epoch": 2.565725588922856, + "grad_norm": 2.952590666270628, + "learning_rate": 2.5474673408953344e-06, + "loss": 0.049, + "step": 3567 + }, + { + "epoch": 2.5664448840136664, + "grad_norm": 1.5528305246308567, + "learning_rate": 2.547223768862001e-06, + "loss": 0.0258, + "step": 3568 + }, + { + "epoch": 2.5671641791044775, + "grad_norm": 3.2861827407112476, + "learning_rate": 2.5469801429471667e-06, + "loss": 0.0542, + "step": 3569 + }, + { + "epoch": 2.5678834741952885, + "grad_norm": 3.987059589041141, + "learning_rate": 2.546736463163366e-06, + "loss": 0.1344, + "step": 3570 + }, + { + "epoch": 2.5686027692860995, + "grad_norm": 4.0469800427662985, + "learning_rate": 2.5464927295231367e-06, + "loss": 0.0384, + "step": 3571 + }, + { + "epoch": 2.5693220643769106, + "grad_norm": 3.687294745418768, + "learning_rate": 2.5462489420390187e-06, + "loss": 0.0492, + "step": 3572 + }, + { + "epoch": 2.5700413594677216, + "grad_norm": 3.350669057530878, + "learning_rate": 2.546005100723557e-06, + "loss": 0.0305, + "step": 3573 + }, + { + "epoch": 2.5707606545585326, + "grad_norm": 3.0989712046710984, + "learning_rate": 2.545761205589296e-06, + "loss": 0.0598, + "step": 3574 + }, + { + "epoch": 2.5714799496493437, + "grad_norm": 3.1113822826441266, + "learning_rate": 2.5455172566487857e-06, + "loss": 0.0717, + "step": 3575 + }, + { + "epoch": 2.5721992447401547, + "grad_norm": 4.247450222425722, + "learning_rate": 2.545273253914577e-06, + "loss": 0.0451, + "step": 3576 + }, + { + "epoch": 2.5729185398309657, + "grad_norm": 4.344623508312981, + "learning_rate": 2.5450291973992245e-06, + "loss": 0.0803, + "step": 3577 + }, + { + "epoch": 2.573637834921777, + "grad_norm": 6.0443315182528865, + "learning_rate": 2.544785087115285e-06, + "loss": 0.2302, + "step": 3578 + }, + { + "epoch": 2.574357130012588, + "grad_norm": 6.532146424687256, + "learning_rate": 2.5445409230753193e-06, + "loss": 0.3271, + "step": 3579 + }, + { + "epoch": 2.575076425103399, + "grad_norm": 3.3577906606487242, + "learning_rate": 2.544296705291889e-06, + "loss": 0.1773, + "step": 3580 + }, + { + "epoch": 2.5757957201942094, + "grad_norm": 2.2909592260368346, + "learning_rate": 2.54405243377756e-06, + "loss": 0.0353, + "step": 3581 + }, + { + "epoch": 2.576515015285021, + "grad_norm": 2.5766040408158126, + "learning_rate": 2.5438081085449004e-06, + "loss": 0.0467, + "step": 3582 + }, + { + "epoch": 2.5772343103758315, + "grad_norm": 5.207479191201509, + "learning_rate": 2.5435637296064816e-06, + "loss": 0.1597, + "step": 3583 + }, + { + "epoch": 2.5779536054666425, + "grad_norm": 4.418553524417093, + "learning_rate": 2.543319296974877e-06, + "loss": 0.1587, + "step": 3584 + }, + { + "epoch": 2.5786729005574536, + "grad_norm": 4.0761017841284595, + "learning_rate": 2.5430748106626633e-06, + "loss": 0.1064, + "step": 3585 + }, + { + "epoch": 2.5793921956482646, + "grad_norm": 3.1550199749966246, + "learning_rate": 2.5428302706824192e-06, + "loss": 0.0458, + "step": 3586 + }, + { + "epoch": 2.5801114907390756, + "grad_norm": 5.958745985124681, + "learning_rate": 2.542585677046727e-06, + "loss": 0.0234, + "step": 3587 + }, + { + "epoch": 2.5808307858298867, + "grad_norm": 2.20615071082052, + "learning_rate": 2.5423410297681715e-06, + "loss": 0.0382, + "step": 3588 + }, + { + "epoch": 2.5815500809206977, + "grad_norm": 5.511574152943852, + "learning_rate": 2.5420963288593403e-06, + "loss": 0.2206, + "step": 3589 + }, + { + "epoch": 2.5822693760115087, + "grad_norm": 2.4973371450672865, + "learning_rate": 2.5418515743328234e-06, + "loss": 0.0186, + "step": 3590 + }, + { + "epoch": 2.5829886711023198, + "grad_norm": 4.51868552850043, + "learning_rate": 2.5416067662012145e-06, + "loss": 0.104, + "step": 3591 + }, + { + "epoch": 2.583707966193131, + "grad_norm": 2.3391649051777574, + "learning_rate": 2.541361904477109e-06, + "loss": 0.0168, + "step": 3592 + }, + { + "epoch": 2.584427261283942, + "grad_norm": 4.156181341218098, + "learning_rate": 2.541116989173105e-06, + "loss": 0.1084, + "step": 3593 + }, + { + "epoch": 2.585146556374753, + "grad_norm": 4.701650207067254, + "learning_rate": 2.5408720203018044e-06, + "loss": 0.1242, + "step": 3594 + }, + { + "epoch": 2.585865851465564, + "grad_norm": 1.9496321604718572, + "learning_rate": 2.5406269978758116e-06, + "loss": 0.0444, + "step": 3595 + }, + { + "epoch": 2.5865851465563745, + "grad_norm": 3.642812884068933, + "learning_rate": 2.5403819219077324e-06, + "loss": 0.0284, + "step": 3596 + }, + { + "epoch": 2.587304441647186, + "grad_norm": 3.717199735116792, + "learning_rate": 2.5401367924101775e-06, + "loss": 0.0954, + "step": 3597 + }, + { + "epoch": 2.5880237367379966, + "grad_norm": 3.8250986849617754, + "learning_rate": 2.5398916093957588e-06, + "loss": 0.1652, + "step": 3598 + }, + { + "epoch": 2.5887430318288076, + "grad_norm": 3.3550735046669327, + "learning_rate": 2.539646372877091e-06, + "loss": 0.1015, + "step": 3599 + }, + { + "epoch": 2.5894623269196186, + "grad_norm": 7.403908730681601, + "learning_rate": 2.539401082866792e-06, + "loss": 0.0419, + "step": 3600 + }, + { + "epoch": 2.5901816220104297, + "grad_norm": 4.458507419027894, + "learning_rate": 2.539155739377483e-06, + "loss": 0.1317, + "step": 3601 + }, + { + "epoch": 2.5909009171012407, + "grad_norm": 5.411549982522412, + "learning_rate": 2.5389103424217872e-06, + "loss": 0.1254, + "step": 3602 + }, + { + "epoch": 2.5916202121920517, + "grad_norm": 5.551914161242035, + "learning_rate": 2.5386648920123307e-06, + "loss": 0.1345, + "step": 3603 + }, + { + "epoch": 2.5923395072828628, + "grad_norm": 3.8350715945936757, + "learning_rate": 2.5384193881617415e-06, + "loss": 0.0654, + "step": 3604 + }, + { + "epoch": 2.593058802373674, + "grad_norm": 0.7387453956564594, + "learning_rate": 2.5381738308826526e-06, + "loss": 0.0078, + "step": 3605 + }, + { + "epoch": 2.593778097464485, + "grad_norm": 4.686956035330403, + "learning_rate": 2.537928220187697e-06, + "loss": 0.1908, + "step": 3606 + }, + { + "epoch": 2.594497392555296, + "grad_norm": 3.1634346554714035, + "learning_rate": 2.537682556089513e-06, + "loss": 0.1345, + "step": 3607 + }, + { + "epoch": 2.595216687646107, + "grad_norm": 4.155984034833928, + "learning_rate": 2.5374368386007395e-06, + "loss": 0.1816, + "step": 3608 + }, + { + "epoch": 2.595935982736918, + "grad_norm": 4.035613536214, + "learning_rate": 2.5371910677340196e-06, + "loss": 0.1298, + "step": 3609 + }, + { + "epoch": 2.596655277827729, + "grad_norm": 4.410615262037329, + "learning_rate": 2.5369452435019984e-06, + "loss": 0.0967, + "step": 3610 + }, + { + "epoch": 2.5973745729185396, + "grad_norm": 4.398167551036084, + "learning_rate": 2.5366993659173243e-06, + "loss": 0.1624, + "step": 3611 + }, + { + "epoch": 2.598093868009351, + "grad_norm": 4.094728965884214, + "learning_rate": 2.5364534349926477e-06, + "loss": 0.0854, + "step": 3612 + }, + { + "epoch": 2.5988131631001616, + "grad_norm": 4.6881330585398455, + "learning_rate": 2.536207450740623e-06, + "loss": 0.1006, + "step": 3613 + }, + { + "epoch": 2.599532458190973, + "grad_norm": 4.999955757286984, + "learning_rate": 2.5359614131739057e-06, + "loss": 0.0947, + "step": 3614 + }, + { + "epoch": 2.6002517532817837, + "grad_norm": 2.2982670107156236, + "learning_rate": 2.535715322305155e-06, + "loss": 0.0807, + "step": 3615 + }, + { + "epoch": 2.6009710483725947, + "grad_norm": 3.2221983666217793, + "learning_rate": 2.535469178147033e-06, + "loss": 0.1313, + "step": 3616 + }, + { + "epoch": 2.6016903434634058, + "grad_norm": 3.61813131024914, + "learning_rate": 2.5352229807122037e-06, + "loss": 0.1815, + "step": 3617 + }, + { + "epoch": 2.602409638554217, + "grad_norm": 3.7769570401644152, + "learning_rate": 2.5349767300133352e-06, + "loss": 0.0803, + "step": 3618 + }, + { + "epoch": 2.603128933645028, + "grad_norm": 5.318290462611663, + "learning_rate": 2.534730426063097e-06, + "loss": 0.0611, + "step": 3619 + }, + { + "epoch": 2.603848228735839, + "grad_norm": 5.459847473252826, + "learning_rate": 2.534484068874162e-06, + "loss": 0.0873, + "step": 3620 + }, + { + "epoch": 2.60456752382665, + "grad_norm": 0.6823841030323171, + "learning_rate": 2.5342376584592056e-06, + "loss": 0.0056, + "step": 3621 + }, + { + "epoch": 2.605286818917461, + "grad_norm": 2.684371990325371, + "learning_rate": 2.5339911948309057e-06, + "loss": 0.1071, + "step": 3622 + }, + { + "epoch": 2.606006114008272, + "grad_norm": 3.6864551386607904, + "learning_rate": 2.533744678001944e-06, + "loss": 0.0622, + "step": 3623 + }, + { + "epoch": 2.606725409099083, + "grad_norm": 1.6205701101000574, + "learning_rate": 2.533498107985004e-06, + "loss": 0.0269, + "step": 3624 + }, + { + "epoch": 2.607444704189894, + "grad_norm": 6.042361537084525, + "learning_rate": 2.5332514847927726e-06, + "loss": 0.1588, + "step": 3625 + }, + { + "epoch": 2.6081639992807046, + "grad_norm": 3.3956691276029956, + "learning_rate": 2.533004808437938e-06, + "loss": 0.0215, + "step": 3626 + }, + { + "epoch": 2.608883294371516, + "grad_norm": 1.8695545031916554, + "learning_rate": 2.532758078933192e-06, + "loss": 0.0139, + "step": 3627 + }, + { + "epoch": 2.6096025894623267, + "grad_norm": 5.4890198792463964, + "learning_rate": 2.532511296291231e-06, + "loss": 0.1388, + "step": 3628 + }, + { + "epoch": 2.610321884553138, + "grad_norm": 2.54397434864001, + "learning_rate": 2.532264460524751e-06, + "loss": 0.0354, + "step": 3629 + }, + { + "epoch": 2.6110411796439488, + "grad_norm": 2.8251835019619747, + "learning_rate": 2.532017571646452e-06, + "loss": 0.0334, + "step": 3630 + }, + { + "epoch": 2.61176047473476, + "grad_norm": 6.623189264211329, + "learning_rate": 2.5317706296690376e-06, + "loss": 0.0528, + "step": 3631 + }, + { + "epoch": 2.612479769825571, + "grad_norm": 6.746329951477701, + "learning_rate": 2.531523634605213e-06, + "loss": 0.1999, + "step": 3632 + }, + { + "epoch": 2.613199064916382, + "grad_norm": 5.7608315074917495, + "learning_rate": 2.531276586467687e-06, + "loss": 0.169, + "step": 3633 + }, + { + "epoch": 2.613918360007193, + "grad_norm": 4.979693590182001, + "learning_rate": 2.53102948526917e-06, + "loss": 0.2363, + "step": 3634 + }, + { + "epoch": 2.614637655098004, + "grad_norm": 2.9257103235188624, + "learning_rate": 2.5307823310223756e-06, + "loss": 0.0452, + "step": 3635 + }, + { + "epoch": 2.615356950188815, + "grad_norm": 2.6623679283048407, + "learning_rate": 2.530535123740021e-06, + "loss": 0.0491, + "step": 3636 + }, + { + "epoch": 2.616076245279626, + "grad_norm": 3.7123501371943477, + "learning_rate": 2.5302878634348255e-06, + "loss": 0.0608, + "step": 3637 + }, + { + "epoch": 2.616795540370437, + "grad_norm": 3.5322104196983006, + "learning_rate": 2.5300405501195107e-06, + "loss": 0.019, + "step": 3638 + }, + { + "epoch": 2.617514835461248, + "grad_norm": 4.9593428058442255, + "learning_rate": 2.5297931838068013e-06, + "loss": 0.4764, + "step": 3639 + }, + { + "epoch": 2.618234130552059, + "grad_norm": 5.775091458388362, + "learning_rate": 2.529545764509425e-06, + "loss": 0.0493, + "step": 3640 + }, + { + "epoch": 2.61895342564287, + "grad_norm": 5.51211232126548, + "learning_rate": 2.5292982922401116e-06, + "loss": 0.1193, + "step": 3641 + }, + { + "epoch": 2.619672720733681, + "grad_norm": 4.183424246739544, + "learning_rate": 2.5290507670115945e-06, + "loss": 0.0862, + "step": 3642 + }, + { + "epoch": 2.6203920158244918, + "grad_norm": 6.2689336496148425, + "learning_rate": 2.528803188836609e-06, + "loss": 0.2448, + "step": 3643 + }, + { + "epoch": 2.6211113109153032, + "grad_norm": 4.683188815343918, + "learning_rate": 2.528555557727894e-06, + "loss": 0.1217, + "step": 3644 + }, + { + "epoch": 2.621830606006114, + "grad_norm": 3.779198395109637, + "learning_rate": 2.528307873698189e-06, + "loss": 0.1743, + "step": 3645 + }, + { + "epoch": 2.622549901096925, + "grad_norm": 4.749523536449059, + "learning_rate": 2.528060136760239e-06, + "loss": 0.1401, + "step": 3646 + }, + { + "epoch": 2.623269196187736, + "grad_norm": 4.608089140342495, + "learning_rate": 2.5278123469267903e-06, + "loss": 0.0204, + "step": 3647 + }, + { + "epoch": 2.623988491278547, + "grad_norm": 0.8244730202703748, + "learning_rate": 2.5275645042105916e-06, + "loss": 0.0013, + "step": 3648 + }, + { + "epoch": 2.624707786369358, + "grad_norm": 4.554578329634901, + "learning_rate": 2.527316608624396e-06, + "loss": 0.1309, + "step": 3649 + }, + { + "epoch": 2.625427081460169, + "grad_norm": 1.8678742867034965, + "learning_rate": 2.5270686601809577e-06, + "loss": 0.0478, + "step": 3650 + }, + { + "epoch": 2.62614637655098, + "grad_norm": 4.25204586476136, + "learning_rate": 2.526820658893033e-06, + "loss": 0.0722, + "step": 3651 + }, + { + "epoch": 2.626865671641791, + "grad_norm": 0.7705422900817839, + "learning_rate": 2.5265726047733837e-06, + "loss": 0.0067, + "step": 3652 + }, + { + "epoch": 2.627584966732602, + "grad_norm": 4.4015151038859734, + "learning_rate": 2.526324497834771e-06, + "loss": 0.2269, + "step": 3653 + }, + { + "epoch": 2.628304261823413, + "grad_norm": 3.428253962330576, + "learning_rate": 2.5260763380899614e-06, + "loss": 0.1341, + "step": 3654 + }, + { + "epoch": 2.629023556914224, + "grad_norm": 2.745567761914923, + "learning_rate": 2.525828125551724e-06, + "loss": 0.0807, + "step": 3655 + }, + { + "epoch": 2.629742852005035, + "grad_norm": 2.09935528166114, + "learning_rate": 2.5255798602328276e-06, + "loss": 0.0307, + "step": 3656 + }, + { + "epoch": 2.6304621470958462, + "grad_norm": 5.762636952307489, + "learning_rate": 2.5253315421460476e-06, + "loss": 0.2204, + "step": 3657 + }, + { + "epoch": 2.631181442186657, + "grad_norm": 4.301575920146319, + "learning_rate": 2.5250831713041596e-06, + "loss": 0.1691, + "step": 3658 + }, + { + "epoch": 2.6319007372774683, + "grad_norm": 0.165754111986202, + "learning_rate": 2.524834747719943e-06, + "loss": 0.0005, + "step": 3659 + }, + { + "epoch": 2.632620032368279, + "grad_norm": 4.2900920314099675, + "learning_rate": 2.52458627140618e-06, + "loss": 0.0677, + "step": 3660 + }, + { + "epoch": 2.63333932745909, + "grad_norm": 4.390895562820762, + "learning_rate": 2.524337742375654e-06, + "loss": 0.1087, + "step": 3661 + }, + { + "epoch": 2.634058622549901, + "grad_norm": 1.2567750183294795, + "learning_rate": 2.5240891606411533e-06, + "loss": 0.0053, + "step": 3662 + }, + { + "epoch": 2.634777917640712, + "grad_norm": 3.5118046317122094, + "learning_rate": 2.523840526215468e-06, + "loss": 0.1538, + "step": 3663 + }, + { + "epoch": 2.635497212731523, + "grad_norm": 3.812563811661294, + "learning_rate": 2.5235918391113895e-06, + "loss": 0.1311, + "step": 3664 + }, + { + "epoch": 2.636216507822334, + "grad_norm": 2.3299051593230944, + "learning_rate": 2.523343099341715e-06, + "loss": 0.0372, + "step": 3665 + }, + { + "epoch": 2.636935802913145, + "grad_norm": 0.5680373022103768, + "learning_rate": 2.5230943069192407e-06, + "loss": 0.0047, + "step": 3666 + }, + { + "epoch": 2.637655098003956, + "grad_norm": 0.99127957395232, + "learning_rate": 2.522845461856769e-06, + "loss": 0.007, + "step": 3667 + }, + { + "epoch": 2.638374393094767, + "grad_norm": 2.4659282685610586, + "learning_rate": 2.5225965641671027e-06, + "loss": 0.0308, + "step": 3668 + }, + { + "epoch": 2.639093688185578, + "grad_norm": 2.5579028925216583, + "learning_rate": 2.522347613863047e-06, + "loss": 0.0541, + "step": 3669 + }, + { + "epoch": 2.6398129832763892, + "grad_norm": 4.180844752048521, + "learning_rate": 2.522098610957413e-06, + "loss": 0.1575, + "step": 3670 + }, + { + "epoch": 2.6405322783672003, + "grad_norm": 0.7525115756979937, + "learning_rate": 2.521849555463011e-06, + "loss": 0.007, + "step": 3671 + }, + { + "epoch": 2.6412515734580113, + "grad_norm": 5.527869819047108, + "learning_rate": 2.521600447392656e-06, + "loss": 0.1649, + "step": 3672 + }, + { + "epoch": 2.641970868548822, + "grad_norm": 5.021716658896101, + "learning_rate": 2.5213512867591637e-06, + "loss": 0.1074, + "step": 3673 + }, + { + "epoch": 2.6426901636396334, + "grad_norm": 7.113543250843125, + "learning_rate": 2.521102073575355e-06, + "loss": 0.1661, + "step": 3674 + }, + { + "epoch": 2.643409458730444, + "grad_norm": 5.486427128211821, + "learning_rate": 2.5208528078540526e-06, + "loss": 0.144, + "step": 3675 + }, + { + "epoch": 2.644128753821255, + "grad_norm": 3.7938829008850474, + "learning_rate": 2.5206034896080806e-06, + "loss": 0.1485, + "step": 3676 + }, + { + "epoch": 2.644848048912066, + "grad_norm": 3.3037568657850587, + "learning_rate": 2.5203541188502676e-06, + "loss": 0.186, + "step": 3677 + }, + { + "epoch": 2.645567344002877, + "grad_norm": 1.5458791820632884, + "learning_rate": 2.5201046955934444e-06, + "loss": 0.047, + "step": 3678 + }, + { + "epoch": 2.646286639093688, + "grad_norm": 4.0669746862308465, + "learning_rate": 2.519855219850444e-06, + "loss": 0.2249, + "step": 3679 + }, + { + "epoch": 2.647005934184499, + "grad_norm": 3.4681655703951138, + "learning_rate": 2.5196056916341016e-06, + "loss": 0.0835, + "step": 3680 + }, + { + "epoch": 2.64772522927531, + "grad_norm": 3.36461425629639, + "learning_rate": 2.519356110957257e-06, + "loss": 0.1012, + "step": 3681 + }, + { + "epoch": 2.648444524366121, + "grad_norm": 3.9286969089688384, + "learning_rate": 2.5191064778327505e-06, + "loss": 0.1335, + "step": 3682 + }, + { + "epoch": 2.6491638194569322, + "grad_norm": 3.80027168525941, + "learning_rate": 2.518856792273427e-06, + "loss": 0.1119, + "step": 3683 + }, + { + "epoch": 2.6498831145477433, + "grad_norm": 6.36085198822989, + "learning_rate": 2.5186070542921335e-06, + "loss": 0.0961, + "step": 3684 + }, + { + "epoch": 2.6506024096385543, + "grad_norm": 3.109014723056777, + "learning_rate": 2.5183572639017182e-06, + "loss": 0.0108, + "step": 3685 + }, + { + "epoch": 2.6513217047293653, + "grad_norm": 4.080856582961475, + "learning_rate": 2.5181074211150345e-06, + "loss": 0.0788, + "step": 3686 + }, + { + "epoch": 2.6520409998201764, + "grad_norm": 4.368595693091915, + "learning_rate": 2.5178575259449364e-06, + "loss": 0.1507, + "step": 3687 + }, + { + "epoch": 2.652760294910987, + "grad_norm": 2.069468663756614, + "learning_rate": 2.517607578404282e-06, + "loss": 0.0476, + "step": 3688 + }, + { + "epoch": 2.6534795900017984, + "grad_norm": 3.670353653272536, + "learning_rate": 2.5173575785059314e-06, + "loss": 0.135, + "step": 3689 + }, + { + "epoch": 2.654198885092609, + "grad_norm": 1.913866513974025, + "learning_rate": 2.5171075262627475e-06, + "loss": 0.0452, + "step": 3690 + }, + { + "epoch": 2.6549181801834205, + "grad_norm": 5.38354193408671, + "learning_rate": 2.5168574216875955e-06, + "loss": 0.1878, + "step": 3691 + }, + { + "epoch": 2.655637475274231, + "grad_norm": 3.8498479946055157, + "learning_rate": 2.5166072647933445e-06, + "loss": 0.1365, + "step": 3692 + }, + { + "epoch": 2.656356770365042, + "grad_norm": 7.240928825153363, + "learning_rate": 2.516357055592865e-06, + "loss": 0.1269, + "step": 3693 + }, + { + "epoch": 2.657076065455853, + "grad_norm": 2.8701860407827517, + "learning_rate": 2.5161067940990307e-06, + "loss": 0.0602, + "step": 3694 + }, + { + "epoch": 2.657795360546664, + "grad_norm": 0.8349571543988786, + "learning_rate": 2.515856480324718e-06, + "loss": 0.0029, + "step": 3695 + }, + { + "epoch": 2.6585146556374752, + "grad_norm": 3.352004991881754, + "learning_rate": 2.515606114282807e-06, + "loss": 0.0565, + "step": 3696 + }, + { + "epoch": 2.6592339507282863, + "grad_norm": 3.8020763122160806, + "learning_rate": 2.515355695986178e-06, + "loss": 0.1456, + "step": 3697 + }, + { + "epoch": 2.6599532458190973, + "grad_norm": 3.1413841286057083, + "learning_rate": 2.515105225447716e-06, + "loss": 0.0573, + "step": 3698 + }, + { + "epoch": 2.6606725409099083, + "grad_norm": 3.3504557985257133, + "learning_rate": 2.514854702680308e-06, + "loss": 0.0397, + "step": 3699 + }, + { + "epoch": 2.6613918360007194, + "grad_norm": 3.8842973516463855, + "learning_rate": 2.514604127696844e-06, + "loss": 0.0605, + "step": 3700 + }, + { + "epoch": 2.6621111310915304, + "grad_norm": 3.4559127078038396, + "learning_rate": 2.514353500510217e-06, + "loss": 0.1296, + "step": 3701 + }, + { + "epoch": 2.6628304261823414, + "grad_norm": 3.4128165104304133, + "learning_rate": 2.5141028211333215e-06, + "loss": 0.0677, + "step": 3702 + }, + { + "epoch": 2.663549721273152, + "grad_norm": 1.4948083931353333, + "learning_rate": 2.513852089579056e-06, + "loss": 0.0332, + "step": 3703 + }, + { + "epoch": 2.6642690163639635, + "grad_norm": 0.3681565698425834, + "learning_rate": 2.5136013058603203e-06, + "loss": 0.0011, + "step": 3704 + }, + { + "epoch": 2.664988311454774, + "grad_norm": 3.0782569704485168, + "learning_rate": 2.513350469990018e-06, + "loss": 0.0464, + "step": 3705 + }, + { + "epoch": 2.6657076065455856, + "grad_norm": 4.508129414508531, + "learning_rate": 2.5130995819810554e-06, + "loss": 0.0997, + "step": 3706 + }, + { + "epoch": 2.666426901636396, + "grad_norm": 4.439008375032999, + "learning_rate": 2.5128486418463405e-06, + "loss": 0.1724, + "step": 3707 + }, + { + "epoch": 2.667146196727207, + "grad_norm": 6.448391018258753, + "learning_rate": 2.5125976495987855e-06, + "loss": 0.1329, + "step": 3708 + }, + { + "epoch": 2.6678654918180182, + "grad_norm": 1.5862208292542819, + "learning_rate": 2.5123466052513037e-06, + "loss": 0.0509, + "step": 3709 + }, + { + "epoch": 2.6685847869088293, + "grad_norm": 6.006311399033885, + "learning_rate": 2.512095508816812e-06, + "loss": 0.2098, + "step": 3710 + }, + { + "epoch": 2.6693040819996403, + "grad_norm": 2.0973880586514104, + "learning_rate": 2.511844360308229e-06, + "loss": 0.0429, + "step": 3711 + }, + { + "epoch": 2.6700233770904513, + "grad_norm": 2.200696129453983, + "learning_rate": 2.5115931597384786e-06, + "loss": 0.0564, + "step": 3712 + }, + { + "epoch": 2.6707426721812624, + "grad_norm": 5.1460394661064885, + "learning_rate": 2.5113419071204837e-06, + "loss": 0.0496, + "step": 3713 + }, + { + "epoch": 2.6714619672720734, + "grad_norm": 3.0807263481660847, + "learning_rate": 2.511090602467172e-06, + "loss": 0.0882, + "step": 3714 + }, + { + "epoch": 2.6721812623628844, + "grad_norm": 2.767989077439596, + "learning_rate": 2.5108392457914745e-06, + "loss": 0.035, + "step": 3715 + }, + { + "epoch": 2.6729005574536955, + "grad_norm": 3.22985158347298, + "learning_rate": 2.510587837106323e-06, + "loss": 0.0918, + "step": 3716 + }, + { + "epoch": 2.6736198525445065, + "grad_norm": 5.685990673290136, + "learning_rate": 2.5103363764246534e-06, + "loss": 0.02, + "step": 3717 + }, + { + "epoch": 2.6743391476353175, + "grad_norm": 1.3526489239489268, + "learning_rate": 2.5100848637594037e-06, + "loss": 0.0276, + "step": 3718 + }, + { + "epoch": 2.6750584427261286, + "grad_norm": 5.080821916705485, + "learning_rate": 2.5098332991235146e-06, + "loss": 0.0773, + "step": 3719 + }, + { + "epoch": 2.675777737816939, + "grad_norm": 4.825549932184532, + "learning_rate": 2.5095816825299296e-06, + "loss": 0.0155, + "step": 3720 + }, + { + "epoch": 2.6764970329077507, + "grad_norm": 3.339456065210547, + "learning_rate": 2.509330013991594e-06, + "loss": 0.0647, + "step": 3721 + }, + { + "epoch": 2.6772163279985612, + "grad_norm": 1.9561245021690106, + "learning_rate": 2.5090782935214582e-06, + "loss": 0.0125, + "step": 3722 + }, + { + "epoch": 2.6779356230893723, + "grad_norm": 6.473213261711899, + "learning_rate": 2.508826521132473e-06, + "loss": 0.0417, + "step": 3723 + }, + { + "epoch": 2.6786549181801833, + "grad_norm": 2.838375328362837, + "learning_rate": 2.5085746968375916e-06, + "loss": 0.0686, + "step": 3724 + }, + { + "epoch": 2.6793742132709943, + "grad_norm": 2.328340876556262, + "learning_rate": 2.508322820649772e-06, + "loss": 0.0062, + "step": 3725 + }, + { + "epoch": 2.6800935083618054, + "grad_norm": 4.180222374359697, + "learning_rate": 2.508070892581973e-06, + "loss": 0.0846, + "step": 3726 + }, + { + "epoch": 2.6808128034526164, + "grad_norm": 2.0105111378374856, + "learning_rate": 2.5078189126471572e-06, + "loss": 0.0299, + "step": 3727 + }, + { + "epoch": 2.6815320985434274, + "grad_norm": 1.8372937207366125, + "learning_rate": 2.5075668808582892e-06, + "loss": 0.0405, + "step": 3728 + }, + { + "epoch": 2.6822513936342385, + "grad_norm": 3.9709861439457126, + "learning_rate": 2.507314797228336e-06, + "loss": 0.172, + "step": 3729 + }, + { + "epoch": 2.6829706887250495, + "grad_norm": 2.9393880250293907, + "learning_rate": 2.5070626617702685e-06, + "loss": 0.1039, + "step": 3730 + }, + { + "epoch": 2.6836899838158605, + "grad_norm": 3.582140188522379, + "learning_rate": 2.5068104744970596e-06, + "loss": 0.0234, + "step": 3731 + }, + { + "epoch": 2.6844092789066716, + "grad_norm": 4.536372856698448, + "learning_rate": 2.5065582354216836e-06, + "loss": 0.2384, + "step": 3732 + }, + { + "epoch": 2.6851285739974826, + "grad_norm": 4.683097014370957, + "learning_rate": 2.50630594455712e-06, + "loss": 0.1698, + "step": 3733 + }, + { + "epoch": 2.6858478690882936, + "grad_norm": 7.623351361379616, + "learning_rate": 2.5060536019163488e-06, + "loss": 0.0651, + "step": 3734 + }, + { + "epoch": 2.6865671641791042, + "grad_norm": 3.436159723634687, + "learning_rate": 2.505801207512354e-06, + "loss": 0.0673, + "step": 3735 + }, + { + "epoch": 2.6872864592699157, + "grad_norm": 3.3445653716355954, + "learning_rate": 2.505548761358121e-06, + "loss": 0.0258, + "step": 3736 + }, + { + "epoch": 2.6880057543607263, + "grad_norm": 5.9275934808945, + "learning_rate": 2.5052962634666394e-06, + "loss": 0.1, + "step": 3737 + }, + { + "epoch": 2.6887250494515373, + "grad_norm": 3.612022371542458, + "learning_rate": 2.5050437138509e-06, + "loss": 0.0311, + "step": 3738 + }, + { + "epoch": 2.6894443445423484, + "grad_norm": 5.101380180247387, + "learning_rate": 2.5047911125238976e-06, + "loss": 0.1947, + "step": 3739 + }, + { + "epoch": 2.6901636396331594, + "grad_norm": 3.324910546004371, + "learning_rate": 2.5045384594986286e-06, + "loss": 0.0614, + "step": 3740 + }, + { + "epoch": 2.6908829347239704, + "grad_norm": 3.563723911119825, + "learning_rate": 2.504285754788092e-06, + "loss": 0.0999, + "step": 3741 + }, + { + "epoch": 2.6916022298147815, + "grad_norm": 1.791112471627173, + "learning_rate": 2.50403299840529e-06, + "loss": 0.0234, + "step": 3742 + }, + { + "epoch": 2.6923215249055925, + "grad_norm": 3.7782141473859157, + "learning_rate": 2.503780190363229e-06, + "loss": 0.1059, + "step": 3743 + }, + { + "epoch": 2.6930408199964035, + "grad_norm": 1.8366305740591597, + "learning_rate": 2.5035273306749142e-06, + "loss": 0.0245, + "step": 3744 + }, + { + "epoch": 2.6937601150872146, + "grad_norm": 2.5378193881548032, + "learning_rate": 2.503274419353357e-06, + "loss": 0.0858, + "step": 3745 + }, + { + "epoch": 2.6944794101780256, + "grad_norm": 4.097122388215333, + "learning_rate": 2.50302145641157e-06, + "loss": 0.0849, + "step": 3746 + }, + { + "epoch": 2.6951987052688366, + "grad_norm": 2.0825888205759426, + "learning_rate": 2.5027684418625677e-06, + "loss": 0.0236, + "step": 3747 + }, + { + "epoch": 2.6959180003596477, + "grad_norm": 1.715268006172488, + "learning_rate": 2.5025153757193687e-06, + "loss": 0.0174, + "step": 3748 + }, + { + "epoch": 2.6966372954504587, + "grad_norm": 4.008159243437323, + "learning_rate": 2.502262257994994e-06, + "loss": 0.1017, + "step": 3749 + }, + { + "epoch": 2.6973565905412693, + "grad_norm": 3.910186110703431, + "learning_rate": 2.5020090887024663e-06, + "loss": 0.0258, + "step": 3750 + }, + { + "epoch": 2.698075885632081, + "grad_norm": 2.6885361033008324, + "learning_rate": 2.5017558678548125e-06, + "loss": 0.081, + "step": 3751 + }, + { + "epoch": 2.6987951807228914, + "grad_norm": 5.532676240066749, + "learning_rate": 2.501502595465061e-06, + "loss": 0.1759, + "step": 3752 + }, + { + "epoch": 2.6995144758137024, + "grad_norm": 3.003324394368566, + "learning_rate": 2.5012492715462422e-06, + "loss": 0.0656, + "step": 3753 + }, + { + "epoch": 2.7002337709045134, + "grad_norm": 1.3094946917644321, + "learning_rate": 2.5009958961113914e-06, + "loss": 0.029, + "step": 3754 + }, + { + "epoch": 2.7009530659953245, + "grad_norm": 6.562976193825908, + "learning_rate": 2.500742469173544e-06, + "loss": 0.157, + "step": 3755 + }, + { + "epoch": 2.7016723610861355, + "grad_norm": 4.3720726567237005, + "learning_rate": 2.50048899074574e-06, + "loss": 0.1266, + "step": 3756 + }, + { + "epoch": 2.7023916561769465, + "grad_norm": 2.3029017667465963, + "learning_rate": 2.5002354608410212e-06, + "loss": 0.0175, + "step": 3757 + }, + { + "epoch": 2.7031109512677576, + "grad_norm": 2.6149334590859934, + "learning_rate": 2.4999818794724324e-06, + "loss": 0.0622, + "step": 3758 + }, + { + "epoch": 2.7038302463585686, + "grad_norm": 1.2967474991398897, + "learning_rate": 2.4997282466530206e-06, + "loss": 0.0275, + "step": 3759 + }, + { + "epoch": 2.7045495414493796, + "grad_norm": 4.003735820358936, + "learning_rate": 2.499474562395835e-06, + "loss": 0.0986, + "step": 3760 + }, + { + "epoch": 2.7052688365401907, + "grad_norm": 3.1107490529969914, + "learning_rate": 2.499220826713928e-06, + "loss": 0.1539, + "step": 3761 + }, + { + "epoch": 2.7059881316310017, + "grad_norm": 1.5447893298975588, + "learning_rate": 2.498967039620357e-06, + "loss": 0.0393, + "step": 3762 + }, + { + "epoch": 2.7067074267218127, + "grad_norm": 1.430421493773461, + "learning_rate": 2.498713201128177e-06, + "loss": 0.0312, + "step": 3763 + }, + { + "epoch": 2.707426721812624, + "grad_norm": 0.7147655038678005, + "learning_rate": 2.4984593112504496e-06, + "loss": 0.0028, + "step": 3764 + }, + { + "epoch": 2.7081460169034344, + "grad_norm": 2.6163023900954605, + "learning_rate": 2.4982053700002387e-06, + "loss": 0.0455, + "step": 3765 + }, + { + "epoch": 2.708865311994246, + "grad_norm": 2.8367525629082797, + "learning_rate": 2.4979513773906086e-06, + "loss": 0.039, + "step": 3766 + }, + { + "epoch": 2.7095846070850564, + "grad_norm": 4.418308938583768, + "learning_rate": 2.4976973334346285e-06, + "loss": 0.1885, + "step": 3767 + }, + { + "epoch": 2.710303902175868, + "grad_norm": 2.2718286888193684, + "learning_rate": 2.4974432381453694e-06, + "loss": 0.0262, + "step": 3768 + }, + { + "epoch": 2.7110231972666785, + "grad_norm": 2.1181053858894985, + "learning_rate": 2.497189091535904e-06, + "loss": 0.0287, + "step": 3769 + }, + { + "epoch": 2.7117424923574895, + "grad_norm": 5.550152053541698, + "learning_rate": 2.4969348936193103e-06, + "loss": 0.0358, + "step": 3770 + }, + { + "epoch": 2.7124617874483006, + "grad_norm": 7.826237432316334, + "learning_rate": 2.4966806444086653e-06, + "loss": 0.2648, + "step": 3771 + }, + { + "epoch": 2.7131810825391116, + "grad_norm": 2.2515669585947684, + "learning_rate": 2.496426343917052e-06, + "loss": 0.0363, + "step": 3772 + }, + { + "epoch": 2.7139003776299226, + "grad_norm": 3.033278647540971, + "learning_rate": 2.4961719921575544e-06, + "loss": 0.0723, + "step": 3773 + }, + { + "epoch": 2.7146196727207337, + "grad_norm": 5.0513808588723474, + "learning_rate": 2.495917589143259e-06, + "loss": 0.1542, + "step": 3774 + }, + { + "epoch": 2.7153389678115447, + "grad_norm": 1.236200234155994, + "learning_rate": 2.4956631348872547e-06, + "loss": 0.0011, + "step": 3775 + }, + { + "epoch": 2.7160582629023557, + "grad_norm": 1.794058497036737, + "learning_rate": 2.4954086294026347e-06, + "loss": 0.0425, + "step": 3776 + }, + { + "epoch": 2.716777557993167, + "grad_norm": 0.7885792545823976, + "learning_rate": 2.4951540727024937e-06, + "loss": 0.0113, + "step": 3777 + }, + { + "epoch": 2.717496853083978, + "grad_norm": 3.9786612271396318, + "learning_rate": 2.4948994647999284e-06, + "loss": 0.1299, + "step": 3778 + }, + { + "epoch": 2.718216148174789, + "grad_norm": 3.800782343998083, + "learning_rate": 2.4946448057080387e-06, + "loss": 0.0195, + "step": 3779 + }, + { + "epoch": 2.7189354432655994, + "grad_norm": 5.517397828314818, + "learning_rate": 2.4943900954399285e-06, + "loss": 0.1374, + "step": 3780 + }, + { + "epoch": 2.719654738356411, + "grad_norm": 4.749014827544546, + "learning_rate": 2.494135334008702e-06, + "loss": 0.0712, + "step": 3781 + }, + { + "epoch": 2.7203740334472215, + "grad_norm": 2.079495561781943, + "learning_rate": 2.4938805214274674e-06, + "loss": 0.0589, + "step": 3782 + }, + { + "epoch": 2.721093328538033, + "grad_norm": 3.7865581853634414, + "learning_rate": 2.493625657709335e-06, + "loss": 0.0141, + "step": 3783 + }, + { + "epoch": 2.7218126236288436, + "grad_norm": 2.7866566240133426, + "learning_rate": 2.4933707428674186e-06, + "loss": 0.0629, + "step": 3784 + }, + { + "epoch": 2.7225319187196546, + "grad_norm": 6.317690132660262, + "learning_rate": 2.4931157769148337e-06, + "loss": 0.1965, + "step": 3785 + }, + { + "epoch": 2.7232512138104656, + "grad_norm": 4.884710160667021, + "learning_rate": 2.4928607598646986e-06, + "loss": 0.1197, + "step": 3786 + }, + { + "epoch": 2.7239705089012767, + "grad_norm": 2.4681905263098405, + "learning_rate": 2.4926056917301344e-06, + "loss": 0.0571, + "step": 3787 + }, + { + "epoch": 2.7246898039920877, + "grad_norm": 4.1517503717633, + "learning_rate": 2.492350572524265e-06, + "loss": 0.22, + "step": 3788 + }, + { + "epoch": 2.7254090990828987, + "grad_norm": 2.5431151256523674, + "learning_rate": 2.4920954022602164e-06, + "loss": 0.0204, + "step": 3789 + }, + { + "epoch": 2.72612839417371, + "grad_norm": 4.02233241232701, + "learning_rate": 2.491840180951118e-06, + "loss": 0.1906, + "step": 3790 + }, + { + "epoch": 2.726847689264521, + "grad_norm": 1.8758054620885445, + "learning_rate": 2.491584908610101e-06, + "loss": 0.0162, + "step": 3791 + }, + { + "epoch": 2.727566984355332, + "grad_norm": 4.11083243647948, + "learning_rate": 2.4913295852503e-06, + "loss": 0.0949, + "step": 3792 + }, + { + "epoch": 2.728286279446143, + "grad_norm": 4.023825793882214, + "learning_rate": 2.4910742108848517e-06, + "loss": 0.0804, + "step": 3793 + }, + { + "epoch": 2.729005574536954, + "grad_norm": 2.093990926581722, + "learning_rate": 2.4908187855268954e-06, + "loss": 0.0066, + "step": 3794 + }, + { + "epoch": 2.729724869627765, + "grad_norm": 1.7491648990941016, + "learning_rate": 2.490563309189573e-06, + "loss": 0.0126, + "step": 3795 + }, + { + "epoch": 2.730444164718576, + "grad_norm": 2.399989557005225, + "learning_rate": 2.4903077818860298e-06, + "loss": 0.045, + "step": 3796 + }, + { + "epoch": 2.7311634598093866, + "grad_norm": 7.659149651078495, + "learning_rate": 2.490052203629412e-06, + "loss": 0.2433, + "step": 3797 + }, + { + "epoch": 2.731882754900198, + "grad_norm": 5.279722924681732, + "learning_rate": 2.489796574432872e-06, + "loss": 0.0777, + "step": 3798 + }, + { + "epoch": 2.7326020499910086, + "grad_norm": 2.779167424058132, + "learning_rate": 2.489540894309559e-06, + "loss": 0.0724, + "step": 3799 + }, + { + "epoch": 2.7333213450818197, + "grad_norm": 3.9183631881981826, + "learning_rate": 2.489285163272631e-06, + "loss": 0.1064, + "step": 3800 + }, + { + "epoch": 2.7340406401726307, + "grad_norm": 1.7996846817942445, + "learning_rate": 2.489029381335244e-06, + "loss": 0.0526, + "step": 3801 + }, + { + "epoch": 2.7347599352634417, + "grad_norm": 2.0556973503417004, + "learning_rate": 2.4887735485105597e-06, + "loss": 0.0331, + "step": 3802 + }, + { + "epoch": 2.7354792303542528, + "grad_norm": 4.095523898768103, + "learning_rate": 2.48851766481174e-06, + "loss": 0.1218, + "step": 3803 + }, + { + "epoch": 2.736198525445064, + "grad_norm": 1.4382698777803322, + "learning_rate": 2.488261730251952e-06, + "loss": 0.037, + "step": 3804 + }, + { + "epoch": 2.736917820535875, + "grad_norm": 2.913472064449454, + "learning_rate": 2.4880057448443625e-06, + "loss": 0.0171, + "step": 3805 + }, + { + "epoch": 2.737637115626686, + "grad_norm": 5.497036022929975, + "learning_rate": 2.487749708602143e-06, + "loss": 0.2356, + "step": 3806 + }, + { + "epoch": 2.738356410717497, + "grad_norm": 2.5011258367401026, + "learning_rate": 2.4874936215384673e-06, + "loss": 0.0705, + "step": 3807 + }, + { + "epoch": 2.739075705808308, + "grad_norm": 4.92438581028327, + "learning_rate": 2.4872374836665114e-06, + "loss": 0.0954, + "step": 3808 + }, + { + "epoch": 2.739795000899119, + "grad_norm": 2.786830921258804, + "learning_rate": 2.4869812949994533e-06, + "loss": 0.0749, + "step": 3809 + }, + { + "epoch": 2.74051429598993, + "grad_norm": 1.8534674508542008, + "learning_rate": 2.486725055550476e-06, + "loss": 0.0069, + "step": 3810 + }, + { + "epoch": 2.741233591080741, + "grad_norm": 6.822405150654127, + "learning_rate": 2.4864687653327615e-06, + "loss": 0.2362, + "step": 3811 + }, + { + "epoch": 2.7419528861715516, + "grad_norm": 3.241730363571939, + "learning_rate": 2.486212424359498e-06, + "loss": 0.1017, + "step": 3812 + }, + { + "epoch": 2.742672181262363, + "grad_norm": 0.9117142688657816, + "learning_rate": 2.4859560326438735e-06, + "loss": 0.0023, + "step": 3813 + }, + { + "epoch": 2.7433914763531737, + "grad_norm": 2.1969584255683046, + "learning_rate": 2.48569959019908e-06, + "loss": 0.0518, + "step": 3814 + }, + { + "epoch": 2.7441107714439847, + "grad_norm": 7.07143618197524, + "learning_rate": 2.4854430970383133e-06, + "loss": 0.1294, + "step": 3815 + }, + { + "epoch": 2.7448300665347958, + "grad_norm": 4.424225818464861, + "learning_rate": 2.485186553174769e-06, + "loss": 0.0107, + "step": 3816 + }, + { + "epoch": 2.745549361625607, + "grad_norm": 3.60659689273812, + "learning_rate": 2.484929958621647e-06, + "loss": 0.0483, + "step": 3817 + }, + { + "epoch": 2.746268656716418, + "grad_norm": 5.526072670512086, + "learning_rate": 2.4846733133921493e-06, + "loss": 0.0575, + "step": 3818 + }, + { + "epoch": 2.746987951807229, + "grad_norm": 2.5986371436460836, + "learning_rate": 2.4844166174994813e-06, + "loss": 0.0567, + "step": 3819 + }, + { + "epoch": 2.74770724689804, + "grad_norm": 6.72525545727892, + "learning_rate": 2.4841598709568505e-06, + "loss": 0.2309, + "step": 3820 + }, + { + "epoch": 2.748426541988851, + "grad_norm": 5.790728760969494, + "learning_rate": 2.4839030737774662e-06, + "loss": 0.1239, + "step": 3821 + }, + { + "epoch": 2.749145837079662, + "grad_norm": 2.932769247696104, + "learning_rate": 2.483646225974542e-06, + "loss": 0.0964, + "step": 3822 + }, + { + "epoch": 2.749865132170473, + "grad_norm": 2.5459656140521436, + "learning_rate": 2.483389327561293e-06, + "loss": 0.0117, + "step": 3823 + }, + { + "epoch": 2.750584427261284, + "grad_norm": 2.0235417859285096, + "learning_rate": 2.483132378550936e-06, + "loss": 0.052, + "step": 3824 + }, + { + "epoch": 2.751303722352095, + "grad_norm": 4.49608334341545, + "learning_rate": 2.482875378956693e-06, + "loss": 0.1671, + "step": 3825 + }, + { + "epoch": 2.752023017442906, + "grad_norm": 5.001709918313664, + "learning_rate": 2.4826183287917862e-06, + "loss": 0.1057, + "step": 3826 + }, + { + "epoch": 2.7527423125337167, + "grad_norm": 0.625589154422185, + "learning_rate": 2.482361228069441e-06, + "loss": 0.0084, + "step": 3827 + }, + { + "epoch": 2.753461607624528, + "grad_norm": 1.6003918171345752, + "learning_rate": 2.4821040768028867e-06, + "loss": 0.0061, + "step": 3828 + }, + { + "epoch": 2.7541809027153388, + "grad_norm": 5.221914451576344, + "learning_rate": 2.4818468750053532e-06, + "loss": 0.2122, + "step": 3829 + }, + { + "epoch": 2.75490019780615, + "grad_norm": 3.8509367562719, + "learning_rate": 2.4815896226900747e-06, + "loss": 0.0365, + "step": 3830 + }, + { + "epoch": 2.755619492896961, + "grad_norm": 1.6401478360540251, + "learning_rate": 2.481332319870287e-06, + "loss": 0.0175, + "step": 3831 + }, + { + "epoch": 2.756338787987772, + "grad_norm": 2.3623839972385743, + "learning_rate": 2.481074966559229e-06, + "loss": 0.0492, + "step": 3832 + }, + { + "epoch": 2.757058083078583, + "grad_norm": 3.2357350982296254, + "learning_rate": 2.480817562770141e-06, + "loss": 0.1159, + "step": 3833 + }, + { + "epoch": 2.757777378169394, + "grad_norm": 2.8424872505661654, + "learning_rate": 2.480560108516268e-06, + "loss": 0.1208, + "step": 3834 + }, + { + "epoch": 2.758496673260205, + "grad_norm": 3.342064015514164, + "learning_rate": 2.480302603810856e-06, + "loss": 0.1189, + "step": 3835 + }, + { + "epoch": 2.759215968351016, + "grad_norm": 5.102169524804131, + "learning_rate": 2.480045048667155e-06, + "loss": 0.128, + "step": 3836 + }, + { + "epoch": 2.759935263441827, + "grad_norm": 3.4758142104540886, + "learning_rate": 2.4797874430984145e-06, + "loss": 0.0851, + "step": 3837 + }, + { + "epoch": 2.760654558532638, + "grad_norm": 3.803944360324067, + "learning_rate": 2.479529787117891e-06, + "loss": 0.145, + "step": 3838 + }, + { + "epoch": 2.761373853623449, + "grad_norm": 0.053240153791519355, + "learning_rate": 2.47927208073884e-06, + "loss": 0.0002, + "step": 3839 + }, + { + "epoch": 2.76209314871426, + "grad_norm": 4.682782318988433, + "learning_rate": 2.479014323974522e-06, + "loss": 0.1588, + "step": 3840 + }, + { + "epoch": 2.762812443805071, + "grad_norm": 5.049956238321539, + "learning_rate": 2.4787565168381977e-06, + "loss": 0.1748, + "step": 3841 + }, + { + "epoch": 2.7635317388958818, + "grad_norm": 1.8721860693308074, + "learning_rate": 2.4784986593431326e-06, + "loss": 0.0042, + "step": 3842 + }, + { + "epoch": 2.7642510339866933, + "grad_norm": 7.490756552189824, + "learning_rate": 2.478240751502594e-06, + "loss": 0.1896, + "step": 3843 + }, + { + "epoch": 2.764970329077504, + "grad_norm": 3.874699148544077, + "learning_rate": 2.4779827933298515e-06, + "loss": 0.0935, + "step": 3844 + }, + { + "epoch": 2.7656896241683153, + "grad_norm": 3.364981346670917, + "learning_rate": 2.477724784838177e-06, + "loss": 0.0555, + "step": 3845 + }, + { + "epoch": 2.766408919259126, + "grad_norm": 2.460734601641832, + "learning_rate": 2.4774667260408467e-06, + "loss": 0.0404, + "step": 3846 + }, + { + "epoch": 2.767128214349937, + "grad_norm": 3.726257485263743, + "learning_rate": 2.4772086169511368e-06, + "loss": 0.0478, + "step": 3847 + }, + { + "epoch": 2.767847509440748, + "grad_norm": 4.982709621343737, + "learning_rate": 2.4769504575823287e-06, + "loss": 0.1154, + "step": 3848 + }, + { + "epoch": 2.768566804531559, + "grad_norm": 4.482995244067132, + "learning_rate": 2.4766922479477044e-06, + "loss": 0.1304, + "step": 3849 + }, + { + "epoch": 2.76928609962237, + "grad_norm": 2.468381460186908, + "learning_rate": 2.476433988060549e-06, + "loss": 0.0441, + "step": 3850 + }, + { + "epoch": 2.770005394713181, + "grad_norm": 2.6766598459936906, + "learning_rate": 2.4761756779341514e-06, + "loss": 0.0742, + "step": 3851 + }, + { + "epoch": 2.770724689803992, + "grad_norm": 7.454960120124403, + "learning_rate": 2.475917317581801e-06, + "loss": 0.1368, + "step": 3852 + }, + { + "epoch": 2.771443984894803, + "grad_norm": 5.028421935334111, + "learning_rate": 2.4756589070167924e-06, + "loss": 0.0116, + "step": 3853 + }, + { + "epoch": 2.772163279985614, + "grad_norm": 1.81474267889335, + "learning_rate": 2.4754004462524194e-06, + "loss": 0.0026, + "step": 3854 + }, + { + "epoch": 2.772882575076425, + "grad_norm": 3.4418421952476472, + "learning_rate": 2.4751419353019816e-06, + "loss": 0.0774, + "step": 3855 + }, + { + "epoch": 2.7736018701672362, + "grad_norm": 3.722658637863967, + "learning_rate": 2.474883374178779e-06, + "loss": 0.1103, + "step": 3856 + }, + { + "epoch": 2.774321165258047, + "grad_norm": 2.5819807755254662, + "learning_rate": 2.4746247628961158e-06, + "loss": 0.092, + "step": 3857 + }, + { + "epoch": 2.7750404603488583, + "grad_norm": 1.5906114731577325, + "learning_rate": 2.4743661014672976e-06, + "loss": 0.014, + "step": 3858 + }, + { + "epoch": 2.775759755439669, + "grad_norm": 1.4831716114139566, + "learning_rate": 2.474107389905633e-06, + "loss": 0.0396, + "step": 3859 + }, + { + "epoch": 2.7764790505304804, + "grad_norm": 3.783161035595401, + "learning_rate": 2.4738486282244334e-06, + "loss": 0.162, + "step": 3860 + }, + { + "epoch": 2.777198345621291, + "grad_norm": 4.850337177229191, + "learning_rate": 2.473589816437012e-06, + "loss": 0.2067, + "step": 3861 + }, + { + "epoch": 2.777917640712102, + "grad_norm": 3.2986567958396296, + "learning_rate": 2.4733309545566857e-06, + "loss": 0.0812, + "step": 3862 + }, + { + "epoch": 2.778636935802913, + "grad_norm": 7.190863343943745, + "learning_rate": 2.473072042596773e-06, + "loss": 0.221, + "step": 3863 + }, + { + "epoch": 2.779356230893724, + "grad_norm": 4.647389894934905, + "learning_rate": 2.4728130805705956e-06, + "loss": 0.1893, + "step": 3864 + }, + { + "epoch": 2.780075525984535, + "grad_norm": 0.6471652289611521, + "learning_rate": 2.4725540684914774e-06, + "loss": 0.0062, + "step": 3865 + }, + { + "epoch": 2.780794821075346, + "grad_norm": 2.2653525905085967, + "learning_rate": 2.4722950063727452e-06, + "loss": 0.0132, + "step": 3866 + }, + { + "epoch": 2.781514116166157, + "grad_norm": 3.94115588378899, + "learning_rate": 2.4720358942277284e-06, + "loss": 0.1455, + "step": 3867 + }, + { + "epoch": 2.782233411256968, + "grad_norm": 1.4484118116893585, + "learning_rate": 2.471776732069758e-06, + "loss": 0.0312, + "step": 3868 + }, + { + "epoch": 2.7829527063477792, + "grad_norm": 5.164363175950282, + "learning_rate": 2.471517519912169e-06, + "loss": 0.1202, + "step": 3869 + }, + { + "epoch": 2.7836720014385903, + "grad_norm": 4.085297370762143, + "learning_rate": 2.4712582577682983e-06, + "loss": 0.1215, + "step": 3870 + }, + { + "epoch": 2.7843912965294013, + "grad_norm": 2.1976435355200423, + "learning_rate": 2.4709989456514852e-06, + "loss": 0.023, + "step": 3871 + }, + { + "epoch": 2.7851105916202124, + "grad_norm": 3.795978867425413, + "learning_rate": 2.4707395835750715e-06, + "loss": 0.0536, + "step": 3872 + }, + { + "epoch": 2.7858298867110234, + "grad_norm": 0.49831318191236784, + "learning_rate": 2.4704801715524023e-06, + "loss": 0.0016, + "step": 3873 + }, + { + "epoch": 2.786549181801834, + "grad_norm": 2.735670696013772, + "learning_rate": 2.4702207095968246e-06, + "loss": 0.0972, + "step": 3874 + }, + { + "epoch": 2.7872684768926455, + "grad_norm": 1.5306207695904406, + "learning_rate": 2.469961197721689e-06, + "loss": 0.0094, + "step": 3875 + }, + { + "epoch": 2.787987771983456, + "grad_norm": 7.485317038192657, + "learning_rate": 2.469701635940346e-06, + "loss": 0.0977, + "step": 3876 + }, + { + "epoch": 2.788707067074267, + "grad_norm": 0.8017605506729603, + "learning_rate": 2.469442024266152e-06, + "loss": 0.0075, + "step": 3877 + }, + { + "epoch": 2.789426362165078, + "grad_norm": 1.5315999796239386, + "learning_rate": 2.469182362712464e-06, + "loss": 0.0486, + "step": 3878 + }, + { + "epoch": 2.790145657255889, + "grad_norm": 3.0094171232032885, + "learning_rate": 2.468922651292642e-06, + "loss": 0.0973, + "step": 3879 + }, + { + "epoch": 2.7908649523467, + "grad_norm": 5.495940362829175, + "learning_rate": 2.468662890020049e-06, + "loss": 0.2341, + "step": 3880 + }, + { + "epoch": 2.791584247437511, + "grad_norm": 0.03222792807784245, + "learning_rate": 2.4684030789080497e-06, + "loss": 0.0002, + "step": 3881 + }, + { + "epoch": 2.7923035425283222, + "grad_norm": 1.6522192177849573, + "learning_rate": 2.468143217970012e-06, + "loss": 0.0216, + "step": 3882 + }, + { + "epoch": 2.7930228376191333, + "grad_norm": 4.278584963512131, + "learning_rate": 2.4678833072193064e-06, + "loss": 0.0904, + "step": 3883 + }, + { + "epoch": 2.7937421327099443, + "grad_norm": 1.8658198901715342, + "learning_rate": 2.467623346669305e-06, + "loss": 0.0646, + "step": 3884 + }, + { + "epoch": 2.7944614278007553, + "grad_norm": 4.165368900570307, + "learning_rate": 2.467363336333385e-06, + "loss": 0.0574, + "step": 3885 + }, + { + "epoch": 2.7951807228915664, + "grad_norm": 6.042793593636978, + "learning_rate": 2.467103276224922e-06, + "loss": 0.1367, + "step": 3886 + }, + { + "epoch": 2.7959000179823774, + "grad_norm": 2.5061249024464836, + "learning_rate": 2.4668431663572983e-06, + "loss": 0.0082, + "step": 3887 + }, + { + "epoch": 2.7966193130731885, + "grad_norm": 3.5663650094448576, + "learning_rate": 2.4665830067438963e-06, + "loss": 0.1184, + "step": 3888 + }, + { + "epoch": 2.797338608163999, + "grad_norm": 5.498811650309177, + "learning_rate": 2.4663227973981017e-06, + "loss": 0.0741, + "step": 3889 + }, + { + "epoch": 2.7980579032548105, + "grad_norm": 3.1247564428907095, + "learning_rate": 2.466062538333303e-06, + "loss": 0.1138, + "step": 3890 + }, + { + "epoch": 2.798777198345621, + "grad_norm": 1.8035930668096747, + "learning_rate": 2.4658022295628908e-06, + "loss": 0.0081, + "step": 3891 + }, + { + "epoch": 2.799496493436432, + "grad_norm": 2.3399103368230056, + "learning_rate": 2.4655418711002585e-06, + "loss": 0.0818, + "step": 3892 + }, + { + "epoch": 2.800215788527243, + "grad_norm": 3.7871763473950506, + "learning_rate": 2.465281462958802e-06, + "loss": 0.046, + "step": 3893 + }, + { + "epoch": 2.800935083618054, + "grad_norm": 2.826320122319046, + "learning_rate": 2.4650210051519196e-06, + "loss": 0.0279, + "step": 3894 + }, + { + "epoch": 2.8016543787088652, + "grad_norm": 2.7598554463993406, + "learning_rate": 2.4647604976930126e-06, + "loss": 0.1152, + "step": 3895 + }, + { + "epoch": 2.8023736737996763, + "grad_norm": 4.017108707566826, + "learning_rate": 2.4644999405954846e-06, + "loss": 0.0171, + "step": 3896 + }, + { + "epoch": 2.8030929688904873, + "grad_norm": 2.5880644865828164, + "learning_rate": 2.464239333872741e-06, + "loss": 0.0483, + "step": 3897 + }, + { + "epoch": 2.8038122639812983, + "grad_norm": 0.020342216381505664, + "learning_rate": 2.4639786775381917e-06, + "loss": 0.0001, + "step": 3898 + }, + { + "epoch": 2.8045315590721094, + "grad_norm": 3.279351052693282, + "learning_rate": 2.4637179716052464e-06, + "loss": 0.0659, + "step": 3899 + }, + { + "epoch": 2.8052508541629204, + "grad_norm": 3.747689614132019, + "learning_rate": 2.4634572160873202e-06, + "loss": 0.1245, + "step": 3900 + }, + { + "epoch": 2.8059701492537314, + "grad_norm": 2.556847198583854, + "learning_rate": 2.4631964109978293e-06, + "loss": 0.0805, + "step": 3901 + }, + { + "epoch": 2.8066894443445425, + "grad_norm": 1.487752215800543, + "learning_rate": 2.4629355563501913e-06, + "loss": 0.0379, + "step": 3902 + }, + { + "epoch": 2.8074087394353535, + "grad_norm": 3.5541552559427996, + "learning_rate": 2.462674652157829e-06, + "loss": 0.1612, + "step": 3903 + }, + { + "epoch": 2.808128034526164, + "grad_norm": 4.497174420593064, + "learning_rate": 2.4624136984341665e-06, + "loss": 0.0715, + "step": 3904 + }, + { + "epoch": 2.8088473296169756, + "grad_norm": 5.354153560445609, + "learning_rate": 2.4621526951926288e-06, + "loss": 0.2761, + "step": 3905 + }, + { + "epoch": 2.809566624707786, + "grad_norm": 0.16774901689798485, + "learning_rate": 2.4618916424466465e-06, + "loss": 0.0007, + "step": 3906 + }, + { + "epoch": 2.810285919798597, + "grad_norm": 1.3001457227868953, + "learning_rate": 2.46163054020965e-06, + "loss": 0.0353, + "step": 3907 + }, + { + "epoch": 2.8110052148894082, + "grad_norm": 3.36958623910022, + "learning_rate": 2.461369388495075e-06, + "loss": 0.1269, + "step": 3908 + }, + { + "epoch": 2.8117245099802193, + "grad_norm": 3.7189422319004803, + "learning_rate": 2.461108187316357e-06, + "loss": 0.0856, + "step": 3909 + }, + { + "epoch": 2.8124438050710303, + "grad_norm": 2.1581560593632, + "learning_rate": 2.4608469366869355e-06, + "loss": 0.0534, + "step": 3910 + }, + { + "epoch": 2.8131631001618413, + "grad_norm": 4.074745164373334, + "learning_rate": 2.4605856366202524e-06, + "loss": 0.0265, + "step": 3911 + }, + { + "epoch": 2.8138823952526524, + "grad_norm": 4.702900835399429, + "learning_rate": 2.4603242871297524e-06, + "loss": 0.1406, + "step": 3912 + }, + { + "epoch": 2.8146016903434634, + "grad_norm": 3.7609312800917847, + "learning_rate": 2.460062888228881e-06, + "loss": 0.1355, + "step": 3913 + }, + { + "epoch": 2.8153209854342744, + "grad_norm": 1.9643504726316963, + "learning_rate": 2.4598014399310895e-06, + "loss": 0.029, + "step": 3914 + }, + { + "epoch": 2.8160402805250855, + "grad_norm": 3.899946623151346, + "learning_rate": 2.4595399422498287e-06, + "loss": 0.2234, + "step": 3915 + }, + { + "epoch": 2.8167595756158965, + "grad_norm": 4.158502327403848, + "learning_rate": 2.459278395198554e-06, + "loss": 0.1442, + "step": 3916 + }, + { + "epoch": 2.8174788707067075, + "grad_norm": 0.3814063460861912, + "learning_rate": 2.4590167987907216e-06, + "loss": 0.0008, + "step": 3917 + }, + { + "epoch": 2.8181981657975186, + "grad_norm": 3.792684106102148, + "learning_rate": 2.458755153039791e-06, + "loss": 0.105, + "step": 3918 + }, + { + "epoch": 2.818917460888329, + "grad_norm": 3.037930935556303, + "learning_rate": 2.458493457959225e-06, + "loss": 0.0505, + "step": 3919 + }, + { + "epoch": 2.8196367559791407, + "grad_norm": 4.051596450862193, + "learning_rate": 2.4582317135624885e-06, + "loss": 0.0599, + "step": 3920 + }, + { + "epoch": 2.8203560510699512, + "grad_norm": 0.6171836207977903, + "learning_rate": 2.4579699198630472e-06, + "loss": 0.0019, + "step": 3921 + }, + { + "epoch": 2.8210753461607627, + "grad_norm": 0.13068336295729466, + "learning_rate": 2.457708076874373e-06, + "loss": 0.0006, + "step": 3922 + }, + { + "epoch": 2.8217946412515733, + "grad_norm": 4.043825817052798, + "learning_rate": 2.4574461846099364e-06, + "loss": 0.1228, + "step": 3923 + }, + { + "epoch": 2.8225139363423843, + "grad_norm": 2.822966298703546, + "learning_rate": 2.457184243083213e-06, + "loss": 0.0971, + "step": 3924 + }, + { + "epoch": 2.8232332314331954, + "grad_norm": 3.629541350056613, + "learning_rate": 2.4569222523076796e-06, + "loss": 0.1121, + "step": 3925 + }, + { + "epoch": 2.8239525265240064, + "grad_norm": 3.074149761417383, + "learning_rate": 2.456660212296817e-06, + "loss": 0.0718, + "step": 3926 + }, + { + "epoch": 2.8246718216148174, + "grad_norm": 0.5343892705709685, + "learning_rate": 2.4563981230641072e-06, + "loss": 0.0031, + "step": 3927 + }, + { + "epoch": 2.8253911167056285, + "grad_norm": 3.4434637464799733, + "learning_rate": 2.456135984623035e-06, + "loss": 0.0941, + "step": 3928 + }, + { + "epoch": 2.8261104117964395, + "grad_norm": 0.07857673604378634, + "learning_rate": 2.455873796987087e-06, + "loss": 0.0003, + "step": 3929 + }, + { + "epoch": 2.8268297068872505, + "grad_norm": 3.7174050141573796, + "learning_rate": 2.4556115601697554e-06, + "loss": 0.11, + "step": 3930 + }, + { + "epoch": 2.8275490019780616, + "grad_norm": 1.2891632747551618, + "learning_rate": 2.455349274184531e-06, + "loss": 0.0018, + "step": 3931 + }, + { + "epoch": 2.8282682970688726, + "grad_norm": 5.876042409452094, + "learning_rate": 2.4550869390449094e-06, + "loss": 0.0384, + "step": 3932 + }, + { + "epoch": 2.8289875921596837, + "grad_norm": 4.036109615461958, + "learning_rate": 2.4548245547643885e-06, + "loss": 0.1664, + "step": 3933 + }, + { + "epoch": 2.8297068872504942, + "grad_norm": 4.02423685266137, + "learning_rate": 2.4545621213564678e-06, + "loss": 0.1349, + "step": 3934 + }, + { + "epoch": 2.8304261823413057, + "grad_norm": 7.121058934795864, + "learning_rate": 2.4542996388346503e-06, + "loss": 0.0881, + "step": 3935 + }, + { + "epoch": 2.8311454774321163, + "grad_norm": 3.4641058028701472, + "learning_rate": 2.454037107212441e-06, + "loss": 0.0375, + "step": 3936 + }, + { + "epoch": 2.831864772522928, + "grad_norm": 2.165451777711609, + "learning_rate": 2.4537745265033483e-06, + "loss": 0.0453, + "step": 3937 + }, + { + "epoch": 2.8325840676137384, + "grad_norm": 2.5579313152664014, + "learning_rate": 2.453511896720881e-06, + "loss": 0.0986, + "step": 3938 + }, + { + "epoch": 2.8333033627045494, + "grad_norm": 4.266215110747072, + "learning_rate": 2.4532492178785534e-06, + "loss": 0.1384, + "step": 3939 + }, + { + "epoch": 2.8340226577953604, + "grad_norm": 5.1746847339273625, + "learning_rate": 2.4529864899898805e-06, + "loss": 0.171, + "step": 3940 + }, + { + "epoch": 2.8347419528861715, + "grad_norm": 5.762053673650539, + "learning_rate": 2.4527237130683795e-06, + "loss": 0.1993, + "step": 3941 + }, + { + "epoch": 2.8354612479769825, + "grad_norm": 4.186070567031017, + "learning_rate": 2.4524608871275703e-06, + "loss": 0.1241, + "step": 3942 + }, + { + "epoch": 2.8361805430677935, + "grad_norm": 3.1773694056042703, + "learning_rate": 2.452198012180977e-06, + "loss": 0.0856, + "step": 3943 + }, + { + "epoch": 2.8368998381586046, + "grad_norm": 1.5064016810534515, + "learning_rate": 2.4519350882421248e-06, + "loss": 0.0609, + "step": 3944 + }, + { + "epoch": 2.8376191332494156, + "grad_norm": 1.750157835421884, + "learning_rate": 2.451672115324541e-06, + "loss": 0.0434, + "step": 3945 + }, + { + "epoch": 2.8383384283402266, + "grad_norm": 3.1831157950776383, + "learning_rate": 2.4514090934417563e-06, + "loss": 0.0419, + "step": 3946 + }, + { + "epoch": 2.8390577234310377, + "grad_norm": 1.0870018209080667, + "learning_rate": 2.451146022607303e-06, + "loss": 0.01, + "step": 3947 + }, + { + "epoch": 2.8397770185218487, + "grad_norm": 7.362506528824175, + "learning_rate": 2.4508829028347176e-06, + "loss": 0.0892, + "step": 3948 + }, + { + "epoch": 2.8404963136126598, + "grad_norm": 2.6623095274022117, + "learning_rate": 2.450619734137538e-06, + "loss": 0.0925, + "step": 3949 + }, + { + "epoch": 2.841215608703471, + "grad_norm": 5.4033203665308545, + "learning_rate": 2.450356516529304e-06, + "loss": 0.0848, + "step": 3950 + }, + { + "epoch": 2.8419349037942814, + "grad_norm": 4.701511893088685, + "learning_rate": 2.450093250023559e-06, + "loss": 0.1343, + "step": 3951 + }, + { + "epoch": 2.842654198885093, + "grad_norm": 6.710058632805847, + "learning_rate": 2.449829934633848e-06, + "loss": 0.214, + "step": 3952 + }, + { + "epoch": 2.8433734939759034, + "grad_norm": 0.5164955730981341, + "learning_rate": 2.44956657037372e-06, + "loss": 0.0013, + "step": 3953 + }, + { + "epoch": 2.8440927890667145, + "grad_norm": 2.2719414017304467, + "learning_rate": 2.449303157256725e-06, + "loss": 0.0545, + "step": 3954 + }, + { + "epoch": 2.8448120841575255, + "grad_norm": 3.3097190162063135, + "learning_rate": 2.449039695296416e-06, + "loss": 0.0617, + "step": 3955 + }, + { + "epoch": 2.8455313792483365, + "grad_norm": 3.661703921111953, + "learning_rate": 2.448776184506349e-06, + "loss": 0.1348, + "step": 3956 + }, + { + "epoch": 2.8462506743391476, + "grad_norm": 3.9434513878931234, + "learning_rate": 2.448512624900081e-06, + "loss": 0.0482, + "step": 3957 + }, + { + "epoch": 2.8469699694299586, + "grad_norm": 2.061810728510171, + "learning_rate": 2.448249016491174e-06, + "loss": 0.0449, + "step": 3958 + }, + { + "epoch": 2.8476892645207696, + "grad_norm": 2.9805563445526033, + "learning_rate": 2.4479853592931908e-06, + "loss": 0.0717, + "step": 3959 + }, + { + "epoch": 2.8484085596115807, + "grad_norm": 3.79019769136605, + "learning_rate": 2.4477216533196956e-06, + "loss": 0.102, + "step": 3960 + }, + { + "epoch": 2.8491278547023917, + "grad_norm": 2.931977244647889, + "learning_rate": 2.4474578985842586e-06, + "loss": 0.0402, + "step": 3961 + }, + { + "epoch": 2.8498471497932027, + "grad_norm": 3.6379791922182214, + "learning_rate": 2.447194095100449e-06, + "loss": 0.1069, + "step": 3962 + }, + { + "epoch": 2.850566444884014, + "grad_norm": 4.267252625106105, + "learning_rate": 2.4469302428818414e-06, + "loss": 0.1547, + "step": 3963 + }, + { + "epoch": 2.851285739974825, + "grad_norm": 0.05056332357411624, + "learning_rate": 2.4466663419420095e-06, + "loss": 0.0002, + "step": 3964 + }, + { + "epoch": 2.852005035065636, + "grad_norm": 3.288978529667222, + "learning_rate": 2.446402392294533e-06, + "loss": 0.1552, + "step": 3965 + }, + { + "epoch": 2.8527243301564464, + "grad_norm": 1.8203500233120882, + "learning_rate": 2.4461383939529917e-06, + "loss": 0.0107, + "step": 3966 + }, + { + "epoch": 2.853443625247258, + "grad_norm": 3.7618450611360825, + "learning_rate": 2.4458743469309693e-06, + "loss": 0.0988, + "step": 3967 + }, + { + "epoch": 2.8541629203380685, + "grad_norm": 4.304674200966189, + "learning_rate": 2.445610251242051e-06, + "loss": 0.1306, + "step": 3968 + }, + { + "epoch": 2.8548822154288795, + "grad_norm": 4.316998971394456, + "learning_rate": 2.445346106899826e-06, + "loss": 0.1338, + "step": 3969 + }, + { + "epoch": 2.8556015105196906, + "grad_norm": 6.114940909869115, + "learning_rate": 2.445081913917884e-06, + "loss": 0.1813, + "step": 3970 + }, + { + "epoch": 2.8563208056105016, + "grad_norm": 5.57808243356584, + "learning_rate": 2.444817672309818e-06, + "loss": 0.1616, + "step": 3971 + }, + { + "epoch": 2.8570401007013126, + "grad_norm": 1.1535881143341444, + "learning_rate": 2.444553382089225e-06, + "loss": 0.0237, + "step": 3972 + }, + { + "epoch": 2.8577593957921237, + "grad_norm": 7.024083904122868, + "learning_rate": 2.444289043269702e-06, + "loss": 0.2559, + "step": 3973 + }, + { + "epoch": 2.8584786908829347, + "grad_norm": 3.469631292689222, + "learning_rate": 2.4440246558648503e-06, + "loss": 0.0238, + "step": 3974 + }, + { + "epoch": 2.8591979859737457, + "grad_norm": 2.472121495050298, + "learning_rate": 2.443760219888273e-06, + "loss": 0.0513, + "step": 3975 + }, + { + "epoch": 2.859917281064557, + "grad_norm": 2.5089600059955406, + "learning_rate": 2.4434957353535747e-06, + "loss": 0.0438, + "step": 3976 + }, + { + "epoch": 2.860636576155368, + "grad_norm": 4.500106520565204, + "learning_rate": 2.4432312022743655e-06, + "loss": 0.0984, + "step": 3977 + }, + { + "epoch": 2.861355871246179, + "grad_norm": 4.2500380728354346, + "learning_rate": 2.442966620664255e-06, + "loss": 0.0946, + "step": 3978 + }, + { + "epoch": 2.86207516633699, + "grad_norm": 2.3293505078068515, + "learning_rate": 2.442701990536856e-06, + "loss": 0.0509, + "step": 3979 + }, + { + "epoch": 2.862794461427801, + "grad_norm": 4.326611013798776, + "learning_rate": 2.4424373119057853e-06, + "loss": 0.0682, + "step": 3980 + }, + { + "epoch": 2.8635137565186115, + "grad_norm": 3.086132242994011, + "learning_rate": 2.44217258478466e-06, + "loss": 0.1492, + "step": 3981 + }, + { + "epoch": 2.864233051609423, + "grad_norm": 3.410951034652264, + "learning_rate": 2.441907809187102e-06, + "loss": 0.0952, + "step": 3982 + }, + { + "epoch": 2.8649523467002336, + "grad_norm": 7.174790656874843, + "learning_rate": 2.441642985126733e-06, + "loss": 0.1243, + "step": 3983 + }, + { + "epoch": 2.8656716417910446, + "grad_norm": 2.978051351629623, + "learning_rate": 2.44137811261718e-06, + "loss": 0.0929, + "step": 3984 + }, + { + "epoch": 2.8663909368818556, + "grad_norm": 4.6915648873351, + "learning_rate": 2.4411131916720706e-06, + "loss": 0.1022, + "step": 3985 + }, + { + "epoch": 2.8671102319726667, + "grad_norm": 3.4057457818581196, + "learning_rate": 2.440848222305035e-06, + "loss": 0.0313, + "step": 3986 + }, + { + "epoch": 2.8678295270634777, + "grad_norm": 2.6204260333164884, + "learning_rate": 2.4405832045297065e-06, + "loss": 0.0688, + "step": 3987 + }, + { + "epoch": 2.8685488221542887, + "grad_norm": 3.997649918011928, + "learning_rate": 2.4403181383597214e-06, + "loss": 0.1027, + "step": 3988 + }, + { + "epoch": 2.8692681172451, + "grad_norm": 0.40027512507999896, + "learning_rate": 2.4400530238087175e-06, + "loss": 0.0012, + "step": 3989 + }, + { + "epoch": 2.869987412335911, + "grad_norm": 2.927866164123594, + "learning_rate": 2.439787860890335e-06, + "loss": 0.0696, + "step": 3990 + }, + { + "epoch": 2.870706707426722, + "grad_norm": 7.310923557898916, + "learning_rate": 2.4395226496182176e-06, + "loss": 0.2647, + "step": 3991 + }, + { + "epoch": 2.871426002517533, + "grad_norm": 5.384865722855087, + "learning_rate": 2.439257390006011e-06, + "loss": 0.1457, + "step": 3992 + }, + { + "epoch": 2.872145297608344, + "grad_norm": 1.9490482562957219, + "learning_rate": 2.438992082067362e-06, + "loss": 0.0288, + "step": 3993 + }, + { + "epoch": 2.872864592699155, + "grad_norm": 0.6213338602229772, + "learning_rate": 2.438726725815922e-06, + "loss": 0.003, + "step": 3994 + }, + { + "epoch": 2.873583887789966, + "grad_norm": 1.1257479061043243, + "learning_rate": 2.4384613212653447e-06, + "loss": 0.0047, + "step": 3995 + }, + { + "epoch": 2.8743031828807766, + "grad_norm": 1.5047134781980294, + "learning_rate": 2.4381958684292852e-06, + "loss": 0.0232, + "step": 3996 + }, + { + "epoch": 2.875022477971588, + "grad_norm": 2.8033836154383045, + "learning_rate": 2.437930367321401e-06, + "loss": 0.0474, + "step": 3997 + }, + { + "epoch": 2.8757417730623986, + "grad_norm": 2.191107533495528, + "learning_rate": 2.4376648179553525e-06, + "loss": 0.0501, + "step": 3998 + }, + { + "epoch": 2.87646106815321, + "grad_norm": 4.106771730558472, + "learning_rate": 2.4373992203448033e-06, + "loss": 0.093, + "step": 3999 + }, + { + "epoch": 2.8771803632440207, + "grad_norm": 4.079746737069633, + "learning_rate": 2.4371335745034187e-06, + "loss": 0.1954, + "step": 4000 + }, + { + "epoch": 2.8778996583348317, + "grad_norm": 5.605464788763568, + "learning_rate": 2.436867880444867e-06, + "loss": 0.082, + "step": 4001 + }, + { + "epoch": 2.878618953425643, + "grad_norm": 1.871823648892638, + "learning_rate": 2.4366021381828177e-06, + "loss": 0.0428, + "step": 4002 + }, + { + "epoch": 2.879338248516454, + "grad_norm": 0.06567784626870671, + "learning_rate": 2.4363363477309446e-06, + "loss": 0.0002, + "step": 4003 + }, + { + "epoch": 2.880057543607265, + "grad_norm": 2.8194330758604726, + "learning_rate": 2.4360705091029225e-06, + "loss": 0.0395, + "step": 4004 + }, + { + "epoch": 2.880776838698076, + "grad_norm": 2.703882846444028, + "learning_rate": 2.43580462231243e-06, + "loss": 0.0175, + "step": 4005 + }, + { + "epoch": 2.881496133788887, + "grad_norm": 4.932939455022974, + "learning_rate": 2.4355386873731465e-06, + "loss": 0.1408, + "step": 4006 + }, + { + "epoch": 2.882215428879698, + "grad_norm": 3.0557873329567626, + "learning_rate": 2.435272704298755e-06, + "loss": 0.1338, + "step": 4007 + }, + { + "epoch": 2.882934723970509, + "grad_norm": 6.306904775007282, + "learning_rate": 2.435006673102941e-06, + "loss": 0.1457, + "step": 4008 + }, + { + "epoch": 2.88365401906132, + "grad_norm": 0.8317877104009143, + "learning_rate": 2.4347405937993928e-06, + "loss": 0.0089, + "step": 4009 + }, + { + "epoch": 2.884373314152131, + "grad_norm": 4.8348194107271985, + "learning_rate": 2.4344744664018e-06, + "loss": 0.2038, + "step": 4010 + }, + { + "epoch": 2.8850926092429416, + "grad_norm": 4.047458302412651, + "learning_rate": 2.434208290923855e-06, + "loss": 0.2075, + "step": 4011 + }, + { + "epoch": 2.885811904333753, + "grad_norm": 9.477334823381563, + "learning_rate": 2.4339420673792547e-06, + "loss": 0.2474, + "step": 4012 + }, + { + "epoch": 2.8865311994245637, + "grad_norm": 5.367446599454089, + "learning_rate": 2.4336757957816946e-06, + "loss": 0.2452, + "step": 4013 + }, + { + "epoch": 2.887250494515375, + "grad_norm": 2.3864766391157275, + "learning_rate": 2.433409476144876e-06, + "loss": 0.0419, + "step": 4014 + }, + { + "epoch": 2.8879697896061858, + "grad_norm": 3.6073024638919846, + "learning_rate": 2.433143108482501e-06, + "loss": 0.047, + "step": 4015 + }, + { + "epoch": 2.888689084696997, + "grad_norm": 3.2261438181962165, + "learning_rate": 2.4328766928082754e-06, + "loss": 0.1055, + "step": 4016 + }, + { + "epoch": 2.889408379787808, + "grad_norm": 3.953972455532349, + "learning_rate": 2.4326102291359064e-06, + "loss": 0.0472, + "step": 4017 + }, + { + "epoch": 2.890127674878619, + "grad_norm": 4.349582275879489, + "learning_rate": 2.4323437174791035e-06, + "loss": 0.1768, + "step": 4018 + }, + { + "epoch": 2.89084696996943, + "grad_norm": 0.23608956034537712, + "learning_rate": 2.43207715785158e-06, + "loss": 0.0009, + "step": 4019 + }, + { + "epoch": 2.891566265060241, + "grad_norm": 4.629453085014154, + "learning_rate": 2.4318105502670505e-06, + "loss": 0.2065, + "step": 4020 + }, + { + "epoch": 2.892285560151052, + "grad_norm": 2.8321606790857743, + "learning_rate": 2.431543894739232e-06, + "loss": 0.0741, + "step": 4021 + }, + { + "epoch": 2.893004855241863, + "grad_norm": 3.857402773719244, + "learning_rate": 2.431277191281846e-06, + "loss": 0.0169, + "step": 4022 + }, + { + "epoch": 2.893724150332674, + "grad_norm": 2.949516060112828, + "learning_rate": 2.431010439908612e-06, + "loss": 0.0266, + "step": 4023 + }, + { + "epoch": 2.894443445423485, + "grad_norm": 1.7563058374528309, + "learning_rate": 2.430743640633258e-06, + "loss": 0.0319, + "step": 4024 + }, + { + "epoch": 2.895162740514296, + "grad_norm": 2.13374950469749, + "learning_rate": 2.4304767934695094e-06, + "loss": 0.0803, + "step": 4025 + }, + { + "epoch": 2.895882035605107, + "grad_norm": 5.829500039613066, + "learning_rate": 2.4302098984310963e-06, + "loss": 0.102, + "step": 4026 + }, + { + "epoch": 2.896601330695918, + "grad_norm": 1.000359921258487, + "learning_rate": 2.4299429555317506e-06, + "loss": 0.0095, + "step": 4027 + }, + { + "epoch": 2.8973206257867288, + "grad_norm": 2.273987849521096, + "learning_rate": 2.429675964785208e-06, + "loss": 0.0982, + "step": 4028 + }, + { + "epoch": 2.8980399208775403, + "grad_norm": 4.377851940181564, + "learning_rate": 2.4294089262052047e-06, + "loss": 0.1295, + "step": 4029 + }, + { + "epoch": 2.898759215968351, + "grad_norm": 1.2304655173003995, + "learning_rate": 2.4291418398054805e-06, + "loss": 0.0252, + "step": 4030 + }, + { + "epoch": 2.899478511059162, + "grad_norm": 7.747474029538773, + "learning_rate": 2.428874705599778e-06, + "loss": 0.2892, + "step": 4031 + }, + { + "epoch": 2.900197806149973, + "grad_norm": 3.5166445876424732, + "learning_rate": 2.4286075236018413e-06, + "loss": 0.0571, + "step": 4032 + }, + { + "epoch": 2.900917101240784, + "grad_norm": 7.080811210758189, + "learning_rate": 2.4283402938254167e-06, + "loss": 0.1347, + "step": 4033 + }, + { + "epoch": 2.901636396331595, + "grad_norm": 3.6410186993151767, + "learning_rate": 2.428073016284255e-06, + "loss": 0.0872, + "step": 4034 + }, + { + "epoch": 2.902355691422406, + "grad_norm": 0.5075037744799848, + "learning_rate": 2.427805690992107e-06, + "loss": 0.0009, + "step": 4035 + }, + { + "epoch": 2.903074986513217, + "grad_norm": 4.944652242433811, + "learning_rate": 2.427538317962728e-06, + "loss": 0.1087, + "step": 4036 + }, + { + "epoch": 2.903794281604028, + "grad_norm": 2.71077036535615, + "learning_rate": 2.4272708972098743e-06, + "loss": 0.0961, + "step": 4037 + }, + { + "epoch": 2.904513576694839, + "grad_norm": 3.177434253173833, + "learning_rate": 2.4270034287473045e-06, + "loss": 0.0752, + "step": 4038 + }, + { + "epoch": 2.90523287178565, + "grad_norm": 5.699450624542296, + "learning_rate": 2.4267359125887813e-06, + "loss": 0.2963, + "step": 4039 + }, + { + "epoch": 2.905952166876461, + "grad_norm": 3.3445526437650117, + "learning_rate": 2.4264683487480685e-06, + "loss": 0.0884, + "step": 4040 + }, + { + "epoch": 2.906671461967272, + "grad_norm": 9.414000578171455, + "learning_rate": 2.426200737238933e-06, + "loss": 0.1655, + "step": 4041 + }, + { + "epoch": 2.9073907570580833, + "grad_norm": 5.457660715804781, + "learning_rate": 2.4259330780751435e-06, + "loss": 0.0939, + "step": 4042 + }, + { + "epoch": 2.908110052148894, + "grad_norm": 5.06964764595076, + "learning_rate": 2.425665371270472e-06, + "loss": 0.0987, + "step": 4043 + }, + { + "epoch": 2.9088293472397053, + "grad_norm": 1.8295566895353454, + "learning_rate": 2.4253976168386917e-06, + "loss": 0.0274, + "step": 4044 + }, + { + "epoch": 2.909548642330516, + "grad_norm": 2.3415930447779005, + "learning_rate": 2.42512981479358e-06, + "loss": 0.0749, + "step": 4045 + }, + { + "epoch": 2.910267937421327, + "grad_norm": 2.6951652858207837, + "learning_rate": 2.424861965148915e-06, + "loss": 0.0604, + "step": 4046 + }, + { + "epoch": 2.910987232512138, + "grad_norm": 2.109786192311443, + "learning_rate": 2.4245940679184787e-06, + "loss": 0.046, + "step": 4047 + }, + { + "epoch": 2.911706527602949, + "grad_norm": 6.558879078971595, + "learning_rate": 2.4243261231160543e-06, + "loss": 0.2115, + "step": 4048 + }, + { + "epoch": 2.91242582269376, + "grad_norm": 3.884864128916162, + "learning_rate": 2.424058130755428e-06, + "loss": 0.1546, + "step": 4049 + }, + { + "epoch": 2.913145117784571, + "grad_norm": 4.692880484306014, + "learning_rate": 2.4237900908503896e-06, + "loss": 0.1718, + "step": 4050 + }, + { + "epoch": 2.913864412875382, + "grad_norm": 3.838108473194209, + "learning_rate": 2.4235220034147286e-06, + "loss": 0.1605, + "step": 4051 + }, + { + "epoch": 2.914583707966193, + "grad_norm": 3.7062481520301174, + "learning_rate": 2.423253868462239e-06, + "loss": 0.1137, + "step": 4052 + }, + { + "epoch": 2.915303003057004, + "grad_norm": 2.3533508807442103, + "learning_rate": 2.422985686006718e-06, + "loss": 0.0316, + "step": 4053 + }, + { + "epoch": 2.916022298147815, + "grad_norm": 4.785612029882363, + "learning_rate": 2.422717456061963e-06, + "loss": 0.1106, + "step": 4054 + }, + { + "epoch": 2.9167415932386263, + "grad_norm": 2.436704157359387, + "learning_rate": 2.422449178641775e-06, + "loss": 0.0337, + "step": 4055 + }, + { + "epoch": 2.9174608883294373, + "grad_norm": 3.5728247008084706, + "learning_rate": 2.4221808537599577e-06, + "loss": 0.1106, + "step": 4056 + }, + { + "epoch": 2.9181801834202483, + "grad_norm": 5.389854919641401, + "learning_rate": 2.421912481430316e-06, + "loss": 0.2431, + "step": 4057 + }, + { + "epoch": 2.918899478511059, + "grad_norm": 3.687793091354124, + "learning_rate": 2.4216440616666593e-06, + "loss": 0.0967, + "step": 4058 + }, + { + "epoch": 2.9196187736018704, + "grad_norm": 5.042936774376561, + "learning_rate": 2.4213755944827978e-06, + "loss": 0.1393, + "step": 4059 + }, + { + "epoch": 2.920338068692681, + "grad_norm": 2.916097533936283, + "learning_rate": 2.421107079892544e-06, + "loss": 0.0247, + "step": 4060 + }, + { + "epoch": 2.921057363783492, + "grad_norm": 2.4429610804838493, + "learning_rate": 2.4208385179097143e-06, + "loss": 0.0585, + "step": 4061 + }, + { + "epoch": 2.921776658874303, + "grad_norm": 3.999142528298892, + "learning_rate": 2.4205699085481263e-06, + "loss": 0.1089, + "step": 4062 + }, + { + "epoch": 2.922495953965114, + "grad_norm": 1.2621150856212513, + "learning_rate": 2.4203012518216003e-06, + "loss": 0.0026, + "step": 4063 + }, + { + "epoch": 2.923215249055925, + "grad_norm": 4.286648128816217, + "learning_rate": 2.42003254774396e-06, + "loss": 0.176, + "step": 4064 + }, + { + "epoch": 2.923934544146736, + "grad_norm": 1.404824649710889, + "learning_rate": 2.4197637963290296e-06, + "loss": 0.0032, + "step": 4065 + }, + { + "epoch": 2.924653839237547, + "grad_norm": 2.7698107332444546, + "learning_rate": 2.419494997590637e-06, + "loss": 0.0968, + "step": 4066 + }, + { + "epoch": 2.925373134328358, + "grad_norm": 2.3330046691836754, + "learning_rate": 2.4192261515426137e-06, + "loss": 0.074, + "step": 4067 + }, + { + "epoch": 2.9260924294191692, + "grad_norm": 5.4132219419039895, + "learning_rate": 2.4189572581987904e-06, + "loss": 0.0865, + "step": 4068 + }, + { + "epoch": 2.9268117245099803, + "grad_norm": 4.36362305725598, + "learning_rate": 2.418688317573003e-06, + "loss": 0.1926, + "step": 4069 + }, + { + "epoch": 2.9275310196007913, + "grad_norm": 3.9338084921022762, + "learning_rate": 2.418419329679089e-06, + "loss": 0.1855, + "step": 4070 + }, + { + "epoch": 2.9282503146916024, + "grad_norm": 5.427174204663504, + "learning_rate": 2.4181502945308884e-06, + "loss": 0.2282, + "step": 4071 + }, + { + "epoch": 2.9289696097824134, + "grad_norm": 6.510650692511353, + "learning_rate": 2.4178812121422433e-06, + "loss": 0.209, + "step": 4072 + }, + { + "epoch": 2.929688904873224, + "grad_norm": 3.3937190120295377, + "learning_rate": 2.4176120825269984e-06, + "loss": 0.132, + "step": 4073 + }, + { + "epoch": 2.9304081999640355, + "grad_norm": 4.374561865419099, + "learning_rate": 2.4173429056990015e-06, + "loss": 0.0982, + "step": 4074 + }, + { + "epoch": 2.931127495054846, + "grad_norm": 6.1255230612342135, + "learning_rate": 2.4170736816721014e-06, + "loss": 0.0456, + "step": 4075 + }, + { + "epoch": 2.9318467901456575, + "grad_norm": 3.9525807601560303, + "learning_rate": 2.4168044104601506e-06, + "loss": 0.1562, + "step": 4076 + }, + { + "epoch": 2.932566085236468, + "grad_norm": 2.1090112562012178, + "learning_rate": 2.4165350920770032e-06, + "loss": 0.061, + "step": 4077 + }, + { + "epoch": 2.933285380327279, + "grad_norm": 3.8221194495351583, + "learning_rate": 2.416265726536517e-06, + "loss": 0.1289, + "step": 4078 + }, + { + "epoch": 2.93400467541809, + "grad_norm": 5.441863010607799, + "learning_rate": 2.4159963138525504e-06, + "loss": 0.0359, + "step": 4079 + }, + { + "epoch": 2.934723970508901, + "grad_norm": 3.5601061736795105, + "learning_rate": 2.415726854038965e-06, + "loss": 0.0479, + "step": 4080 + }, + { + "epoch": 2.9354432655997122, + "grad_norm": 1.6543937266672544, + "learning_rate": 2.4154573471096264e-06, + "loss": 0.0463, + "step": 4081 + }, + { + "epoch": 2.9361625606905233, + "grad_norm": 4.292833439786673, + "learning_rate": 2.4151877930783995e-06, + "loss": 0.11, + "step": 4082 + }, + { + "epoch": 2.9368818557813343, + "grad_norm": 0.10876797326340275, + "learning_rate": 2.4149181919591543e-06, + "loss": 0.0004, + "step": 4083 + }, + { + "epoch": 2.9376011508721454, + "grad_norm": 7.594886590458589, + "learning_rate": 2.414648543765762e-06, + "loss": 0.2957, + "step": 4084 + }, + { + "epoch": 2.9383204459629564, + "grad_norm": 4.326872976514883, + "learning_rate": 2.4143788485120967e-06, + "loss": 0.0915, + "step": 4085 + }, + { + "epoch": 2.9390397410537674, + "grad_norm": 1.7994498309696922, + "learning_rate": 2.4141091062120345e-06, + "loss": 0.0382, + "step": 4086 + }, + { + "epoch": 2.9397590361445785, + "grad_norm": 0.015075957072079137, + "learning_rate": 2.413839316879454e-06, + "loss": 0.0001, + "step": 4087 + }, + { + "epoch": 2.940478331235389, + "grad_norm": 4.7113750136763155, + "learning_rate": 2.4135694805282363e-06, + "loss": 0.1854, + "step": 4088 + }, + { + "epoch": 2.9411976263262005, + "grad_norm": 5.076628862837255, + "learning_rate": 2.4132995971722657e-06, + "loss": 0.0923, + "step": 4089 + }, + { + "epoch": 2.941916921417011, + "grad_norm": 5.084939676213631, + "learning_rate": 2.413029666825427e-06, + "loss": 0.1131, + "step": 4090 + }, + { + "epoch": 2.9426362165078226, + "grad_norm": 1.483487642108858, + "learning_rate": 2.412759689501609e-06, + "loss": 0.0252, + "step": 4091 + }, + { + "epoch": 2.943355511598633, + "grad_norm": 0.5106460693582094, + "learning_rate": 2.412489665214703e-06, + "loss": 0.0009, + "step": 4092 + }, + { + "epoch": 2.944074806689444, + "grad_norm": 1.9776705996474475, + "learning_rate": 2.412219593978602e-06, + "loss": 0.009, + "step": 4093 + }, + { + "epoch": 2.9447941017802552, + "grad_norm": 2.5431989767215004, + "learning_rate": 2.4119494758072014e-06, + "loss": 0.0105, + "step": 4094 + }, + { + "epoch": 2.9455133968710663, + "grad_norm": 0.13345312251002414, + "learning_rate": 2.4116793107143997e-06, + "loss": 0.0003, + "step": 4095 + }, + { + "epoch": 2.9462326919618773, + "grad_norm": 2.277061590681093, + "learning_rate": 2.4114090987140963e-06, + "loss": 0.0198, + "step": 4096 + }, + { + "epoch": 2.9469519870526883, + "grad_norm": 1.9406600013171222, + "learning_rate": 2.411138839820195e-06, + "loss": 0.0568, + "step": 4097 + }, + { + "epoch": 2.9476712821434994, + "grad_norm": 4.042199765571634, + "learning_rate": 2.4108685340466013e-06, + "loss": 0.179, + "step": 4098 + }, + { + "epoch": 2.9483905772343104, + "grad_norm": 3.1920518577664327, + "learning_rate": 2.410598181407223e-06, + "loss": 0.0625, + "step": 4099 + }, + { + "epoch": 2.9491098723251215, + "grad_norm": 2.06695238138367, + "learning_rate": 2.4103277819159686e-06, + "loss": 0.0603, + "step": 4100 + }, + { + "epoch": 2.9498291674159325, + "grad_norm": 1.662737001137136, + "learning_rate": 2.410057335586753e-06, + "loss": 0.0036, + "step": 4101 + }, + { + "epoch": 2.9505484625067435, + "grad_norm": 3.2805203402024303, + "learning_rate": 2.4097868424334896e-06, + "loss": 0.1487, + "step": 4102 + }, + { + "epoch": 2.9512677575975546, + "grad_norm": 4.528886246289247, + "learning_rate": 2.4095163024700957e-06, + "loss": 0.0906, + "step": 4103 + }, + { + "epoch": 2.9519870526883656, + "grad_norm": 0.8292465952872653, + "learning_rate": 2.4092457157104913e-06, + "loss": 0.002, + "step": 4104 + }, + { + "epoch": 2.952706347779176, + "grad_norm": 2.8809842182877103, + "learning_rate": 2.408975082168599e-06, + "loss": 0.0827, + "step": 4105 + }, + { + "epoch": 2.9534256428699877, + "grad_norm": 5.588086998950786, + "learning_rate": 2.408704401858343e-06, + "loss": 0.126, + "step": 4106 + }, + { + "epoch": 2.9541449379607982, + "grad_norm": 1.8061928646134922, + "learning_rate": 2.4084336747936505e-06, + "loss": 0.0644, + "step": 4107 + }, + { + "epoch": 2.9548642330516093, + "grad_norm": 3.5734074429837968, + "learning_rate": 2.4081629009884504e-06, + "loss": 0.1727, + "step": 4108 + }, + { + "epoch": 2.9555835281424203, + "grad_norm": 2.178123590626089, + "learning_rate": 2.407892080456675e-06, + "loss": 0.0556, + "step": 4109 + }, + { + "epoch": 2.9563028232332313, + "grad_norm": 2.6174168141438092, + "learning_rate": 2.4076212132122586e-06, + "loss": 0.1022, + "step": 4110 + }, + { + "epoch": 2.9570221183240424, + "grad_norm": 4.69771767275535, + "learning_rate": 2.4073502992691374e-06, + "loss": 0.1327, + "step": 4111 + }, + { + "epoch": 2.9577414134148534, + "grad_norm": 2.220444234192975, + "learning_rate": 2.40707933864125e-06, + "loss": 0.0328, + "step": 4112 + }, + { + "epoch": 2.9584607085056644, + "grad_norm": 2.6102616753250474, + "learning_rate": 2.4068083313425387e-06, + "loss": 0.0048, + "step": 4113 + }, + { + "epoch": 2.9591800035964755, + "grad_norm": 6.103492458754118, + "learning_rate": 2.406537277386947e-06, + "loss": 0.1393, + "step": 4114 + }, + { + "epoch": 2.9598992986872865, + "grad_norm": 4.016951525446478, + "learning_rate": 2.4062661767884207e-06, + "loss": 0.0847, + "step": 4115 + }, + { + "epoch": 2.9606185937780976, + "grad_norm": 3.433703290909037, + "learning_rate": 2.405995029560909e-06, + "loss": 0.0888, + "step": 4116 + }, + { + "epoch": 2.9613378888689086, + "grad_norm": 3.5172940546188016, + "learning_rate": 2.405723835718362e-06, + "loss": 0.0991, + "step": 4117 + }, + { + "epoch": 2.9620571839597196, + "grad_norm": 3.728253007345405, + "learning_rate": 2.4054525952747345e-06, + "loss": 0.1529, + "step": 4118 + }, + { + "epoch": 2.9627764790505307, + "grad_norm": 3.5878716793670042, + "learning_rate": 2.4051813082439813e-06, + "loss": 0.1431, + "step": 4119 + }, + { + "epoch": 2.9634957741413412, + "grad_norm": 2.0671486323487356, + "learning_rate": 2.40490997464006e-06, + "loss": 0.0621, + "step": 4120 + }, + { + "epoch": 2.9642150692321527, + "grad_norm": 4.068389413588118, + "learning_rate": 2.404638594476933e-06, + "loss": 0.0798, + "step": 4121 + }, + { + "epoch": 2.9649343643229633, + "grad_norm": 1.6161380640706662, + "learning_rate": 2.404367167768562e-06, + "loss": 0.0346, + "step": 4122 + }, + { + "epoch": 2.9656536594137743, + "grad_norm": 3.357229265397604, + "learning_rate": 2.4040956945289126e-06, + "loss": 0.0074, + "step": 4123 + }, + { + "epoch": 2.9663729545045854, + "grad_norm": 5.914602916531723, + "learning_rate": 2.4038241747719528e-06, + "loss": 0.3287, + "step": 4124 + }, + { + "epoch": 2.9670922495953964, + "grad_norm": 1.9470666319214112, + "learning_rate": 2.403552608511652e-06, + "loss": 0.0156, + "step": 4125 + }, + { + "epoch": 2.9678115446862074, + "grad_norm": 12.440589763005336, + "learning_rate": 2.4032809957619838e-06, + "loss": 0.1262, + "step": 4126 + }, + { + "epoch": 2.9685308397770185, + "grad_norm": 3.663332747719266, + "learning_rate": 2.403009336536923e-06, + "loss": 0.1644, + "step": 4127 + }, + { + "epoch": 2.9692501348678295, + "grad_norm": 2.646464936106486, + "learning_rate": 2.402737630850446e-06, + "loss": 0.0884, + "step": 4128 + }, + { + "epoch": 2.9699694299586405, + "grad_norm": 4.9178212709685685, + "learning_rate": 2.4024658787165335e-06, + "loss": 0.1932, + "step": 4129 + }, + { + "epoch": 2.9706887250494516, + "grad_norm": 3.7730786708412256, + "learning_rate": 2.402194080149167e-06, + "loss": 0.0667, + "step": 4130 + }, + { + "epoch": 2.9714080201402626, + "grad_norm": 4.759544024134562, + "learning_rate": 2.401922235162332e-06, + "loss": 0.1953, + "step": 4131 + }, + { + "epoch": 2.9721273152310737, + "grad_norm": 3.1101232914634624, + "learning_rate": 2.401650343770014e-06, + "loss": 0.0312, + "step": 4132 + }, + { + "epoch": 2.9728466103218847, + "grad_norm": 2.3170645809851673, + "learning_rate": 2.4013784059862038e-06, + "loss": 0.075, + "step": 4133 + }, + { + "epoch": 2.9735659054126957, + "grad_norm": 2.056849305860185, + "learning_rate": 2.401106421824892e-06, + "loss": 0.08, + "step": 4134 + }, + { + "epoch": 2.9742852005035063, + "grad_norm": 6.622587287512343, + "learning_rate": 2.4008343913000727e-06, + "loss": 0.1385, + "step": 4135 + }, + { + "epoch": 2.975004495594318, + "grad_norm": 1.9454634426209558, + "learning_rate": 2.4005623144257423e-06, + "loss": 0.046, + "step": 4136 + }, + { + "epoch": 2.9757237906851284, + "grad_norm": 2.5591183213031656, + "learning_rate": 2.4002901912159005e-06, + "loss": 0.0302, + "step": 4137 + }, + { + "epoch": 2.97644308577594, + "grad_norm": 3.3934713244465393, + "learning_rate": 2.400018021684548e-06, + "loss": 0.1026, + "step": 4138 + }, + { + "epoch": 2.9771623808667504, + "grad_norm": 3.7655848725890886, + "learning_rate": 2.399745805845688e-06, + "loss": 0.1699, + "step": 4139 + }, + { + "epoch": 2.9778816759575615, + "grad_norm": 2.6937700686001893, + "learning_rate": 2.399473543713327e-06, + "loss": 0.0716, + "step": 4140 + }, + { + "epoch": 2.9786009710483725, + "grad_norm": 13.1724771866324, + "learning_rate": 2.399201235301473e-06, + "loss": 0.2103, + "step": 4141 + }, + { + "epoch": 2.9793202661391835, + "grad_norm": 3.174580183326379, + "learning_rate": 2.3989288806241372e-06, + "loss": 0.0949, + "step": 4142 + }, + { + "epoch": 2.9800395612299946, + "grad_norm": 2.368162761869619, + "learning_rate": 2.398656479695332e-06, + "loss": 0.0288, + "step": 4143 + }, + { + "epoch": 2.9807588563208056, + "grad_norm": 4.826055368228739, + "learning_rate": 2.3983840325290744e-06, + "loss": 0.0422, + "step": 4144 + }, + { + "epoch": 2.9814781514116167, + "grad_norm": 1.4393542045677177, + "learning_rate": 2.3981115391393805e-06, + "loss": 0.0049, + "step": 4145 + }, + { + "epoch": 2.9821974465024277, + "grad_norm": 2.8331877704508943, + "learning_rate": 2.3978389995402718e-06, + "loss": 0.0906, + "step": 4146 + }, + { + "epoch": 2.9829167415932387, + "grad_norm": 3.698643699422827, + "learning_rate": 2.39756641374577e-06, + "loss": 0.1112, + "step": 4147 + }, + { + "epoch": 2.9836360366840498, + "grad_norm": 4.043395873149081, + "learning_rate": 2.397293781769901e-06, + "loss": 0.157, + "step": 4148 + }, + { + "epoch": 2.984355331774861, + "grad_norm": 7.529882959656265, + "learning_rate": 2.397021103626691e-06, + "loss": 0.1253, + "step": 4149 + }, + { + "epoch": 2.9850746268656714, + "grad_norm": 5.1465982499934375, + "learning_rate": 2.396748379330171e-06, + "loss": 0.0571, + "step": 4150 + }, + { + "epoch": 2.985793921956483, + "grad_norm": 4.1784428736381, + "learning_rate": 2.3964756088943726e-06, + "loss": 0.1348, + "step": 4151 + }, + { + "epoch": 2.9865132170472934, + "grad_norm": 4.2778857605226825, + "learning_rate": 2.3962027923333308e-06, + "loss": 0.0214, + "step": 4152 + }, + { + "epoch": 2.987232512138105, + "grad_norm": 2.0542372568580376, + "learning_rate": 2.395929929661082e-06, + "loss": 0.0288, + "step": 4153 + }, + { + "epoch": 2.9879518072289155, + "grad_norm": 3.854234059046007, + "learning_rate": 2.395657020891665e-06, + "loss": 0.1617, + "step": 4154 + }, + { + "epoch": 2.9886711023197265, + "grad_norm": 2.8655471471596616, + "learning_rate": 2.3953840660391227e-06, + "loss": 0.1316, + "step": 4155 + }, + { + "epoch": 2.9893903974105376, + "grad_norm": 4.3688363590003165, + "learning_rate": 2.395111065117498e-06, + "loss": 0.1842, + "step": 4156 + }, + { + "epoch": 2.9901096925013486, + "grad_norm": 3.1108634363265653, + "learning_rate": 2.394838018140838e-06, + "loss": 0.0865, + "step": 4157 + }, + { + "epoch": 2.9908289875921596, + "grad_norm": 4.690860341645325, + "learning_rate": 2.3945649251231912e-06, + "loss": 0.0129, + "step": 4158 + }, + { + "epoch": 2.9915482826829707, + "grad_norm": 2.745800892612085, + "learning_rate": 2.3942917860786078e-06, + "loss": 0.0111, + "step": 4159 + }, + { + "epoch": 2.9922675777737817, + "grad_norm": 1.3408737233608998, + "learning_rate": 2.3940186010211427e-06, + "loss": 0.0077, + "step": 4160 + }, + { + "epoch": 2.9929868728645928, + "grad_norm": 1.9225141314379461, + "learning_rate": 2.393745369964851e-06, + "loss": 0.0089, + "step": 4161 + }, + { + "epoch": 2.993706167955404, + "grad_norm": 1.652740339238617, + "learning_rate": 2.3934720929237915e-06, + "loss": 0.0123, + "step": 4162 + }, + { + "epoch": 2.994425463046215, + "grad_norm": 3.463711466785171, + "learning_rate": 2.3931987699120244e-06, + "loss": 0.0234, + "step": 4163 + }, + { + "epoch": 2.995144758137026, + "grad_norm": 5.3191335017786985, + "learning_rate": 2.3929254009436123e-06, + "loss": 0.1326, + "step": 4164 + }, + { + "epoch": 2.9958640532278364, + "grad_norm": 2.639773402631255, + "learning_rate": 2.392651986032621e-06, + "loss": 0.0689, + "step": 4165 + }, + { + "epoch": 2.996583348318648, + "grad_norm": 0.2142367408113003, + "learning_rate": 2.392378525193118e-06, + "loss": 0.0005, + "step": 4166 + }, + { + "epoch": 2.9973026434094585, + "grad_norm": 2.5295504667030615, + "learning_rate": 2.392105018439173e-06, + "loss": 0.0459, + "step": 4167 + }, + { + "epoch": 2.99802193850027, + "grad_norm": 4.44897061246736, + "learning_rate": 2.3918314657848593e-06, + "loss": 0.0974, + "step": 4168 + }, + { + "epoch": 2.9987412335910806, + "grad_norm": 3.359496912409121, + "learning_rate": 2.3915578672442507e-06, + "loss": 0.0505, + "step": 4169 + }, + { + "epoch": 2.9994605286818916, + "grad_norm": 2.470114387979346, + "learning_rate": 2.3912842228314247e-06, + "loss": 0.0072, + "step": 4170 + }, + { + "epoch": 3.0001798237727026, + "grad_norm": 3.59122860591066, + "learning_rate": 2.3910105325604615e-06, + "loss": 0.0332, + "step": 4171 + }, + { + "epoch": 3.0008991188635137, + "grad_norm": 0.6169503159338277, + "learning_rate": 2.3907367964454414e-06, + "loss": 0.0023, + "step": 4172 + }, + { + "epoch": 3.0016184139543247, + "grad_norm": 0.9037409611200752, + "learning_rate": 2.39046301450045e-06, + "loss": 0.0036, + "step": 4173 + }, + { + "epoch": 3.0023377090451357, + "grad_norm": 3.9201807625746623, + "learning_rate": 2.3901891867395736e-06, + "loss": 0.2522, + "step": 4174 + }, + { + "epoch": 3.003057004135947, + "grad_norm": 1.7476328669061456, + "learning_rate": 2.3899153131769004e-06, + "loss": 0.0408, + "step": 4175 + }, + { + "epoch": 3.003776299226758, + "grad_norm": 5.552091647776159, + "learning_rate": 2.3896413938265225e-06, + "loss": 0.243, + "step": 4176 + }, + { + "epoch": 3.004495594317569, + "grad_norm": 1.8745584996485445, + "learning_rate": 2.389367428702533e-06, + "loss": 0.0279, + "step": 4177 + }, + { + "epoch": 3.00521488940838, + "grad_norm": 5.520559230484063, + "learning_rate": 2.3890934178190278e-06, + "loss": 0.1771, + "step": 4178 + }, + { + "epoch": 3.005934184499191, + "grad_norm": 2.2436343713028415, + "learning_rate": 2.388819361190106e-06, + "loss": 0.0779, + "step": 4179 + }, + { + "epoch": 3.006653479590002, + "grad_norm": 3.941050956450359, + "learning_rate": 2.388545258829867e-06, + "loss": 0.1301, + "step": 4180 + }, + { + "epoch": 3.007372774680813, + "grad_norm": 1.2230568613484318, + "learning_rate": 2.3882711107524152e-06, + "loss": 0.0191, + "step": 4181 + }, + { + "epoch": 3.008092069771624, + "grad_norm": 2.134468316102497, + "learning_rate": 2.387996916971855e-06, + "loss": 0.0313, + "step": 4182 + }, + { + "epoch": 3.0088113648624346, + "grad_norm": 2.6230265454737904, + "learning_rate": 2.3877226775022956e-06, + "loss": 0.0465, + "step": 4183 + }, + { + "epoch": 3.0095306599532456, + "grad_norm": 2.6253203439982187, + "learning_rate": 2.3874483923578456e-06, + "loss": 0.0632, + "step": 4184 + }, + { + "epoch": 3.0102499550440567, + "grad_norm": 3.860021219245699, + "learning_rate": 2.3871740615526174e-06, + "loss": 0.0858, + "step": 4185 + }, + { + "epoch": 3.0109692501348677, + "grad_norm": 4.717155763948642, + "learning_rate": 2.386899685100727e-06, + "loss": 0.1256, + "step": 4186 + }, + { + "epoch": 3.0116885452256787, + "grad_norm": 1.1979552935091249, + "learning_rate": 2.3866252630162906e-06, + "loss": 0.0015, + "step": 4187 + }, + { + "epoch": 3.01240784031649, + "grad_norm": 1.3444719069735305, + "learning_rate": 2.3863507953134277e-06, + "loss": 0.032, + "step": 4188 + }, + { + "epoch": 3.013127135407301, + "grad_norm": 3.0218792562505294, + "learning_rate": 2.3860762820062613e-06, + "loss": 0.0957, + "step": 4189 + }, + { + "epoch": 3.013846430498112, + "grad_norm": 5.539820067894367, + "learning_rate": 2.385801723108914e-06, + "loss": 0.1122, + "step": 4190 + }, + { + "epoch": 3.014565725588923, + "grad_norm": 2.839439274299694, + "learning_rate": 2.3855271186355133e-06, + "loss": 0.14, + "step": 4191 + }, + { + "epoch": 3.015285020679734, + "grad_norm": 6.025425588496945, + "learning_rate": 2.385252468600188e-06, + "loss": 0.1919, + "step": 4192 + }, + { + "epoch": 3.016004315770545, + "grad_norm": 2.651902982370502, + "learning_rate": 2.384977773017069e-06, + "loss": 0.0765, + "step": 4193 + }, + { + "epoch": 3.016723610861356, + "grad_norm": 5.584723435995463, + "learning_rate": 2.38470303190029e-06, + "loss": 0.243, + "step": 4194 + }, + { + "epoch": 3.017442905952167, + "grad_norm": 0.21702919331759238, + "learning_rate": 2.384428245263987e-06, + "loss": 0.0007, + "step": 4195 + }, + { + "epoch": 3.018162201042978, + "grad_norm": 3.959544896838745, + "learning_rate": 2.3841534131222983e-06, + "loss": 0.0944, + "step": 4196 + }, + { + "epoch": 3.018881496133789, + "grad_norm": 8.091757801544954, + "learning_rate": 2.3838785354893643e-06, + "loss": 0.0883, + "step": 4197 + }, + { + "epoch": 3.0196007912245997, + "grad_norm": 5.334334448935455, + "learning_rate": 2.3836036123793284e-06, + "loss": 0.1804, + "step": 4198 + }, + { + "epoch": 3.0203200863154107, + "grad_norm": 2.4739603545097917, + "learning_rate": 2.3833286438063347e-06, + "loss": 0.0344, + "step": 4199 + }, + { + "epoch": 3.0210393814062217, + "grad_norm": 1.8935032969526147, + "learning_rate": 2.383053629784533e-06, + "loss": 0.0413, + "step": 4200 + }, + { + "epoch": 3.021758676497033, + "grad_norm": 1.7693898253045113, + "learning_rate": 2.3827785703280705e-06, + "loss": 0.049, + "step": 4201 + }, + { + "epoch": 3.022477971587844, + "grad_norm": 0.9341041953345984, + "learning_rate": 2.3825034654511016e-06, + "loss": 0.0051, + "step": 4202 + }, + { + "epoch": 3.023197266678655, + "grad_norm": 3.376126724265184, + "learning_rate": 2.3822283151677804e-06, + "loss": 0.044, + "step": 4203 + }, + { + "epoch": 3.023916561769466, + "grad_norm": 6.430254996723841, + "learning_rate": 2.381953119492263e-06, + "loss": 0.0397, + "step": 4204 + }, + { + "epoch": 3.024635856860277, + "grad_norm": 3.146090713379502, + "learning_rate": 2.3816778784387097e-06, + "loss": 0.1337, + "step": 4205 + }, + { + "epoch": 3.025355151951088, + "grad_norm": 0.9404201372790487, + "learning_rate": 2.381402592021282e-06, + "loss": 0.0013, + "step": 4206 + }, + { + "epoch": 3.026074447041899, + "grad_norm": 3.040531749485178, + "learning_rate": 2.3811272602541435e-06, + "loss": 0.1072, + "step": 4207 + }, + { + "epoch": 3.02679374213271, + "grad_norm": 2.7331294212449406, + "learning_rate": 2.380851883151461e-06, + "loss": 0.1086, + "step": 4208 + }, + { + "epoch": 3.027513037223521, + "grad_norm": 0.7813001789305579, + "learning_rate": 2.3805764607274025e-06, + "loss": 0.0107, + "step": 4209 + }, + { + "epoch": 3.028232332314332, + "grad_norm": 0.8467279756161453, + "learning_rate": 2.3803009929961392e-06, + "loss": 0.0107, + "step": 4210 + }, + { + "epoch": 3.028951627405143, + "grad_norm": 1.9139865378183534, + "learning_rate": 2.380025479971845e-06, + "loss": 0.0684, + "step": 4211 + }, + { + "epoch": 3.029670922495954, + "grad_norm": 4.271522356206117, + "learning_rate": 2.3797499216686944e-06, + "loss": 0.0632, + "step": 4212 + }, + { + "epoch": 3.030390217586765, + "grad_norm": 3.2269966076346774, + "learning_rate": 2.379474318100866e-06, + "loss": 0.0057, + "step": 4213 + }, + { + "epoch": 3.031109512677576, + "grad_norm": 4.38574000228918, + "learning_rate": 2.3791986692825403e-06, + "loss": 0.0705, + "step": 4214 + }, + { + "epoch": 3.031828807768387, + "grad_norm": 4.599974964173051, + "learning_rate": 2.3789229752279e-06, + "loss": 0.0479, + "step": 4215 + }, + { + "epoch": 3.032548102859198, + "grad_norm": 3.428312100683051, + "learning_rate": 2.3786472359511294e-06, + "loss": 0.0975, + "step": 4216 + }, + { + "epoch": 3.033267397950009, + "grad_norm": 2.890842723546386, + "learning_rate": 2.378371451466416e-06, + "loss": 0.0647, + "step": 4217 + }, + { + "epoch": 3.03398669304082, + "grad_norm": 0.5021541169120519, + "learning_rate": 2.3780956217879497e-06, + "loss": 0.0036, + "step": 4218 + }, + { + "epoch": 3.034705988131631, + "grad_norm": 0.9431453515040041, + "learning_rate": 2.377819746929922e-06, + "loss": 0.004, + "step": 4219 + }, + { + "epoch": 3.035425283222442, + "grad_norm": 2.80218662382666, + "learning_rate": 2.3775438269065277e-06, + "loss": 0.085, + "step": 4220 + }, + { + "epoch": 3.036144578313253, + "grad_norm": 2.194163801698843, + "learning_rate": 2.3772678617319625e-06, + "loss": 0.0053, + "step": 4221 + }, + { + "epoch": 3.036863873404064, + "grad_norm": 0.943174449177465, + "learning_rate": 2.376991851420426e-06, + "loss": 0.0035, + "step": 4222 + }, + { + "epoch": 3.037583168494875, + "grad_norm": 1.1345296273371606, + "learning_rate": 2.3767157959861194e-06, + "loss": 0.0135, + "step": 4223 + }, + { + "epoch": 3.038302463585686, + "grad_norm": 0.947695332204924, + "learning_rate": 2.3764396954432463e-06, + "loss": 0.0026, + "step": 4224 + }, + { + "epoch": 3.039021758676497, + "grad_norm": 2.085304078368595, + "learning_rate": 2.3761635498060116e-06, + "loss": 0.0468, + "step": 4225 + }, + { + "epoch": 3.039741053767308, + "grad_norm": 2.747362085775327, + "learning_rate": 2.375887359088625e-06, + "loss": 0.0409, + "step": 4226 + }, + { + "epoch": 3.040460348858119, + "grad_norm": 3.118929709318799, + "learning_rate": 2.3756111233052956e-06, + "loss": 0.0739, + "step": 4227 + }, + { + "epoch": 3.0411796439489303, + "grad_norm": 5.440501693345429, + "learning_rate": 2.3753348424702375e-06, + "loss": 0.1631, + "step": 4228 + }, + { + "epoch": 3.041898939039741, + "grad_norm": 4.482552556565129, + "learning_rate": 2.3750585165976645e-06, + "loss": 0.0973, + "step": 4229 + }, + { + "epoch": 3.042618234130552, + "grad_norm": 2.864436996747955, + "learning_rate": 2.374782145701795e-06, + "loss": 0.0919, + "step": 4230 + }, + { + "epoch": 3.043337529221363, + "grad_norm": 2.461399628082142, + "learning_rate": 2.374505729796848e-06, + "loss": 0.0846, + "step": 4231 + }, + { + "epoch": 3.044056824312174, + "grad_norm": 2.2924828328619595, + "learning_rate": 2.374229268897047e-06, + "loss": 0.0085, + "step": 4232 + }, + { + "epoch": 3.044776119402985, + "grad_norm": 2.618717050607476, + "learning_rate": 2.373952763016615e-06, + "loss": 0.0666, + "step": 4233 + }, + { + "epoch": 3.045495414493796, + "grad_norm": 1.14162621878143, + "learning_rate": 2.373676212169779e-06, + "loss": 0.0029, + "step": 4234 + }, + { + "epoch": 3.046214709584607, + "grad_norm": 2.7276185291154365, + "learning_rate": 2.3733996163707682e-06, + "loss": 0.0464, + "step": 4235 + }, + { + "epoch": 3.046934004675418, + "grad_norm": 3.208819015070432, + "learning_rate": 2.373122975633814e-06, + "loss": 0.1133, + "step": 4236 + }, + { + "epoch": 3.047653299766229, + "grad_norm": 3.697283228708233, + "learning_rate": 2.3728462899731503e-06, + "loss": 0.1071, + "step": 4237 + }, + { + "epoch": 3.04837259485704, + "grad_norm": 4.019303212686577, + "learning_rate": 2.372569559403012e-06, + "loss": 0.1478, + "step": 4238 + }, + { + "epoch": 3.049091889947851, + "grad_norm": 2.9207523777158064, + "learning_rate": 2.372292783937639e-06, + "loss": 0.0204, + "step": 4239 + }, + { + "epoch": 3.049811185038662, + "grad_norm": 2.9299556598537184, + "learning_rate": 2.37201596359127e-06, + "loss": 0.0083, + "step": 4240 + }, + { + "epoch": 3.0505304801294733, + "grad_norm": 2.9216239530004393, + "learning_rate": 2.3717390983781496e-06, + "loss": 0.0317, + "step": 4241 + }, + { + "epoch": 3.0512497752202843, + "grad_norm": 2.608884176859807, + "learning_rate": 2.3714621883125223e-06, + "loss": 0.0697, + "step": 4242 + }, + { + "epoch": 3.0519690703110953, + "grad_norm": 1.2258199667983332, + "learning_rate": 2.3711852334086353e-06, + "loss": 0.0158, + "step": 4243 + }, + { + "epoch": 3.052688365401906, + "grad_norm": 2.4133042469503865, + "learning_rate": 2.3709082336807394e-06, + "loss": 0.0997, + "step": 4244 + }, + { + "epoch": 3.053407660492717, + "grad_norm": 0.37056642016267966, + "learning_rate": 2.3706311891430852e-06, + "loss": 0.0025, + "step": 4245 + }, + { + "epoch": 3.054126955583528, + "grad_norm": 2.076519312693626, + "learning_rate": 2.3703540998099285e-06, + "loss": 0.0499, + "step": 4246 + }, + { + "epoch": 3.054846250674339, + "grad_norm": 3.0397354064361455, + "learning_rate": 2.370076965695525e-06, + "loss": 0.0074, + "step": 4247 + }, + { + "epoch": 3.05556554576515, + "grad_norm": 4.750150026330861, + "learning_rate": 2.3697997868141346e-06, + "loss": 0.0148, + "step": 4248 + }, + { + "epoch": 3.056284840855961, + "grad_norm": 0.7002074750284241, + "learning_rate": 2.3695225631800188e-06, + "loss": 0.0018, + "step": 4249 + }, + { + "epoch": 3.057004135946772, + "grad_norm": 3.513798130996827, + "learning_rate": 2.3692452948074395e-06, + "loss": 0.1049, + "step": 4250 + }, + { + "epoch": 3.057723431037583, + "grad_norm": 1.5761236159852587, + "learning_rate": 2.3689679817106648e-06, + "loss": 0.0227, + "step": 4251 + }, + { + "epoch": 3.058442726128394, + "grad_norm": 1.8652839386456317, + "learning_rate": 2.368690623903962e-06, + "loss": 0.0555, + "step": 4252 + }, + { + "epoch": 3.059162021219205, + "grad_norm": 1.2630611348212153, + "learning_rate": 2.3684132214016012e-06, + "loss": 0.0208, + "step": 4253 + }, + { + "epoch": 3.0598813163100163, + "grad_norm": 4.613969455167519, + "learning_rate": 2.368135774217856e-06, + "loss": 0.1331, + "step": 4254 + }, + { + "epoch": 3.0606006114008273, + "grad_norm": 7.9408672917409655, + "learning_rate": 2.367858282367001e-06, + "loss": 0.2847, + "step": 4255 + }, + { + "epoch": 3.0613199064916383, + "grad_norm": 2.611169267826939, + "learning_rate": 2.367580745863314e-06, + "loss": 0.0473, + "step": 4256 + }, + { + "epoch": 3.0620392015824494, + "grad_norm": 4.657516864019772, + "learning_rate": 2.367303164721075e-06, + "loss": 0.1567, + "step": 4257 + }, + { + "epoch": 3.0627584966732604, + "grad_norm": 5.592713514327501, + "learning_rate": 2.367025538954565e-06, + "loss": 0.1874, + "step": 4258 + }, + { + "epoch": 3.0634777917640714, + "grad_norm": 3.344231730713646, + "learning_rate": 2.3667478685780696e-06, + "loss": 0.0883, + "step": 4259 + }, + { + "epoch": 3.064197086854882, + "grad_norm": 4.46359162643568, + "learning_rate": 2.366470153605875e-06, + "loss": 0.111, + "step": 4260 + }, + { + "epoch": 3.064916381945693, + "grad_norm": 3.4810248485826074, + "learning_rate": 2.3661923940522688e-06, + "loss": 0.0874, + "step": 4261 + }, + { + "epoch": 3.065635677036504, + "grad_norm": 3.5521754248150783, + "learning_rate": 2.3659145899315443e-06, + "loss": 0.1066, + "step": 4262 + }, + { + "epoch": 3.066354972127315, + "grad_norm": 2.1156323313741363, + "learning_rate": 2.3656367412579944e-06, + "loss": 0.0219, + "step": 4263 + }, + { + "epoch": 3.067074267218126, + "grad_norm": 2.8901210133463264, + "learning_rate": 2.365358848045914e-06, + "loss": 0.1019, + "step": 4264 + }, + { + "epoch": 3.067793562308937, + "grad_norm": 0.2997566649560629, + "learning_rate": 2.365080910309602e-06, + "loss": 0.0006, + "step": 4265 + }, + { + "epoch": 3.068512857399748, + "grad_norm": 4.486524192683793, + "learning_rate": 2.3648029280633583e-06, + "loss": 0.0114, + "step": 4266 + }, + { + "epoch": 3.0692321524905593, + "grad_norm": 6.182350158815735, + "learning_rate": 2.3645249013214863e-06, + "loss": 0.028, + "step": 4267 + }, + { + "epoch": 3.0699514475813703, + "grad_norm": 4.254354420486257, + "learning_rate": 2.36424683009829e-06, + "loss": 0.1259, + "step": 4268 + }, + { + "epoch": 3.0706707426721813, + "grad_norm": 1.9498966249832805, + "learning_rate": 2.363968714408077e-06, + "loss": 0.0569, + "step": 4269 + }, + { + "epoch": 3.0713900377629924, + "grad_norm": 4.125931196586699, + "learning_rate": 2.3636905542651575e-06, + "loss": 0.143, + "step": 4270 + }, + { + "epoch": 3.0721093328538034, + "grad_norm": 5.334595462075598, + "learning_rate": 2.3634123496838426e-06, + "loss": 0.1689, + "step": 4271 + }, + { + "epoch": 3.0728286279446144, + "grad_norm": 1.5484097879141911, + "learning_rate": 2.363134100678447e-06, + "loss": 0.0023, + "step": 4272 + }, + { + "epoch": 3.0735479230354255, + "grad_norm": 3.7753831885329063, + "learning_rate": 2.3628558072632864e-06, + "loss": 0.0657, + "step": 4273 + }, + { + "epoch": 3.0742672181262365, + "grad_norm": 2.839757893740574, + "learning_rate": 2.3625774694526798e-06, + "loss": 0.0154, + "step": 4274 + }, + { + "epoch": 3.074986513217047, + "grad_norm": 3.802846975250464, + "learning_rate": 2.3622990872609484e-06, + "loss": 0.036, + "step": 4275 + }, + { + "epoch": 3.075705808307858, + "grad_norm": 1.0863605541048587, + "learning_rate": 2.362020660702415e-06, + "loss": 0.0125, + "step": 4276 + }, + { + "epoch": 3.076425103398669, + "grad_norm": 1.3829114362419619, + "learning_rate": 2.3617421897914056e-06, + "loss": 0.0178, + "step": 4277 + }, + { + "epoch": 3.07714439848948, + "grad_norm": 2.1706024767608665, + "learning_rate": 2.3614636745422475e-06, + "loss": 0.0454, + "step": 4278 + }, + { + "epoch": 3.077863693580291, + "grad_norm": 4.155052721344708, + "learning_rate": 2.3611851149692716e-06, + "loss": 0.0134, + "step": 4279 + }, + { + "epoch": 3.0785829886711022, + "grad_norm": 4.3679023295653865, + "learning_rate": 2.3609065110868087e-06, + "loss": 0.1243, + "step": 4280 + }, + { + "epoch": 3.0793022837619133, + "grad_norm": 3.2743271295106804, + "learning_rate": 2.360627862909196e-06, + "loss": 0.1273, + "step": 4281 + }, + { + "epoch": 3.0800215788527243, + "grad_norm": 3.0484755149030303, + "learning_rate": 2.360349170450768e-06, + "loss": 0.0696, + "step": 4282 + }, + { + "epoch": 3.0807408739435354, + "grad_norm": 3.7898927680087153, + "learning_rate": 2.3600704337258647e-06, + "loss": 0.1295, + "step": 4283 + }, + { + "epoch": 3.0814601690343464, + "grad_norm": 5.614572298420928, + "learning_rate": 2.359791652748829e-06, + "loss": 0.0927, + "step": 4284 + }, + { + "epoch": 3.0821794641251574, + "grad_norm": 3.055719870226799, + "learning_rate": 2.3595128275340023e-06, + "loss": 0.0787, + "step": 4285 + }, + { + "epoch": 3.0828987592159685, + "grad_norm": 3.0982478577483175, + "learning_rate": 2.359233958095732e-06, + "loss": 0.0227, + "step": 4286 + }, + { + "epoch": 3.0836180543067795, + "grad_norm": 1.3530381316982012, + "learning_rate": 2.358955044448367e-06, + "loss": 0.0042, + "step": 4287 + }, + { + "epoch": 3.0843373493975905, + "grad_norm": 2.3582436814956123, + "learning_rate": 2.358676086606256e-06, + "loss": 0.059, + "step": 4288 + }, + { + "epoch": 3.0850566444884016, + "grad_norm": 4.760299677220257, + "learning_rate": 2.3583970845837532e-06, + "loss": 0.1986, + "step": 4289 + }, + { + "epoch": 3.0857759395792126, + "grad_norm": 3.2049452017410696, + "learning_rate": 2.3581180383952136e-06, + "loss": 0.0372, + "step": 4290 + }, + { + "epoch": 3.086495234670023, + "grad_norm": 2.417617986286057, + "learning_rate": 2.357838948054995e-06, + "loss": 0.0622, + "step": 4291 + }, + { + "epoch": 3.087214529760834, + "grad_norm": 3.997759165906392, + "learning_rate": 2.3575598135774566e-06, + "loss": 0.2331, + "step": 4292 + }, + { + "epoch": 3.0879338248516452, + "grad_norm": 2.8098320071384792, + "learning_rate": 2.35728063497696e-06, + "loss": 0.0641, + "step": 4293 + }, + { + "epoch": 3.0886531199424563, + "grad_norm": 3.3666155757235465, + "learning_rate": 2.3570014122678704e-06, + "loss": 0.0883, + "step": 4294 + }, + { + "epoch": 3.0893724150332673, + "grad_norm": 3.025132373950285, + "learning_rate": 2.3567221454645527e-06, + "loss": 0.1054, + "step": 4295 + }, + { + "epoch": 3.0900917101240784, + "grad_norm": 2.5349951182898525, + "learning_rate": 2.3564428345813777e-06, + "loss": 0.0416, + "step": 4296 + }, + { + "epoch": 3.0908110052148894, + "grad_norm": 3.175242898859685, + "learning_rate": 2.3561634796327147e-06, + "loss": 0.1184, + "step": 4297 + }, + { + "epoch": 3.0915303003057004, + "grad_norm": 3.264749656039814, + "learning_rate": 2.355884080632938e-06, + "loss": 0.0854, + "step": 4298 + }, + { + "epoch": 3.0922495953965115, + "grad_norm": 4.211797112789473, + "learning_rate": 2.355604637596423e-06, + "loss": 0.0912, + "step": 4299 + }, + { + "epoch": 3.0929688904873225, + "grad_norm": 1.9526795931431653, + "learning_rate": 2.3553251505375475e-06, + "loss": 0.0346, + "step": 4300 + }, + { + "epoch": 3.0936881855781335, + "grad_norm": 2.891672190616186, + "learning_rate": 2.3550456194706914e-06, + "loss": 0.0583, + "step": 4301 + }, + { + "epoch": 3.0944074806689446, + "grad_norm": 2.1798682428435843, + "learning_rate": 2.354766044410237e-06, + "loss": 0.0691, + "step": 4302 + }, + { + "epoch": 3.0951267757597556, + "grad_norm": 2.8837444574227895, + "learning_rate": 2.354486425370569e-06, + "loss": 0.0173, + "step": 4303 + }, + { + "epoch": 3.0958460708505666, + "grad_norm": 3.996177179250047, + "learning_rate": 2.354206762366075e-06, + "loss": 0.1788, + "step": 4304 + }, + { + "epoch": 3.0965653659413777, + "grad_norm": 0.3590702367727158, + "learning_rate": 2.3539270554111435e-06, + "loss": 0.0029, + "step": 4305 + }, + { + "epoch": 3.0972846610321882, + "grad_norm": 2.0487003853320527, + "learning_rate": 2.3536473045201658e-06, + "loss": 0.0401, + "step": 4306 + }, + { + "epoch": 3.0980039561229993, + "grad_norm": 1.546319294010759, + "learning_rate": 2.3533675097075355e-06, + "loss": 0.0259, + "step": 4307 + }, + { + "epoch": 3.0987232512138103, + "grad_norm": 1.077268305263352, + "learning_rate": 2.3530876709876492e-06, + "loss": 0.0073, + "step": 4308 + }, + { + "epoch": 3.0994425463046213, + "grad_norm": 3.038848824000709, + "learning_rate": 2.3528077883749044e-06, + "loss": 0.1274, + "step": 4309 + }, + { + "epoch": 3.1001618413954324, + "grad_norm": 1.3542669452799583, + "learning_rate": 2.352527861883702e-06, + "loss": 0.031, + "step": 4310 + }, + { + "epoch": 3.1008811364862434, + "grad_norm": 2.630013177638445, + "learning_rate": 2.3522478915284443e-06, + "loss": 0.023, + "step": 4311 + }, + { + "epoch": 3.1016004315770545, + "grad_norm": 3.393761631108634, + "learning_rate": 2.351967877323537e-06, + "loss": 0.0583, + "step": 4312 + }, + { + "epoch": 3.1023197266678655, + "grad_norm": 0.9049014106597464, + "learning_rate": 2.351687819283386e-06, + "loss": 0.0044, + "step": 4313 + }, + { + "epoch": 3.1030390217586765, + "grad_norm": 6.652411489218541, + "learning_rate": 2.351407717422402e-06, + "loss": 0.1481, + "step": 4314 + }, + { + "epoch": 3.1037583168494876, + "grad_norm": 0.3874367078146442, + "learning_rate": 2.351127571754997e-06, + "loss": 0.0012, + "step": 4315 + }, + { + "epoch": 3.1044776119402986, + "grad_norm": 3.737806608133162, + "learning_rate": 2.3508473822955833e-06, + "loss": 0.0637, + "step": 4316 + }, + { + "epoch": 3.1051969070311096, + "grad_norm": 4.028230804251777, + "learning_rate": 2.350567149058579e-06, + "loss": 0.0655, + "step": 4317 + }, + { + "epoch": 3.1059162021219207, + "grad_norm": 4.619850208644465, + "learning_rate": 2.3502868720584012e-06, + "loss": 0.1221, + "step": 4318 + }, + { + "epoch": 3.1066354972127317, + "grad_norm": 3.1554251194268703, + "learning_rate": 2.350006551309471e-06, + "loss": 0.0741, + "step": 4319 + }, + { + "epoch": 3.1073547923035427, + "grad_norm": 2.8972149475938838, + "learning_rate": 2.3497261868262125e-06, + "loss": 0.0521, + "step": 4320 + }, + { + "epoch": 3.1080740873943533, + "grad_norm": 4.270402225725747, + "learning_rate": 2.3494457786230496e-06, + "loss": 0.198, + "step": 4321 + }, + { + "epoch": 3.1087933824851643, + "grad_norm": 2.3661587186136623, + "learning_rate": 2.3491653267144097e-06, + "loss": 0.0751, + "step": 4322 + }, + { + "epoch": 3.1095126775759754, + "grad_norm": 1.6209350706745373, + "learning_rate": 2.348884831114724e-06, + "loss": 0.0394, + "step": 4323 + }, + { + "epoch": 3.1102319726667864, + "grad_norm": 3.401284027484061, + "learning_rate": 2.3486042918384227e-06, + "loss": 0.1487, + "step": 4324 + }, + { + "epoch": 3.1109512677575974, + "grad_norm": 2.5718903872531147, + "learning_rate": 2.3483237088999416e-06, + "loss": 0.0495, + "step": 4325 + }, + { + "epoch": 3.1116705628484085, + "grad_norm": 4.056533512076855, + "learning_rate": 2.3480430823137164e-06, + "loss": 0.0835, + "step": 4326 + }, + { + "epoch": 3.1123898579392195, + "grad_norm": 4.296967812275655, + "learning_rate": 2.347762412094185e-06, + "loss": 0.0833, + "step": 4327 + }, + { + "epoch": 3.1131091530300306, + "grad_norm": 3.0100353057211184, + "learning_rate": 2.3474816982557907e-06, + "loss": 0.0452, + "step": 4328 + }, + { + "epoch": 3.1138284481208416, + "grad_norm": 4.925011069746328, + "learning_rate": 2.3472009408129744e-06, + "loss": 0.0975, + "step": 4329 + }, + { + "epoch": 3.1145477432116526, + "grad_norm": 2.8372176835274674, + "learning_rate": 2.346920139780183e-06, + "loss": 0.0761, + "step": 4330 + }, + { + "epoch": 3.1152670383024637, + "grad_norm": 5.141305497682005, + "learning_rate": 2.346639295171863e-06, + "loss": 0.1165, + "step": 4331 + }, + { + "epoch": 3.1159863333932747, + "grad_norm": 2.6174047495789323, + "learning_rate": 2.3463584070024654e-06, + "loss": 0.0171, + "step": 4332 + }, + { + "epoch": 3.1167056284840857, + "grad_norm": 4.752582155482094, + "learning_rate": 2.3460774752864424e-06, + "loss": 0.0586, + "step": 4333 + }, + { + "epoch": 3.1174249235748968, + "grad_norm": 3.049226264561142, + "learning_rate": 2.345796500038248e-06, + "loss": 0.0458, + "step": 4334 + }, + { + "epoch": 3.118144218665708, + "grad_norm": 0.38172098568531765, + "learning_rate": 2.3455154812723384e-06, + "loss": 0.0009, + "step": 4335 + }, + { + "epoch": 3.118863513756519, + "grad_norm": 2.1839789865994983, + "learning_rate": 2.345234419003173e-06, + "loss": 0.0073, + "step": 4336 + }, + { + "epoch": 3.1195828088473294, + "grad_norm": 5.145329011124619, + "learning_rate": 2.3449533132452134e-06, + "loss": 0.1399, + "step": 4337 + }, + { + "epoch": 3.1203021039381404, + "grad_norm": 2.912096667122734, + "learning_rate": 2.3446721640129223e-06, + "loss": 0.0631, + "step": 4338 + }, + { + "epoch": 3.1210213990289515, + "grad_norm": 3.0175501768625073, + "learning_rate": 2.344390971320766e-06, + "loss": 0.0729, + "step": 4339 + }, + { + "epoch": 3.1217406941197625, + "grad_norm": 4.234504107732753, + "learning_rate": 2.344109735183211e-06, + "loss": 0.0736, + "step": 4340 + }, + { + "epoch": 3.1224599892105735, + "grad_norm": 3.084957838339434, + "learning_rate": 2.3438284556147294e-06, + "loss": 0.0761, + "step": 4341 + }, + { + "epoch": 3.1231792843013846, + "grad_norm": 3.2653839732704286, + "learning_rate": 2.3435471326297923e-06, + "loss": 0.0202, + "step": 4342 + }, + { + "epoch": 3.1238985793921956, + "grad_norm": 1.4975803740093616, + "learning_rate": 2.343265766242874e-06, + "loss": 0.0055, + "step": 4343 + }, + { + "epoch": 3.1246178744830067, + "grad_norm": 5.944314430720115, + "learning_rate": 2.342984356468452e-06, + "loss": 0.1112, + "step": 4344 + }, + { + "epoch": 3.1253371695738177, + "grad_norm": 2.115896244822855, + "learning_rate": 2.342702903321005e-06, + "loss": 0.0578, + "step": 4345 + }, + { + "epoch": 3.1260564646646287, + "grad_norm": 2.6737930817274354, + "learning_rate": 2.342421406815014e-06, + "loss": 0.0674, + "step": 4346 + }, + { + "epoch": 3.1267757597554398, + "grad_norm": 5.484989361988779, + "learning_rate": 2.342139866964963e-06, + "loss": 0.1758, + "step": 4347 + }, + { + "epoch": 3.127495054846251, + "grad_norm": 4.471772851325045, + "learning_rate": 2.3418582837853376e-06, + "loss": 0.0179, + "step": 4348 + }, + { + "epoch": 3.128214349937062, + "grad_norm": 1.3428885128036792, + "learning_rate": 2.341576657290626e-06, + "loss": 0.0156, + "step": 4349 + }, + { + "epoch": 3.128933645027873, + "grad_norm": 2.791134095875226, + "learning_rate": 2.3412949874953174e-06, + "loss": 0.0832, + "step": 4350 + }, + { + "epoch": 3.129652940118684, + "grad_norm": 1.542315512294783, + "learning_rate": 2.341013274413905e-06, + "loss": 0.004, + "step": 4351 + }, + { + "epoch": 3.130372235209495, + "grad_norm": 4.910016549068694, + "learning_rate": 2.3407315180608835e-06, + "loss": 0.1427, + "step": 4352 + }, + { + "epoch": 3.1310915303003055, + "grad_norm": 4.93607016673045, + "learning_rate": 2.3404497184507492e-06, + "loss": 0.1069, + "step": 4353 + }, + { + "epoch": 3.1318108253911165, + "grad_norm": 2.9150605732507944, + "learning_rate": 2.340167875598002e-06, + "loss": 0.059, + "step": 4354 + }, + { + "epoch": 3.1325301204819276, + "grad_norm": 3.991647708689319, + "learning_rate": 2.339885989517142e-06, + "loss": 0.2022, + "step": 4355 + }, + { + "epoch": 3.1332494155727386, + "grad_norm": 5.586382933642049, + "learning_rate": 2.3396040602226743e-06, + "loss": 0.1204, + "step": 4356 + }, + { + "epoch": 3.1339687106635497, + "grad_norm": 5.914467360284644, + "learning_rate": 2.3393220877291035e-06, + "loss": 0.2169, + "step": 4357 + }, + { + "epoch": 3.1346880057543607, + "grad_norm": 3.0540154205588688, + "learning_rate": 2.3390400720509377e-06, + "loss": 0.0445, + "step": 4358 + }, + { + "epoch": 3.1354073008451717, + "grad_norm": 1.3041172786915693, + "learning_rate": 2.3387580132026882e-06, + "loss": 0.0183, + "step": 4359 + }, + { + "epoch": 3.1361265959359828, + "grad_norm": 3.2906619112387876, + "learning_rate": 2.3384759111988657e-06, + "loss": 0.0948, + "step": 4360 + }, + { + "epoch": 3.136845891026794, + "grad_norm": 2.3115473671685565, + "learning_rate": 2.338193766053985e-06, + "loss": 0.0688, + "step": 4361 + }, + { + "epoch": 3.137565186117605, + "grad_norm": 2.6231877080535253, + "learning_rate": 2.337911577782565e-06, + "loss": 0.0862, + "step": 4362 + }, + { + "epoch": 3.138284481208416, + "grad_norm": 3.759699379289737, + "learning_rate": 2.337629346399123e-06, + "loss": 0.0673, + "step": 4363 + }, + { + "epoch": 3.139003776299227, + "grad_norm": 2.5690403586303616, + "learning_rate": 2.33734707191818e-06, + "loss": 0.0613, + "step": 4364 + }, + { + "epoch": 3.139723071390038, + "grad_norm": 2.7753052910416987, + "learning_rate": 2.3370647543542612e-06, + "loss": 0.0267, + "step": 4365 + }, + { + "epoch": 3.140442366480849, + "grad_norm": 2.7417495295000798, + "learning_rate": 2.336782393721891e-06, + "loss": 0.0842, + "step": 4366 + }, + { + "epoch": 3.14116166157166, + "grad_norm": 3.954977435689569, + "learning_rate": 2.336499990035597e-06, + "loss": 0.0303, + "step": 4367 + }, + { + "epoch": 3.1418809566624706, + "grad_norm": 2.6735883113450196, + "learning_rate": 2.3362175433099115e-06, + "loss": 0.0181, + "step": 4368 + }, + { + "epoch": 3.1426002517532816, + "grad_norm": 2.783632486656075, + "learning_rate": 2.3359350535593644e-06, + "loss": 0.0243, + "step": 4369 + }, + { + "epoch": 3.1433195468440926, + "grad_norm": 4.187529340339971, + "learning_rate": 2.3356525207984917e-06, + "loss": 0.1168, + "step": 4370 + }, + { + "epoch": 3.1440388419349037, + "grad_norm": 2.488807835596682, + "learning_rate": 2.3353699450418297e-06, + "loss": 0.0391, + "step": 4371 + }, + { + "epoch": 3.1447581370257147, + "grad_norm": 2.275658666436562, + "learning_rate": 2.335087326303918e-06, + "loss": 0.0438, + "step": 4372 + }, + { + "epoch": 3.1454774321165258, + "grad_norm": 4.183793661469565, + "learning_rate": 2.334804664599297e-06, + "loss": 0.1694, + "step": 4373 + }, + { + "epoch": 3.146196727207337, + "grad_norm": 0.6376790921047801, + "learning_rate": 2.3345219599425105e-06, + "loss": 0.0027, + "step": 4374 + }, + { + "epoch": 3.146916022298148, + "grad_norm": 2.3987108888760424, + "learning_rate": 2.3342392123481047e-06, + "loss": 0.0507, + "step": 4375 + }, + { + "epoch": 3.147635317388959, + "grad_norm": 0.17687590556014357, + "learning_rate": 2.3339564218306266e-06, + "loss": 0.0008, + "step": 4376 + }, + { + "epoch": 3.14835461247977, + "grad_norm": 2.364436737289925, + "learning_rate": 2.333673588404627e-06, + "loss": 0.0047, + "step": 4377 + }, + { + "epoch": 3.149073907570581, + "grad_norm": 3.785599503105051, + "learning_rate": 2.333390712084657e-06, + "loss": 0.0402, + "step": 4378 + }, + { + "epoch": 3.149793202661392, + "grad_norm": 4.104086656378021, + "learning_rate": 2.3331077928852723e-06, + "loss": 0.033, + "step": 4379 + }, + { + "epoch": 3.150512497752203, + "grad_norm": 2.2335032453804646, + "learning_rate": 2.3328248308210295e-06, + "loss": 0.0317, + "step": 4380 + }, + { + "epoch": 3.151231792843014, + "grad_norm": 6.85313650262855, + "learning_rate": 2.3325418259064867e-06, + "loss": 0.0987, + "step": 4381 + }, + { + "epoch": 3.151951087933825, + "grad_norm": 4.338090647555055, + "learning_rate": 2.332258778156206e-06, + "loss": 0.0687, + "step": 4382 + }, + { + "epoch": 3.1526703830246356, + "grad_norm": 4.947901042791069, + "learning_rate": 2.3319756875847495e-06, + "loss": 0.223, + "step": 4383 + }, + { + "epoch": 3.1533896781154467, + "grad_norm": 2.094919411961174, + "learning_rate": 2.331692554206684e-06, + "loss": 0.0449, + "step": 4384 + }, + { + "epoch": 3.1541089732062577, + "grad_norm": 5.742208622773294, + "learning_rate": 2.3314093780365763e-06, + "loss": 0.0373, + "step": 4385 + }, + { + "epoch": 3.1548282682970687, + "grad_norm": 3.617450356904854, + "learning_rate": 2.3311261590889966e-06, + "loss": 0.0782, + "step": 4386 + }, + { + "epoch": 3.15554756338788, + "grad_norm": 4.714092859564531, + "learning_rate": 2.3308428973785173e-06, + "loss": 0.1354, + "step": 4387 + }, + { + "epoch": 3.156266858478691, + "grad_norm": 2.043927054606574, + "learning_rate": 2.330559592919712e-06, + "loss": 0.0324, + "step": 4388 + }, + { + "epoch": 3.156986153569502, + "grad_norm": 3.354761684526233, + "learning_rate": 2.330276245727158e-06, + "loss": 0.0628, + "step": 4389 + }, + { + "epoch": 3.157705448660313, + "grad_norm": 2.8181267759493362, + "learning_rate": 2.3299928558154333e-06, + "loss": 0.0306, + "step": 4390 + }, + { + "epoch": 3.158424743751124, + "grad_norm": 1.189782749982172, + "learning_rate": 2.3297094231991195e-06, + "loss": 0.027, + "step": 4391 + }, + { + "epoch": 3.159144038841935, + "grad_norm": 10.402515550770174, + "learning_rate": 2.3294259478927993e-06, + "loss": 0.1305, + "step": 4392 + }, + { + "epoch": 3.159863333932746, + "grad_norm": 5.9075749703713045, + "learning_rate": 2.329142429911058e-06, + "loss": 0.0738, + "step": 4393 + }, + { + "epoch": 3.160582629023557, + "grad_norm": 2.470905455932841, + "learning_rate": 2.3288588692684832e-06, + "loss": 0.0617, + "step": 4394 + }, + { + "epoch": 3.161301924114368, + "grad_norm": 5.257654640936029, + "learning_rate": 2.3285752659796647e-06, + "loss": 0.0687, + "step": 4395 + }, + { + "epoch": 3.162021219205179, + "grad_norm": 1.072471334251246, + "learning_rate": 2.3282916200591944e-06, + "loss": 0.0213, + "step": 4396 + }, + { + "epoch": 3.16274051429599, + "grad_norm": 3.0844900738070473, + "learning_rate": 2.3280079315216656e-06, + "loss": 0.0098, + "step": 4397 + }, + { + "epoch": 3.1634598093868007, + "grad_norm": 3.3694254660500405, + "learning_rate": 2.3277242003816755e-06, + "loss": 0.08, + "step": 4398 + }, + { + "epoch": 3.1641791044776117, + "grad_norm": 1.0663286297831176, + "learning_rate": 2.327440426653823e-06, + "loss": 0.0091, + "step": 4399 + }, + { + "epoch": 3.164898399568423, + "grad_norm": 4.808314710642797, + "learning_rate": 2.3271566103527065e-06, + "loss": 0.1251, + "step": 4400 + }, + { + "epoch": 3.165617694659234, + "grad_norm": 1.6913797189511743, + "learning_rate": 2.3268727514929317e-06, + "loss": 0.0039, + "step": 4401 + }, + { + "epoch": 3.166336989750045, + "grad_norm": 0.5088426229237254, + "learning_rate": 2.326588850089102e-06, + "loss": 0.0056, + "step": 4402 + }, + { + "epoch": 3.167056284840856, + "grad_norm": 1.00465384953645, + "learning_rate": 2.3263049061558243e-06, + "loss": 0.0174, + "step": 4403 + }, + { + "epoch": 3.167775579931667, + "grad_norm": 3.146771588246694, + "learning_rate": 2.326020919707709e-06, + "loss": 0.0736, + "step": 4404 + }, + { + "epoch": 3.168494875022478, + "grad_norm": 6.461581125668157, + "learning_rate": 2.3257368907593674e-06, + "loss": 0.1716, + "step": 4405 + }, + { + "epoch": 3.169214170113289, + "grad_norm": 3.9975187908012844, + "learning_rate": 2.325452819325413e-06, + "loss": 0.0562, + "step": 4406 + }, + { + "epoch": 3.1699334652041, + "grad_norm": 4.842077619584255, + "learning_rate": 2.3251687054204626e-06, + "loss": 0.076, + "step": 4407 + }, + { + "epoch": 3.170652760294911, + "grad_norm": 1.6013897776608923, + "learning_rate": 2.324884549059133e-06, + "loss": 0.0528, + "step": 4408 + }, + { + "epoch": 3.171372055385722, + "grad_norm": 7.85392116682029, + "learning_rate": 2.324600350256046e-06, + "loss": 0.3062, + "step": 4409 + }, + { + "epoch": 3.172091350476533, + "grad_norm": 1.3042384112023762, + "learning_rate": 2.3243161090258233e-06, + "loss": 0.0039, + "step": 4410 + }, + { + "epoch": 3.172810645567344, + "grad_norm": 2.278091582036591, + "learning_rate": 2.3240318253830895e-06, + "loss": 0.0567, + "step": 4411 + }, + { + "epoch": 3.173529940658155, + "grad_norm": 3.211394369583977, + "learning_rate": 2.323747499342472e-06, + "loss": 0.0613, + "step": 4412 + }, + { + "epoch": 3.174249235748966, + "grad_norm": 2.8824392541265307, + "learning_rate": 2.3234631309185993e-06, + "loss": 0.0418, + "step": 4413 + }, + { + "epoch": 3.174968530839777, + "grad_norm": 1.6915780414511505, + "learning_rate": 2.3231787201261037e-06, + "loss": 0.0084, + "step": 4414 + }, + { + "epoch": 3.175687825930588, + "grad_norm": 5.258206607465898, + "learning_rate": 2.322894266979617e-06, + "loss": 0.1322, + "step": 4415 + }, + { + "epoch": 3.176407121021399, + "grad_norm": 3.223558723258728, + "learning_rate": 2.322609771493776e-06, + "loss": 0.09, + "step": 4416 + }, + { + "epoch": 3.17712641611221, + "grad_norm": 2.4456818557466606, + "learning_rate": 2.322325233683218e-06, + "loss": 0.028, + "step": 4417 + }, + { + "epoch": 3.177845711203021, + "grad_norm": 0.5984818678664977, + "learning_rate": 2.322040653562584e-06, + "loss": 0.0027, + "step": 4418 + }, + { + "epoch": 3.178565006293832, + "grad_norm": 3.2196869190183617, + "learning_rate": 2.321756031146515e-06, + "loss": 0.0867, + "step": 4419 + }, + { + "epoch": 3.179284301384643, + "grad_norm": 2.0150142720916913, + "learning_rate": 2.3214713664496553e-06, + "loss": 0.0229, + "step": 4420 + }, + { + "epoch": 3.180003596475454, + "grad_norm": 0.23136971926811617, + "learning_rate": 2.3211866594866517e-06, + "loss": 0.0005, + "step": 4421 + }, + { + "epoch": 3.180722891566265, + "grad_norm": 1.8908780644179548, + "learning_rate": 2.3209019102721534e-06, + "loss": 0.035, + "step": 4422 + }, + { + "epoch": 3.181442186657076, + "grad_norm": 4.551876932182956, + "learning_rate": 2.3206171188208104e-06, + "loss": 0.0171, + "step": 4423 + }, + { + "epoch": 3.182161481747887, + "grad_norm": 0.21177469250078276, + "learning_rate": 2.320332285147276e-06, + "loss": 0.0008, + "step": 4424 + }, + { + "epoch": 3.182880776838698, + "grad_norm": 2.736776600710655, + "learning_rate": 2.3200474092662057e-06, + "loss": 0.0612, + "step": 4425 + }, + { + "epoch": 3.1836000719295092, + "grad_norm": 5.346517619869282, + "learning_rate": 2.3197624911922568e-06, + "loss": 0.0711, + "step": 4426 + }, + { + "epoch": 3.1843193670203203, + "grad_norm": 2.523345662587272, + "learning_rate": 2.3194775309400884e-06, + "loss": 0.0308, + "step": 4427 + }, + { + "epoch": 3.1850386621111313, + "grad_norm": 0.31850409570219446, + "learning_rate": 2.319192528524363e-06, + "loss": 0.0018, + "step": 4428 + }, + { + "epoch": 3.1857579572019423, + "grad_norm": 6.595269192651395, + "learning_rate": 2.318907483959743e-06, + "loss": 0.0984, + "step": 4429 + }, + { + "epoch": 3.186477252292753, + "grad_norm": 4.045564026113133, + "learning_rate": 2.3186223972608963e-06, + "loss": 0.1256, + "step": 4430 + }, + { + "epoch": 3.187196547383564, + "grad_norm": 1.807476724480253, + "learning_rate": 2.3183372684424896e-06, + "loss": 0.0333, + "step": 4431 + }, + { + "epoch": 3.187915842474375, + "grad_norm": 2.752883557467025, + "learning_rate": 2.3180520975191944e-06, + "loss": 0.0585, + "step": 4432 + }, + { + "epoch": 3.188635137565186, + "grad_norm": 3.2346090928380984, + "learning_rate": 2.3177668845056827e-06, + "loss": 0.0783, + "step": 4433 + }, + { + "epoch": 3.189354432655997, + "grad_norm": 3.0956889653819144, + "learning_rate": 2.317481629416629e-06, + "loss": 0.0867, + "step": 4434 + }, + { + "epoch": 3.190073727746808, + "grad_norm": 1.4527593343471032, + "learning_rate": 2.31719633226671e-06, + "loss": 0.0391, + "step": 4435 + }, + { + "epoch": 3.190793022837619, + "grad_norm": 0.06607369291953118, + "learning_rate": 2.316910993070606e-06, + "loss": 0.0003, + "step": 4436 + }, + { + "epoch": 3.19151231792843, + "grad_norm": 0.2958334965677924, + "learning_rate": 2.3166256118429966e-06, + "loss": 0.0023, + "step": 4437 + }, + { + "epoch": 3.192231613019241, + "grad_norm": 2.8391639476371613, + "learning_rate": 2.3163401885985663e-06, + "loss": 0.0784, + "step": 4438 + }, + { + "epoch": 3.192950908110052, + "grad_norm": 3.8482289008831425, + "learning_rate": 2.3160547233520006e-06, + "loss": 0.0589, + "step": 4439 + }, + { + "epoch": 3.1936702032008633, + "grad_norm": 2.9866771657659994, + "learning_rate": 2.315769216117986e-06, + "loss": 0.094, + "step": 4440 + }, + { + "epoch": 3.1943894982916743, + "grad_norm": 1.9922775740843797, + "learning_rate": 2.315483666911214e-06, + "loss": 0.0321, + "step": 4441 + }, + { + "epoch": 3.1951087933824853, + "grad_norm": 3.2256003173956, + "learning_rate": 2.3151980757463756e-06, + "loss": 0.1114, + "step": 4442 + }, + { + "epoch": 3.1958280884732964, + "grad_norm": 1.8775875489575369, + "learning_rate": 2.314912442638165e-06, + "loss": 0.0552, + "step": 4443 + }, + { + "epoch": 3.1965473835641074, + "grad_norm": 4.578461062774031, + "learning_rate": 2.3146267676012787e-06, + "loss": 0.0781, + "step": 4444 + }, + { + "epoch": 3.197266678654918, + "grad_norm": 3.5597560614703454, + "learning_rate": 2.314341050650415e-06, + "loss": 0.1009, + "step": 4445 + }, + { + "epoch": 3.197985973745729, + "grad_norm": 3.475688725970508, + "learning_rate": 2.314055291800275e-06, + "loss": 0.0742, + "step": 4446 + }, + { + "epoch": 3.19870526883654, + "grad_norm": 0.4769903742144726, + "learning_rate": 2.3137694910655614e-06, + "loss": 0.0014, + "step": 4447 + }, + { + "epoch": 3.199424563927351, + "grad_norm": 3.0777961065453936, + "learning_rate": 2.3134836484609786e-06, + "loss": 0.0576, + "step": 4448 + }, + { + "epoch": 3.200143859018162, + "grad_norm": 3.5262820400683257, + "learning_rate": 2.313197764001234e-06, + "loss": 0.0686, + "step": 4449 + }, + { + "epoch": 3.200863154108973, + "grad_norm": 6.804225922747237, + "learning_rate": 2.312911837701037e-06, + "loss": 0.1149, + "step": 4450 + }, + { + "epoch": 3.201582449199784, + "grad_norm": 2.3001825882476306, + "learning_rate": 2.312625869575099e-06, + "loss": 0.0091, + "step": 4451 + }, + { + "epoch": 3.202301744290595, + "grad_norm": 0.7355766387388128, + "learning_rate": 2.312339859638134e-06, + "loss": 0.0028, + "step": 4452 + }, + { + "epoch": 3.2030210393814063, + "grad_norm": 2.914060450606174, + "learning_rate": 2.3120538079048566e-06, + "loss": 0.0847, + "step": 4453 + }, + { + "epoch": 3.2037403344722173, + "grad_norm": 3.1174857545574137, + "learning_rate": 2.3117677143899854e-06, + "loss": 0.1301, + "step": 4454 + }, + { + "epoch": 3.2044596295630283, + "grad_norm": 2.068425161934889, + "learning_rate": 2.31148157910824e-06, + "loss": 0.0559, + "step": 4455 + }, + { + "epoch": 3.2051789246538394, + "grad_norm": 3.6944000626291995, + "learning_rate": 2.3111954020743433e-06, + "loss": 0.015, + "step": 4456 + }, + { + "epoch": 3.2058982197446504, + "grad_norm": 4.724724367187219, + "learning_rate": 2.3109091833030194e-06, + "loss": 0.1656, + "step": 4457 + }, + { + "epoch": 3.2066175148354614, + "grad_norm": 1.4975532713284245, + "learning_rate": 2.310622922808994e-06, + "loss": 0.0175, + "step": 4458 + }, + { + "epoch": 3.2073368099262725, + "grad_norm": 2.565549422071911, + "learning_rate": 2.310336620606996e-06, + "loss": 0.0842, + "step": 4459 + }, + { + "epoch": 3.208056105017083, + "grad_norm": 3.538279043331068, + "learning_rate": 2.3100502767117564e-06, + "loss": 0.0488, + "step": 4460 + }, + { + "epoch": 3.208775400107894, + "grad_norm": 3.7405877425216785, + "learning_rate": 2.3097638911380082e-06, + "loss": 0.087, + "step": 4461 + }, + { + "epoch": 3.209494695198705, + "grad_norm": 4.413247130370564, + "learning_rate": 2.309477463900487e-06, + "loss": 0.094, + "step": 4462 + }, + { + "epoch": 3.210213990289516, + "grad_norm": 3.137204079538125, + "learning_rate": 2.3091909950139285e-06, + "loss": 0.0584, + "step": 4463 + }, + { + "epoch": 3.210933285380327, + "grad_norm": 0.33025369181503383, + "learning_rate": 2.308904484493073e-06, + "loss": 0.0011, + "step": 4464 + }, + { + "epoch": 3.211652580471138, + "grad_norm": 4.331290380990014, + "learning_rate": 2.3086179323526622e-06, + "loss": 0.1194, + "step": 4465 + }, + { + "epoch": 3.2123718755619493, + "grad_norm": 4.480447079152821, + "learning_rate": 2.3083313386074386e-06, + "loss": 0.133, + "step": 4466 + }, + { + "epoch": 3.2130911706527603, + "grad_norm": 2.075713118990339, + "learning_rate": 2.308044703272149e-06, + "loss": 0.049, + "step": 4467 + }, + { + "epoch": 3.2138104657435713, + "grad_norm": 1.3081725493856065, + "learning_rate": 2.307758026361541e-06, + "loss": 0.0268, + "step": 4468 + }, + { + "epoch": 3.2145297608343824, + "grad_norm": 2.798791305170333, + "learning_rate": 2.3074713078903647e-06, + "loss": 0.0724, + "step": 4469 + }, + { + "epoch": 3.2152490559251934, + "grad_norm": 5.410810379485853, + "learning_rate": 2.3071845478733724e-06, + "loss": 0.0911, + "step": 4470 + }, + { + "epoch": 3.2159683510160044, + "grad_norm": 2.0910150816236084, + "learning_rate": 2.306897746325317e-06, + "loss": 0.0423, + "step": 4471 + }, + { + "epoch": 3.2166876461068155, + "grad_norm": 0.6258747737965367, + "learning_rate": 2.306610903260957e-06, + "loss": 0.0014, + "step": 4472 + }, + { + "epoch": 3.2174069411976265, + "grad_norm": 0.7972727761803204, + "learning_rate": 2.3063240186950498e-06, + "loss": 0.0023, + "step": 4473 + }, + { + "epoch": 3.2181262362884375, + "grad_norm": 3.3915993658670294, + "learning_rate": 2.3060370926423562e-06, + "loss": 0.0958, + "step": 4474 + }, + { + "epoch": 3.218845531379248, + "grad_norm": 3.94959226914782, + "learning_rate": 2.3057501251176396e-06, + "loss": 0.0748, + "step": 4475 + }, + { + "epoch": 3.219564826470059, + "grad_norm": 3.883366501246911, + "learning_rate": 2.3054631161356644e-06, + "loss": 0.0979, + "step": 4476 + }, + { + "epoch": 3.22028412156087, + "grad_norm": 3.1993830122167055, + "learning_rate": 2.305176065711198e-06, + "loss": 0.0478, + "step": 4477 + }, + { + "epoch": 3.221003416651681, + "grad_norm": 2.1813816984212995, + "learning_rate": 2.30488897385901e-06, + "loss": 0.0077, + "step": 4478 + }, + { + "epoch": 3.2217227117424923, + "grad_norm": 0.3713677942767126, + "learning_rate": 2.3046018405938703e-06, + "loss": 0.0015, + "step": 4479 + }, + { + "epoch": 3.2224420068333033, + "grad_norm": 3.771772288692133, + "learning_rate": 2.304314665930554e-06, + "loss": 0.0433, + "step": 4480 + }, + { + "epoch": 3.2231613019241143, + "grad_norm": 5.365918324831739, + "learning_rate": 2.3040274498838368e-06, + "loss": 0.1204, + "step": 4481 + }, + { + "epoch": 3.2238805970149254, + "grad_norm": 0.03214763778544739, + "learning_rate": 2.303740192468495e-06, + "loss": 0.0002, + "step": 4482 + }, + { + "epoch": 3.2245998921057364, + "grad_norm": 4.317335105175406, + "learning_rate": 2.3034528936993097e-06, + "loss": 0.0212, + "step": 4483 + }, + { + "epoch": 3.2253191871965474, + "grad_norm": 4.584782371559871, + "learning_rate": 2.303165553591063e-06, + "loss": 0.0604, + "step": 4484 + }, + { + "epoch": 3.2260384822873585, + "grad_norm": 4.475533312134028, + "learning_rate": 2.302878172158538e-06, + "loss": 0.1119, + "step": 4485 + }, + { + "epoch": 3.2267577773781695, + "grad_norm": 2.838723315491718, + "learning_rate": 2.302590749416522e-06, + "loss": 0.0079, + "step": 4486 + }, + { + "epoch": 3.2274770724689805, + "grad_norm": 3.0813757998138915, + "learning_rate": 2.3023032853798027e-06, + "loss": 0.0764, + "step": 4487 + }, + { + "epoch": 3.2281963675597916, + "grad_norm": 3.1761811516910146, + "learning_rate": 2.3020157800631715e-06, + "loss": 0.0858, + "step": 4488 + }, + { + "epoch": 3.2289156626506026, + "grad_norm": 4.015051287524155, + "learning_rate": 2.30172823348142e-06, + "loss": 0.0715, + "step": 4489 + }, + { + "epoch": 3.229634957741413, + "grad_norm": 1.6266441911200107, + "learning_rate": 2.301440645649344e-06, + "loss": 0.0056, + "step": 4490 + }, + { + "epoch": 3.230354252832224, + "grad_norm": 5.880698040183779, + "learning_rate": 2.3011530165817396e-06, + "loss": 0.1065, + "step": 4491 + }, + { + "epoch": 3.2310735479230352, + "grad_norm": 3.950368788874015, + "learning_rate": 2.300865346293406e-06, + "loss": 0.0552, + "step": 4492 + }, + { + "epoch": 3.2317928430138463, + "grad_norm": 2.0590123928379023, + "learning_rate": 2.3005776347991447e-06, + "loss": 0.043, + "step": 4493 + }, + { + "epoch": 3.2325121381046573, + "grad_norm": 1.58608227069417, + "learning_rate": 2.300289882113759e-06, + "loss": 0.0394, + "step": 4494 + }, + { + "epoch": 3.2332314331954684, + "grad_norm": 7.333014031893844, + "learning_rate": 2.3000020882520532e-06, + "loss": 0.13, + "step": 4495 + }, + { + "epoch": 3.2339507282862794, + "grad_norm": 5.6578992083860635, + "learning_rate": 2.2997142532288362e-06, + "loss": 0.159, + "step": 4496 + }, + { + "epoch": 3.2346700233770904, + "grad_norm": 3.0283983287122624, + "learning_rate": 2.2994263770589165e-06, + "loss": 0.1066, + "step": 4497 + }, + { + "epoch": 3.2353893184679015, + "grad_norm": 1.9122741314876381, + "learning_rate": 2.299138459757107e-06, + "loss": 0.0181, + "step": 4498 + }, + { + "epoch": 3.2361086135587125, + "grad_norm": 2.1077813933946623, + "learning_rate": 2.298850501338221e-06, + "loss": 0.0619, + "step": 4499 + }, + { + "epoch": 3.2368279086495235, + "grad_norm": 2.8305511425373227, + "learning_rate": 2.298562501817074e-06, + "loss": 0.0378, + "step": 4500 + }, + { + "epoch": 3.2375472037403346, + "grad_norm": 1.6833399133911264, + "learning_rate": 2.2982744612084845e-06, + "loss": 0.0031, + "step": 4501 + }, + { + "epoch": 3.2382664988311456, + "grad_norm": 1.5574718042363924, + "learning_rate": 2.297986379527272e-06, + "loss": 0.031, + "step": 4502 + }, + { + "epoch": 3.2389857939219566, + "grad_norm": 5.587475936995155, + "learning_rate": 2.2976982567882604e-06, + "loss": 0.0804, + "step": 4503 + }, + { + "epoch": 3.2397050890127677, + "grad_norm": 3.9502177006772508, + "learning_rate": 2.297410093006273e-06, + "loss": 0.0142, + "step": 4504 + }, + { + "epoch": 3.2404243841035787, + "grad_norm": 3.10472429272831, + "learning_rate": 2.297121888196136e-06, + "loss": 0.0842, + "step": 4505 + }, + { + "epoch": 3.2411436791943897, + "grad_norm": 4.399228698845738, + "learning_rate": 2.2968336423726793e-06, + "loss": 0.1163, + "step": 4506 + }, + { + "epoch": 3.2418629742852003, + "grad_norm": 3.613437795295939, + "learning_rate": 2.296545355550732e-06, + "loss": 0.0516, + "step": 4507 + }, + { + "epoch": 3.2425822693760114, + "grad_norm": 1.3604763218491365, + "learning_rate": 2.2962570277451285e-06, + "loss": 0.0032, + "step": 4508 + }, + { + "epoch": 3.2433015644668224, + "grad_norm": 5.82123442505743, + "learning_rate": 2.2959686589707033e-06, + "loss": 0.2941, + "step": 4509 + }, + { + "epoch": 3.2440208595576334, + "grad_norm": 1.240475221497884, + "learning_rate": 2.2956802492422922e-06, + "loss": 0.0056, + "step": 4510 + }, + { + "epoch": 3.2447401546484445, + "grad_norm": 2.482300430094419, + "learning_rate": 2.2953917985747366e-06, + "loss": 0.0809, + "step": 4511 + }, + { + "epoch": 3.2454594497392555, + "grad_norm": 1.1974042821033857, + "learning_rate": 2.2951033069828765e-06, + "loss": 0.0043, + "step": 4512 + }, + { + "epoch": 3.2461787448300665, + "grad_norm": 4.178142093388856, + "learning_rate": 2.2948147744815552e-06, + "loss": 0.0921, + "step": 4513 + }, + { + "epoch": 3.2468980399208776, + "grad_norm": 1.6573401438851636, + "learning_rate": 2.294526201085619e-06, + "loss": 0.0244, + "step": 4514 + }, + { + "epoch": 3.2476173350116886, + "grad_norm": 1.6868882414163784, + "learning_rate": 2.294237586809915e-06, + "loss": 0.0226, + "step": 4515 + }, + { + "epoch": 3.2483366301024996, + "grad_norm": 2.5941880716242203, + "learning_rate": 2.2939489316692923e-06, + "loss": 0.0872, + "step": 4516 + }, + { + "epoch": 3.2490559251933107, + "grad_norm": 2.3213252904537636, + "learning_rate": 2.2936602356786034e-06, + "loss": 0.0516, + "step": 4517 + }, + { + "epoch": 3.2497752202841217, + "grad_norm": 1.7414896510303517, + "learning_rate": 2.2933714988527022e-06, + "loss": 0.0349, + "step": 4518 + }, + { + "epoch": 3.2504945153749327, + "grad_norm": 0.5481272617545307, + "learning_rate": 2.293082721206445e-06, + "loss": 0.0012, + "step": 4519 + }, + { + "epoch": 3.2512138104657438, + "grad_norm": 2.2856843161152884, + "learning_rate": 2.2927939027546894e-06, + "loss": 0.063, + "step": 4520 + }, + { + "epoch": 3.251933105556555, + "grad_norm": 3.8500958724591814, + "learning_rate": 2.292505043512296e-06, + "loss": 0.0929, + "step": 4521 + }, + { + "epoch": 3.2526524006473654, + "grad_norm": 4.01258132567913, + "learning_rate": 2.292216143494127e-06, + "loss": 0.0796, + "step": 4522 + }, + { + "epoch": 3.2533716957381764, + "grad_norm": 1.9024944093843816, + "learning_rate": 2.291927202715046e-06, + "loss": 0.038, + "step": 4523 + }, + { + "epoch": 3.2540909908289875, + "grad_norm": 4.1204685252936795, + "learning_rate": 2.2916382211899207e-06, + "loss": 0.172, + "step": 4524 + }, + { + "epoch": 3.2548102859197985, + "grad_norm": 3.9832920083420627, + "learning_rate": 2.291349198933619e-06, + "loss": 0.0191, + "step": 4525 + }, + { + "epoch": 3.2555295810106095, + "grad_norm": 4.723811135501269, + "learning_rate": 2.2910601359610125e-06, + "loss": 0.1843, + "step": 4526 + }, + { + "epoch": 3.2562488761014206, + "grad_norm": 3.1517702708389836, + "learning_rate": 2.2907710322869728e-06, + "loss": 0.0552, + "step": 4527 + }, + { + "epoch": 3.2569681711922316, + "grad_norm": 3.010441660491754, + "learning_rate": 2.290481887926376e-06, + "loss": 0.049, + "step": 4528 + }, + { + "epoch": 3.2576874662830426, + "grad_norm": 4.882684002253346, + "learning_rate": 2.2901927028940976e-06, + "loss": 0.0976, + "step": 4529 + }, + { + "epoch": 3.2584067613738537, + "grad_norm": 3.0209006867492363, + "learning_rate": 2.2899034772050177e-06, + "loss": 0.031, + "step": 4530 + }, + { + "epoch": 3.2591260564646647, + "grad_norm": 4.482030342568125, + "learning_rate": 2.289614210874017e-06, + "loss": 0.0085, + "step": 4531 + }, + { + "epoch": 3.2598453515554757, + "grad_norm": 4.693993702004803, + "learning_rate": 2.28932490391598e-06, + "loss": 0.1204, + "step": 4532 + }, + { + "epoch": 3.2605646466462868, + "grad_norm": 4.065221188719599, + "learning_rate": 2.28903555634579e-06, + "loss": 0.0289, + "step": 4533 + }, + { + "epoch": 3.261283941737098, + "grad_norm": 2.6341416286948567, + "learning_rate": 2.288746168178336e-06, + "loss": 0.0955, + "step": 4534 + }, + { + "epoch": 3.262003236827909, + "grad_norm": 3.239883654137051, + "learning_rate": 2.2884567394285073e-06, + "loss": 0.113, + "step": 4535 + }, + { + "epoch": 3.26272253191872, + "grad_norm": 4.264913439965863, + "learning_rate": 2.288167270111195e-06, + "loss": 0.0286, + "step": 4536 + }, + { + "epoch": 3.2634418270095304, + "grad_norm": 4.938021074297976, + "learning_rate": 2.287877760241293e-06, + "loss": 0.1536, + "step": 4537 + }, + { + "epoch": 3.2641611221003415, + "grad_norm": 10.212806225955248, + "learning_rate": 2.2875882098336975e-06, + "loss": 0.1728, + "step": 4538 + }, + { + "epoch": 3.2648804171911525, + "grad_norm": 3.230838492279424, + "learning_rate": 2.287298618903305e-06, + "loss": 0.0409, + "step": 4539 + }, + { + "epoch": 3.2655997122819636, + "grad_norm": 3.0017179474407647, + "learning_rate": 2.2870089874650177e-06, + "loss": 0.0897, + "step": 4540 + }, + { + "epoch": 3.2663190073727746, + "grad_norm": 1.9895188313574754, + "learning_rate": 2.2867193155337357e-06, + "loss": 0.0096, + "step": 4541 + }, + { + "epoch": 3.2670383024635856, + "grad_norm": 0.03845037147874532, + "learning_rate": 2.2864296031243645e-06, + "loss": 0.0001, + "step": 4542 + }, + { + "epoch": 3.2677575975543967, + "grad_norm": 0.8479546687788646, + "learning_rate": 2.2861398502518093e-06, + "loss": 0.01, + "step": 4543 + }, + { + "epoch": 3.2684768926452077, + "grad_norm": 6.196471453208916, + "learning_rate": 2.2858500569309787e-06, + "loss": 0.1443, + "step": 4544 + }, + { + "epoch": 3.2691961877360187, + "grad_norm": 0.6438104350712269, + "learning_rate": 2.2855602231767834e-06, + "loss": 0.0058, + "step": 4545 + }, + { + "epoch": 3.2699154828268298, + "grad_norm": 0.12485292211004564, + "learning_rate": 2.2852703490041355e-06, + "loss": 0.0002, + "step": 4546 + }, + { + "epoch": 3.270634777917641, + "grad_norm": 7.715480287665266, + "learning_rate": 2.2849804344279495e-06, + "loss": 0.3101, + "step": 4547 + }, + { + "epoch": 3.271354073008452, + "grad_norm": 2.899992539018861, + "learning_rate": 2.284690479463142e-06, + "loss": 0.0613, + "step": 4548 + }, + { + "epoch": 3.272073368099263, + "grad_norm": 1.10308044019799, + "learning_rate": 2.2844004841246326e-06, + "loss": 0.0028, + "step": 4549 + }, + { + "epoch": 3.272792663190074, + "grad_norm": 4.677775241500674, + "learning_rate": 2.284110448427341e-06, + "loss": 0.1668, + "step": 4550 + }, + { + "epoch": 3.273511958280885, + "grad_norm": 2.1997057213962234, + "learning_rate": 2.2838203723861905e-06, + "loss": 0.0633, + "step": 4551 + }, + { + "epoch": 3.2742312533716955, + "grad_norm": 2.551868283747727, + "learning_rate": 2.2835302560161054e-06, + "loss": 0.11, + "step": 4552 + }, + { + "epoch": 3.2749505484625065, + "grad_norm": 1.3950783421460902, + "learning_rate": 2.283240099332013e-06, + "loss": 0.0227, + "step": 4553 + }, + { + "epoch": 3.2756698435533176, + "grad_norm": 1.960847236190856, + "learning_rate": 2.282949902348843e-06, + "loss": 0.0313, + "step": 4554 + }, + { + "epoch": 3.2763891386441286, + "grad_norm": 1.6767291378098843, + "learning_rate": 2.282659665081526e-06, + "loss": 0.0573, + "step": 4555 + }, + { + "epoch": 3.2771084337349397, + "grad_norm": 1.7117983309855933, + "learning_rate": 2.2823693875449954e-06, + "loss": 0.0399, + "step": 4556 + }, + { + "epoch": 3.2778277288257507, + "grad_norm": 0.049293936119734004, + "learning_rate": 2.2820790697541866e-06, + "loss": 0.0002, + "step": 4557 + }, + { + "epoch": 3.2785470239165617, + "grad_norm": 3.528735185834412, + "learning_rate": 2.2817887117240367e-06, + "loss": 0.0749, + "step": 4558 + }, + { + "epoch": 3.2792663190073728, + "grad_norm": 2.3572186886854807, + "learning_rate": 2.281498313469485e-06, + "loss": 0.0973, + "step": 4559 + }, + { + "epoch": 3.279985614098184, + "grad_norm": 3.158029135814789, + "learning_rate": 2.281207875005473e-06, + "loss": 0.0851, + "step": 4560 + }, + { + "epoch": 3.280704909188995, + "grad_norm": 0.7366220091713963, + "learning_rate": 2.280917396346945e-06, + "loss": 0.0016, + "step": 4561 + }, + { + "epoch": 3.281424204279806, + "grad_norm": 7.647067954876685, + "learning_rate": 2.280626877508846e-06, + "loss": 0.1104, + "step": 4562 + }, + { + "epoch": 3.282143499370617, + "grad_norm": 1.4819066623474257, + "learning_rate": 2.280336318506123e-06, + "loss": 0.03, + "step": 4563 + }, + { + "epoch": 3.282862794461428, + "grad_norm": 6.302592744530469, + "learning_rate": 2.2800457193537274e-06, + "loss": 0.1614, + "step": 4564 + }, + { + "epoch": 3.283582089552239, + "grad_norm": 1.9464195660864163, + "learning_rate": 2.2797550800666097e-06, + "loss": 0.028, + "step": 4565 + }, + { + "epoch": 3.28430138464305, + "grad_norm": 2.3357279928713117, + "learning_rate": 2.279464400659725e-06, + "loss": 0.0788, + "step": 4566 + }, + { + "epoch": 3.2850206797338606, + "grad_norm": 1.4252714385344964, + "learning_rate": 2.279173681148028e-06, + "loss": 0.0279, + "step": 4567 + }, + { + "epoch": 3.285739974824672, + "grad_norm": 9.437745735072891, + "learning_rate": 2.2788829215464774e-06, + "loss": 0.0497, + "step": 4568 + }, + { + "epoch": 3.2864592699154827, + "grad_norm": 2.070819588304671, + "learning_rate": 2.2785921218700335e-06, + "loss": 0.0281, + "step": 4569 + }, + { + "epoch": 3.2871785650062937, + "grad_norm": 3.296193370164543, + "learning_rate": 2.278301282133658e-06, + "loss": 0.1054, + "step": 4570 + }, + { + "epoch": 3.2878978600971047, + "grad_norm": 2.2553582316870675, + "learning_rate": 2.278010402352315e-06, + "loss": 0.0572, + "step": 4571 + }, + { + "epoch": 3.2886171551879158, + "grad_norm": 1.6190806394079351, + "learning_rate": 2.2777194825409715e-06, + "loss": 0.0277, + "step": 4572 + }, + { + "epoch": 3.289336450278727, + "grad_norm": 4.375624257877794, + "learning_rate": 2.277428522714595e-06, + "loss": 0.1327, + "step": 4573 + }, + { + "epoch": 3.290055745369538, + "grad_norm": 4.60772729249138, + "learning_rate": 2.2771375228881565e-06, + "loss": 0.0091, + "step": 4574 + }, + { + "epoch": 3.290775040460349, + "grad_norm": 3.808832312377162, + "learning_rate": 2.2768464830766282e-06, + "loss": 0.0983, + "step": 4575 + }, + { + "epoch": 3.29149433555116, + "grad_norm": 3.9519047824282008, + "learning_rate": 2.2765554032949844e-06, + "loss": 0.0751, + "step": 4576 + }, + { + "epoch": 3.292213630641971, + "grad_norm": 5.251271528843381, + "learning_rate": 2.276264283558202e-06, + "loss": 0.1656, + "step": 4577 + }, + { + "epoch": 3.292932925732782, + "grad_norm": 1.6159554904377285, + "learning_rate": 2.27597312388126e-06, + "loss": 0.0363, + "step": 4578 + }, + { + "epoch": 3.293652220823593, + "grad_norm": 5.314410382343385, + "learning_rate": 2.275681924279138e-06, + "loss": 0.1606, + "step": 4579 + }, + { + "epoch": 3.294371515914404, + "grad_norm": 3.712655228468354, + "learning_rate": 2.2753906847668197e-06, + "loss": 0.0505, + "step": 4580 + }, + { + "epoch": 3.295090811005215, + "grad_norm": 1.4819984412072205, + "learning_rate": 2.2750994053592892e-06, + "loss": 0.0262, + "step": 4581 + }, + { + "epoch": 3.2958101060960256, + "grad_norm": 1.5768488565343495, + "learning_rate": 2.2748080860715338e-06, + "loss": 0.02, + "step": 4582 + }, + { + "epoch": 3.296529401186837, + "grad_norm": 4.143716206358008, + "learning_rate": 2.2745167269185423e-06, + "loss": 0.0174, + "step": 4583 + }, + { + "epoch": 3.2972486962776477, + "grad_norm": 3.8417429639463854, + "learning_rate": 2.2742253279153053e-06, + "loss": 0.1345, + "step": 4584 + }, + { + "epoch": 3.2979679913684588, + "grad_norm": 3.9513148302901424, + "learning_rate": 2.273933889076816e-06, + "loss": 0.1042, + "step": 4585 + }, + { + "epoch": 3.29868728645927, + "grad_norm": 0.3357264799130588, + "learning_rate": 2.27364241041807e-06, + "loss": 0.0013, + "step": 4586 + }, + { + "epoch": 3.299406581550081, + "grad_norm": 2.5913857987668356, + "learning_rate": 2.273350891954064e-06, + "loss": 0.0202, + "step": 4587 + }, + { + "epoch": 3.300125876640892, + "grad_norm": 3.5526145432577168, + "learning_rate": 2.273059333699797e-06, + "loss": 0.0914, + "step": 4588 + }, + { + "epoch": 3.300845171731703, + "grad_norm": 4.027765050839108, + "learning_rate": 2.2727677356702695e-06, + "loss": 0.1174, + "step": 4589 + }, + { + "epoch": 3.301564466822514, + "grad_norm": 2.6971976355039846, + "learning_rate": 2.272476097880486e-06, + "loss": 0.0114, + "step": 4590 + }, + { + "epoch": 3.302283761913325, + "grad_norm": 3.7877920073943425, + "learning_rate": 2.2721844203454512e-06, + "loss": 0.1523, + "step": 4591 + }, + { + "epoch": 3.303003057004136, + "grad_norm": 1.4624130923754577, + "learning_rate": 2.2718927030801718e-06, + "loss": 0.0375, + "step": 4592 + }, + { + "epoch": 3.303722352094947, + "grad_norm": 2.601114071889803, + "learning_rate": 2.2716009460996587e-06, + "loss": 0.0113, + "step": 4593 + }, + { + "epoch": 3.304441647185758, + "grad_norm": 1.658452112315142, + "learning_rate": 2.2713091494189224e-06, + "loss": 0.0442, + "step": 4594 + }, + { + "epoch": 3.305160942276569, + "grad_norm": 4.687579377825402, + "learning_rate": 2.271017313052976e-06, + "loss": 0.1974, + "step": 4595 + }, + { + "epoch": 3.30588023736738, + "grad_norm": 8.066413590243029, + "learning_rate": 2.270725437016835e-06, + "loss": 0.2049, + "step": 4596 + }, + { + "epoch": 3.306599532458191, + "grad_norm": 1.0743605143054202, + "learning_rate": 2.270433521325518e-06, + "loss": 0.0247, + "step": 4597 + }, + { + "epoch": 3.307318827549002, + "grad_norm": 7.126794679566996, + "learning_rate": 2.270141565994044e-06, + "loss": 0.1819, + "step": 4598 + }, + { + "epoch": 3.308038122639813, + "grad_norm": 1.4065200007617749, + "learning_rate": 2.2698495710374336e-06, + "loss": 0.0335, + "step": 4599 + }, + { + "epoch": 3.308757417730624, + "grad_norm": 5.219751576680809, + "learning_rate": 2.2695575364707123e-06, + "loss": 0.1905, + "step": 4600 + }, + { + "epoch": 3.309476712821435, + "grad_norm": 2.011498576996309, + "learning_rate": 2.2692654623089043e-06, + "loss": 0.0451, + "step": 4601 + }, + { + "epoch": 3.310196007912246, + "grad_norm": 5.471526910226561, + "learning_rate": 2.2689733485670374e-06, + "loss": 0.1386, + "step": 4602 + }, + { + "epoch": 3.310915303003057, + "grad_norm": 2.6324609005193302, + "learning_rate": 2.268681195260143e-06, + "loss": 0.026, + "step": 4603 + }, + { + "epoch": 3.311634598093868, + "grad_norm": 2.8379221293866235, + "learning_rate": 2.2683890024032507e-06, + "loss": 0.0397, + "step": 4604 + }, + { + "epoch": 3.312353893184679, + "grad_norm": 2.401384847474412, + "learning_rate": 2.2680967700113954e-06, + "loss": 0.0414, + "step": 4605 + }, + { + "epoch": 3.31307318827549, + "grad_norm": 4.437594829064407, + "learning_rate": 2.2678044980996138e-06, + "loss": 0.108, + "step": 4606 + }, + { + "epoch": 3.313792483366301, + "grad_norm": 2.7898312208528577, + "learning_rate": 2.267512186682942e-06, + "loss": 0.0615, + "step": 4607 + }, + { + "epoch": 3.314511778457112, + "grad_norm": 3.1202331093819042, + "learning_rate": 2.2672198357764215e-06, + "loss": 0.1272, + "step": 4608 + }, + { + "epoch": 3.315231073547923, + "grad_norm": 4.005148568976252, + "learning_rate": 2.2669274453950936e-06, + "loss": 0.0549, + "step": 4609 + }, + { + "epoch": 3.315950368638734, + "grad_norm": 3.3979289183777737, + "learning_rate": 2.266635015554002e-06, + "loss": 0.0478, + "step": 4610 + }, + { + "epoch": 3.316669663729545, + "grad_norm": 4.255760231466779, + "learning_rate": 2.2663425462681937e-06, + "loss": 0.0402, + "step": 4611 + }, + { + "epoch": 3.3173889588203562, + "grad_norm": 0.14566386006638457, + "learning_rate": 2.2660500375527154e-06, + "loss": 0.0004, + "step": 4612 + }, + { + "epoch": 3.3181082539111673, + "grad_norm": 3.377909714033796, + "learning_rate": 2.2657574894226176e-06, + "loss": 0.0246, + "step": 4613 + }, + { + "epoch": 3.318827549001978, + "grad_norm": 2.7798196755299425, + "learning_rate": 2.2654649018929538e-06, + "loss": 0.1077, + "step": 4614 + }, + { + "epoch": 3.319546844092789, + "grad_norm": 3.4333059745254584, + "learning_rate": 2.265172274978776e-06, + "loss": 0.1422, + "step": 4615 + }, + { + "epoch": 3.3202661391836, + "grad_norm": 4.149765173269704, + "learning_rate": 2.2648796086951424e-06, + "loss": 0.0956, + "step": 4616 + }, + { + "epoch": 3.320985434274411, + "grad_norm": 3.4253001642208973, + "learning_rate": 2.26458690305711e-06, + "loss": 0.1475, + "step": 4617 + }, + { + "epoch": 3.321704729365222, + "grad_norm": 2.411090217356525, + "learning_rate": 2.264294158079739e-06, + "loss": 0.0919, + "step": 4618 + }, + { + "epoch": 3.322424024456033, + "grad_norm": 1.716393759176561, + "learning_rate": 2.2640013737780917e-06, + "loss": 0.0279, + "step": 4619 + }, + { + "epoch": 3.323143319546844, + "grad_norm": 3.5579797535563813, + "learning_rate": 2.2637085501672326e-06, + "loss": 0.08, + "step": 4620 + }, + { + "epoch": 3.323862614637655, + "grad_norm": 5.852410579839579, + "learning_rate": 2.2634156872622283e-06, + "loss": 0.0247, + "step": 4621 + }, + { + "epoch": 3.324581909728466, + "grad_norm": 3.3410967195786423, + "learning_rate": 2.2631227850781464e-06, + "loss": 0.1077, + "step": 4622 + }, + { + "epoch": 3.325301204819277, + "grad_norm": 3.4875794784119467, + "learning_rate": 2.262829843630058e-06, + "loss": 0.0496, + "step": 4623 + }, + { + "epoch": 3.326020499910088, + "grad_norm": 2.9862350985219055, + "learning_rate": 2.2625368629330348e-06, + "loss": 0.112, + "step": 4624 + }, + { + "epoch": 3.3267397950008992, + "grad_norm": 3.5766844424611715, + "learning_rate": 2.2622438430021515e-06, + "loss": 0.0549, + "step": 4625 + }, + { + "epoch": 3.3274590900917103, + "grad_norm": 4.145329247767296, + "learning_rate": 2.2619507838524844e-06, + "loss": 0.1037, + "step": 4626 + }, + { + "epoch": 3.3281783851825213, + "grad_norm": 2.374786616840892, + "learning_rate": 2.2616576854991116e-06, + "loss": 0.0979, + "step": 4627 + }, + { + "epoch": 3.3288976802733323, + "grad_norm": 3.43815562239229, + "learning_rate": 2.261364547957114e-06, + "loss": 0.0667, + "step": 4628 + }, + { + "epoch": 3.329616975364143, + "grad_norm": 3.5825413319629664, + "learning_rate": 2.261071371241574e-06, + "loss": 0.0599, + "step": 4629 + }, + { + "epoch": 3.330336270454954, + "grad_norm": 0.9785888096761388, + "learning_rate": 2.2607781553675764e-06, + "loss": 0.003, + "step": 4630 + }, + { + "epoch": 3.331055565545765, + "grad_norm": 2.7211478599190375, + "learning_rate": 2.2604849003502063e-06, + "loss": 0.0665, + "step": 4631 + }, + { + "epoch": 3.331774860636576, + "grad_norm": 2.5696218567258176, + "learning_rate": 2.2601916062045536e-06, + "loss": 0.078, + "step": 4632 + }, + { + "epoch": 3.332494155727387, + "grad_norm": 1.0607272301614792, + "learning_rate": 2.259898272945708e-06, + "loss": 0.0282, + "step": 4633 + }, + { + "epoch": 3.333213450818198, + "grad_norm": 2.1483468583017515, + "learning_rate": 2.2596049005887624e-06, + "loss": 0.0304, + "step": 4634 + }, + { + "epoch": 3.333932745909009, + "grad_norm": 2.7495025918075258, + "learning_rate": 2.259311489148811e-06, + "loss": 0.0699, + "step": 4635 + }, + { + "epoch": 3.33465204099982, + "grad_norm": 6.8697577740125, + "learning_rate": 2.2590180386409514e-06, + "loss": 0.1266, + "step": 4636 + }, + { + "epoch": 3.335371336090631, + "grad_norm": 2.451761324228775, + "learning_rate": 2.258724549080281e-06, + "loss": 0.0632, + "step": 4637 + }, + { + "epoch": 3.3360906311814422, + "grad_norm": 0.7727714464516373, + "learning_rate": 2.2584310204819005e-06, + "loss": 0.0011, + "step": 4638 + }, + { + "epoch": 3.3368099262722533, + "grad_norm": 1.5251708462637936, + "learning_rate": 2.258137452860912e-06, + "loss": 0.0058, + "step": 4639 + }, + { + "epoch": 3.3375292213630643, + "grad_norm": 2.038861291381106, + "learning_rate": 2.2578438462324214e-06, + "loss": 0.0137, + "step": 4640 + }, + { + "epoch": 3.3382485164538753, + "grad_norm": 4.041237973999525, + "learning_rate": 2.257550200611534e-06, + "loss": 0.0567, + "step": 4641 + }, + { + "epoch": 3.3389678115446864, + "grad_norm": 3.3431803805449647, + "learning_rate": 2.257256516013359e-06, + "loss": 0.0643, + "step": 4642 + }, + { + "epoch": 3.3396871066354974, + "grad_norm": 4.086610789954164, + "learning_rate": 2.256962792453008e-06, + "loss": 0.0952, + "step": 4643 + }, + { + "epoch": 3.340406401726308, + "grad_norm": 2.2747009156079625, + "learning_rate": 2.2566690299455917e-06, + "loss": 0.0477, + "step": 4644 + }, + { + "epoch": 3.3411256968171195, + "grad_norm": 1.7163701090440489, + "learning_rate": 2.256375228506225e-06, + "loss": 0.0114, + "step": 4645 + }, + { + "epoch": 3.34184499190793, + "grad_norm": 2.4754866696384252, + "learning_rate": 2.256081388150026e-06, + "loss": 0.0439, + "step": 4646 + }, + { + "epoch": 3.342564286998741, + "grad_norm": 0.03531110884529071, + "learning_rate": 2.2557875088921117e-06, + "loss": 0.0002, + "step": 4647 + }, + { + "epoch": 3.343283582089552, + "grad_norm": 2.2581208739683722, + "learning_rate": 2.2554935907476036e-06, + "loss": 0.0259, + "step": 4648 + }, + { + "epoch": 3.344002877180363, + "grad_norm": 3.503500988454564, + "learning_rate": 2.2551996337316237e-06, + "loss": 0.0903, + "step": 4649 + }, + { + "epoch": 3.344722172271174, + "grad_norm": 3.9290873109363837, + "learning_rate": 2.2549056378592974e-06, + "loss": 0.1438, + "step": 4650 + }, + { + "epoch": 3.345441467361985, + "grad_norm": 1.6002283641932478, + "learning_rate": 2.2546116031457505e-06, + "loss": 0.0326, + "step": 4651 + }, + { + "epoch": 3.3461607624527963, + "grad_norm": 3.5557990694248285, + "learning_rate": 2.2543175296061123e-06, + "loss": 0.0906, + "step": 4652 + }, + { + "epoch": 3.3468800575436073, + "grad_norm": 6.145418868372815, + "learning_rate": 2.2540234172555126e-06, + "loss": 0.2325, + "step": 4653 + }, + { + "epoch": 3.3475993526344183, + "grad_norm": 3.263152845867906, + "learning_rate": 2.2537292661090845e-06, + "loss": 0.0913, + "step": 4654 + }, + { + "epoch": 3.3483186477252294, + "grad_norm": 1.6448441310083617, + "learning_rate": 2.2534350761819627e-06, + "loss": 0.0277, + "step": 4655 + }, + { + "epoch": 3.3490379428160404, + "grad_norm": 0.8600227807255956, + "learning_rate": 2.2531408474892836e-06, + "loss": 0.0081, + "step": 4656 + }, + { + "epoch": 3.3497572379068514, + "grad_norm": 4.091182238194695, + "learning_rate": 2.2528465800461856e-06, + "loss": 0.0444, + "step": 4657 + }, + { + "epoch": 3.3504765329976625, + "grad_norm": 4.467972655570028, + "learning_rate": 2.2525522738678096e-06, + "loss": 0.1393, + "step": 4658 + }, + { + "epoch": 3.351195828088473, + "grad_norm": 2.885702680797472, + "learning_rate": 2.2522579289692986e-06, + "loss": 0.0129, + "step": 4659 + }, + { + "epoch": 3.3519151231792845, + "grad_norm": 4.4838451862595905, + "learning_rate": 2.251963545365796e-06, + "loss": 0.0588, + "step": 4660 + }, + { + "epoch": 3.352634418270095, + "grad_norm": 5.723855746996082, + "learning_rate": 2.251669123072449e-06, + "loss": 0.114, + "step": 4661 + }, + { + "epoch": 3.353353713360906, + "grad_norm": 5.322550712106414, + "learning_rate": 2.251374662104406e-06, + "loss": 0.1994, + "step": 4662 + }, + { + "epoch": 3.354073008451717, + "grad_norm": 1.7644032474537261, + "learning_rate": 2.251080162476818e-06, + "loss": 0.0064, + "step": 4663 + }, + { + "epoch": 3.354792303542528, + "grad_norm": 2.5606621465875317, + "learning_rate": 2.250785624204837e-06, + "loss": 0.0458, + "step": 4664 + }, + { + "epoch": 3.3555115986333393, + "grad_norm": 1.4876610701893838, + "learning_rate": 2.250491047303618e-06, + "loss": 0.0195, + "step": 4665 + }, + { + "epoch": 3.3562308937241503, + "grad_norm": 7.727388536219961, + "learning_rate": 2.2501964317883166e-06, + "loss": 0.1523, + "step": 4666 + }, + { + "epoch": 3.3569501888149613, + "grad_norm": 2.166299620552467, + "learning_rate": 2.249901777674092e-06, + "loss": 0.0309, + "step": 4667 + }, + { + "epoch": 3.3576694839057724, + "grad_norm": 3.62008146747624, + "learning_rate": 2.2496070849761046e-06, + "loss": 0.1292, + "step": 4668 + }, + { + "epoch": 3.3583887789965834, + "grad_norm": 2.653985730380036, + "learning_rate": 2.2493123537095168e-06, + "loss": 0.0665, + "step": 4669 + }, + { + "epoch": 3.3591080740873944, + "grad_norm": 3.7695583004826085, + "learning_rate": 2.2490175838894925e-06, + "loss": 0.0767, + "step": 4670 + }, + { + "epoch": 3.3598273691782055, + "grad_norm": 0.8552403061605411, + "learning_rate": 2.2487227755311992e-06, + "loss": 0.0089, + "step": 4671 + }, + { + "epoch": 3.3605466642690165, + "grad_norm": 5.442625327055703, + "learning_rate": 2.248427928649805e-06, + "loss": 0.3339, + "step": 4672 + }, + { + "epoch": 3.3612659593598275, + "grad_norm": 3.7944016774693776, + "learning_rate": 2.24813304326048e-06, + "loss": 0.042, + "step": 4673 + }, + { + "epoch": 3.3619852544506386, + "grad_norm": 2.9805827335521, + "learning_rate": 2.247838119378396e-06, + "loss": 0.049, + "step": 4674 + }, + { + "epoch": 3.3627045495414496, + "grad_norm": 2.5559681234181997, + "learning_rate": 2.2475431570187286e-06, + "loss": 0.0656, + "step": 4675 + }, + { + "epoch": 3.36342384463226, + "grad_norm": 3.600913993243652, + "learning_rate": 2.247248156196653e-06, + "loss": 0.0677, + "step": 4676 + }, + { + "epoch": 3.364143139723071, + "grad_norm": 0.9568398140694035, + "learning_rate": 2.246953116927348e-06, + "loss": 0.013, + "step": 4677 + }, + { + "epoch": 3.3648624348138823, + "grad_norm": 5.4087340214151975, + "learning_rate": 2.2466580392259943e-06, + "loss": 0.109, + "step": 4678 + }, + { + "epoch": 3.3655817299046933, + "grad_norm": 4.776792629506669, + "learning_rate": 2.2463629231077736e-06, + "loss": 0.0768, + "step": 4679 + }, + { + "epoch": 3.3663010249955043, + "grad_norm": 0.7862005452467916, + "learning_rate": 2.2460677685878705e-06, + "loss": 0.0026, + "step": 4680 + }, + { + "epoch": 3.3670203200863154, + "grad_norm": 1.2502530924420738, + "learning_rate": 2.245772575681471e-06, + "loss": 0.015, + "step": 4681 + }, + { + "epoch": 3.3677396151771264, + "grad_norm": 2.7579716090563773, + "learning_rate": 2.245477344403763e-06, + "loss": 0.0533, + "step": 4682 + }, + { + "epoch": 3.3684589102679374, + "grad_norm": 2.3366873385121605, + "learning_rate": 2.245182074769937e-06, + "loss": 0.0079, + "step": 4683 + }, + { + "epoch": 3.3691782053587485, + "grad_norm": 2.975293668753321, + "learning_rate": 2.244886766795185e-06, + "loss": 0.0584, + "step": 4684 + }, + { + "epoch": 3.3698975004495595, + "grad_norm": 0.8520672451743246, + "learning_rate": 2.2445914204947023e-06, + "loss": 0.009, + "step": 4685 + }, + { + "epoch": 3.3706167955403705, + "grad_norm": 3.168689809464066, + "learning_rate": 2.2442960358836828e-06, + "loss": 0.0222, + "step": 4686 + }, + { + "epoch": 3.3713360906311816, + "grad_norm": 3.4564929491761203, + "learning_rate": 2.244000612977326e-06, + "loss": 0.0573, + "step": 4687 + }, + { + "epoch": 3.3720553857219926, + "grad_norm": 1.756055278236517, + "learning_rate": 2.243705151790832e-06, + "loss": 0.0457, + "step": 4688 + }, + { + "epoch": 3.3727746808128036, + "grad_norm": 3.67772882765112, + "learning_rate": 2.2434096523394017e-06, + "loss": 0.0645, + "step": 4689 + }, + { + "epoch": 3.3734939759036147, + "grad_norm": 1.62152638725986, + "learning_rate": 2.2431141146382405e-06, + "loss": 0.0038, + "step": 4690 + }, + { + "epoch": 3.3742132709944253, + "grad_norm": 1.2338867089516974, + "learning_rate": 2.242818538702553e-06, + "loss": 0.0188, + "step": 4691 + }, + { + "epoch": 3.3749325660852363, + "grad_norm": 1.2105834463216467, + "learning_rate": 2.2425229245475478e-06, + "loss": 0.0124, + "step": 4692 + }, + { + "epoch": 3.3756518611760473, + "grad_norm": 4.782924634672067, + "learning_rate": 2.242227272188435e-06, + "loss": 0.1057, + "step": 4693 + }, + { + "epoch": 3.3763711562668584, + "grad_norm": 6.56223586524317, + "learning_rate": 2.2419315816404257e-06, + "loss": 0.1561, + "step": 4694 + }, + { + "epoch": 3.3770904513576694, + "grad_norm": 0.874906315979422, + "learning_rate": 2.2416358529187346e-06, + "loss": 0.0081, + "step": 4695 + }, + { + "epoch": 3.3778097464484804, + "grad_norm": 5.931676174794837, + "learning_rate": 2.2413400860385764e-06, + "loss": 0.2462, + "step": 4696 + }, + { + "epoch": 3.3785290415392915, + "grad_norm": 5.629994826764464, + "learning_rate": 2.2410442810151698e-06, + "loss": 0.1874, + "step": 4697 + }, + { + "epoch": 3.3792483366301025, + "grad_norm": 5.238026553500715, + "learning_rate": 2.240748437863734e-06, + "loss": 0.0823, + "step": 4698 + }, + { + "epoch": 3.3799676317209135, + "grad_norm": 4.131621153131269, + "learning_rate": 2.2404525565994904e-06, + "loss": 0.0354, + "step": 4699 + }, + { + "epoch": 3.3806869268117246, + "grad_norm": 3.4533131216732538, + "learning_rate": 2.2401566372376638e-06, + "loss": 0.0553, + "step": 4700 + }, + { + "epoch": 3.3814062219025356, + "grad_norm": 6.023716375523928, + "learning_rate": 2.239860679793478e-06, + "loss": 0.08, + "step": 4701 + }, + { + "epoch": 3.3821255169933466, + "grad_norm": 1.1710310757796447, + "learning_rate": 2.239564684282162e-06, + "loss": 0.0106, + "step": 4702 + }, + { + "epoch": 3.3828448120841577, + "grad_norm": 2.700397450236177, + "learning_rate": 2.2392686507189445e-06, + "loss": 0.0581, + "step": 4703 + }, + { + "epoch": 3.3835641071749687, + "grad_norm": 3.049607382976599, + "learning_rate": 2.238972579119057e-06, + "loss": 0.0396, + "step": 4704 + }, + { + "epoch": 3.3842834022657797, + "grad_norm": 0.5042966511924528, + "learning_rate": 2.2386764694977334e-06, + "loss": 0.0012, + "step": 4705 + }, + { + "epoch": 3.3850026973565903, + "grad_norm": 0.5870613057463729, + "learning_rate": 2.2383803218702087e-06, + "loss": 0.0061, + "step": 4706 + }, + { + "epoch": 3.385721992447402, + "grad_norm": 2.7772328619387063, + "learning_rate": 2.23808413625172e-06, + "loss": 0.0754, + "step": 4707 + }, + { + "epoch": 3.3864412875382124, + "grad_norm": 3.8853509258692345, + "learning_rate": 2.237787912657507e-06, + "loss": 0.101, + "step": 4708 + }, + { + "epoch": 3.3871605826290234, + "grad_norm": 1.2301877074023235, + "learning_rate": 2.2374916511028104e-06, + "loss": 0.019, + "step": 4709 + }, + { + "epoch": 3.3878798777198345, + "grad_norm": 0.06698512511063469, + "learning_rate": 2.237195351602874e-06, + "loss": 0.0002, + "step": 4710 + }, + { + "epoch": 3.3885991728106455, + "grad_norm": 2.1117887089124165, + "learning_rate": 2.236899014172943e-06, + "loss": 0.0193, + "step": 4711 + }, + { + "epoch": 3.3893184679014565, + "grad_norm": 2.3698672501265694, + "learning_rate": 2.236602638828264e-06, + "loss": 0.0588, + "step": 4712 + }, + { + "epoch": 3.3900377629922676, + "grad_norm": 1.477754247353774, + "learning_rate": 2.2363062255840857e-06, + "loss": 0.0325, + "step": 4713 + }, + { + "epoch": 3.3907570580830786, + "grad_norm": 4.545704100593133, + "learning_rate": 2.23600977445566e-06, + "loss": 0.158, + "step": 4714 + }, + { + "epoch": 3.3914763531738896, + "grad_norm": 3.5984078643191704, + "learning_rate": 2.235713285458239e-06, + "loss": 0.1288, + "step": 4715 + }, + { + "epoch": 3.3921956482647007, + "grad_norm": 2.328364900180709, + "learning_rate": 2.2354167586070785e-06, + "loss": 0.0508, + "step": 4716 + }, + { + "epoch": 3.3929149433555117, + "grad_norm": 1.1801729186809167, + "learning_rate": 2.235120193917435e-06, + "loss": 0.023, + "step": 4717 + }, + { + "epoch": 3.3936342384463227, + "grad_norm": 2.9665168857646584, + "learning_rate": 2.234823591404566e-06, + "loss": 0.0739, + "step": 4718 + }, + { + "epoch": 3.3943535335371338, + "grad_norm": 1.6817976416840759, + "learning_rate": 2.234526951083734e-06, + "loss": 0.0232, + "step": 4719 + }, + { + "epoch": 3.395072828627945, + "grad_norm": 5.154914619240526, + "learning_rate": 2.2342302729702003e-06, + "loss": 0.1776, + "step": 4720 + }, + { + "epoch": 3.3957921237187554, + "grad_norm": 0.4536112945459199, + "learning_rate": 2.233933557079231e-06, + "loss": 0.001, + "step": 4721 + }, + { + "epoch": 3.396511418809567, + "grad_norm": 6.357817633434567, + "learning_rate": 2.2336368034260918e-06, + "loss": 0.1256, + "step": 4722 + }, + { + "epoch": 3.3972307139003775, + "grad_norm": 4.672749765779479, + "learning_rate": 2.2333400120260502e-06, + "loss": 0.149, + "step": 4723 + }, + { + "epoch": 3.3979500089911885, + "grad_norm": 3.8458885065188166, + "learning_rate": 2.233043182894379e-06, + "loss": 0.0974, + "step": 4724 + }, + { + "epoch": 3.3986693040819995, + "grad_norm": 1.3360643800555752, + "learning_rate": 2.232746316046348e-06, + "loss": 0.0176, + "step": 4725 + }, + { + "epoch": 3.3993885991728106, + "grad_norm": 3.417572831543109, + "learning_rate": 2.2324494114972335e-06, + "loss": 0.0443, + "step": 4726 + }, + { + "epoch": 3.4001078942636216, + "grad_norm": 1.2847969351838116, + "learning_rate": 2.2321524692623107e-06, + "loss": 0.0186, + "step": 4727 + }, + { + "epoch": 3.4008271893544326, + "grad_norm": 7.080391959934435, + "learning_rate": 2.231855489356858e-06, + "loss": 0.146, + "step": 4728 + }, + { + "epoch": 3.4015464844452437, + "grad_norm": 0.08117420920412446, + "learning_rate": 2.2315584717961557e-06, + "loss": 0.0004, + "step": 4729 + }, + { + "epoch": 3.4022657795360547, + "grad_norm": 3.2496910956414493, + "learning_rate": 2.231261416595486e-06, + "loss": 0.1129, + "step": 4730 + }, + { + "epoch": 3.4029850746268657, + "grad_norm": 2.062150696300699, + "learning_rate": 2.2309643237701327e-06, + "loss": 0.0553, + "step": 4731 + }, + { + "epoch": 3.4037043697176768, + "grad_norm": 2.9589660811934393, + "learning_rate": 2.230667193335382e-06, + "loss": 0.0203, + "step": 4732 + }, + { + "epoch": 3.404423664808488, + "grad_norm": 2.3570763722052503, + "learning_rate": 2.2303700253065208e-06, + "loss": 0.0507, + "step": 4733 + }, + { + "epoch": 3.405142959899299, + "grad_norm": 5.094271613780984, + "learning_rate": 2.2300728196988405e-06, + "loss": 0.1658, + "step": 4734 + }, + { + "epoch": 3.40586225499011, + "grad_norm": 2.452345117430938, + "learning_rate": 2.229775576527632e-06, + "loss": 0.0371, + "step": 4735 + }, + { + "epoch": 3.4065815500809205, + "grad_norm": 0.9341596809003695, + "learning_rate": 2.229478295808188e-06, + "loss": 0.0215, + "step": 4736 + }, + { + "epoch": 3.407300845171732, + "grad_norm": 2.753016232623573, + "learning_rate": 2.2291809775558058e-06, + "loss": 0.0567, + "step": 4737 + }, + { + "epoch": 3.4080201402625425, + "grad_norm": 7.266014821988594, + "learning_rate": 2.2288836217857825e-06, + "loss": 0.0234, + "step": 4738 + }, + { + "epoch": 3.4087394353533536, + "grad_norm": 3.2188200848077315, + "learning_rate": 2.2285862285134173e-06, + "loss": 0.0188, + "step": 4739 + }, + { + "epoch": 3.4094587304441646, + "grad_norm": 2.3016824937207137, + "learning_rate": 2.228288797754011e-06, + "loss": 0.0458, + "step": 4740 + }, + { + "epoch": 3.4101780255349756, + "grad_norm": 2.072983380105102, + "learning_rate": 2.2279913295228682e-06, + "loss": 0.0338, + "step": 4741 + }, + { + "epoch": 3.4108973206257867, + "grad_norm": 3.8986320642007835, + "learning_rate": 2.2276938238352935e-06, + "loss": 0.1289, + "step": 4742 + }, + { + "epoch": 3.4116166157165977, + "grad_norm": 1.9355951217390621, + "learning_rate": 2.227396280706594e-06, + "loss": 0.0637, + "step": 4743 + }, + { + "epoch": 3.4123359108074087, + "grad_norm": 1.7729114699246107, + "learning_rate": 2.227098700152079e-06, + "loss": 0.0447, + "step": 4744 + }, + { + "epoch": 3.4130552058982198, + "grad_norm": 2.938018749472116, + "learning_rate": 2.226801082187059e-06, + "loss": 0.0841, + "step": 4745 + }, + { + "epoch": 3.413774500989031, + "grad_norm": 4.51382066051197, + "learning_rate": 2.2265034268268483e-06, + "loss": 0.1777, + "step": 4746 + }, + { + "epoch": 3.414493796079842, + "grad_norm": 4.276834181330756, + "learning_rate": 2.226205734086761e-06, + "loss": 0.065, + "step": 4747 + }, + { + "epoch": 3.415213091170653, + "grad_norm": 3.1470600200877135, + "learning_rate": 2.2259080039821136e-06, + "loss": 0.0296, + "step": 4748 + }, + { + "epoch": 3.415932386261464, + "grad_norm": 2.3592435874716364, + "learning_rate": 2.2256102365282243e-06, + "loss": 0.0127, + "step": 4749 + }, + { + "epoch": 3.416651681352275, + "grad_norm": 1.5945228046146487, + "learning_rate": 2.2253124317404156e-06, + "loss": 0.0398, + "step": 4750 + }, + { + "epoch": 3.417370976443086, + "grad_norm": 4.126296426157023, + "learning_rate": 2.2250145896340086e-06, + "loss": 0.0123, + "step": 4751 + }, + { + "epoch": 3.418090271533897, + "grad_norm": 4.164000556128089, + "learning_rate": 2.224716710224328e-06, + "loss": 0.0986, + "step": 4752 + }, + { + "epoch": 3.4188095666247076, + "grad_norm": 3.4981189654928304, + "learning_rate": 2.224418793526701e-06, + "loss": 0.0717, + "step": 4753 + }, + { + "epoch": 3.4195288617155186, + "grad_norm": 3.431808144491488, + "learning_rate": 2.2241208395564552e-06, + "loss": 0.1565, + "step": 4754 + }, + { + "epoch": 3.4202481568063297, + "grad_norm": 2.461926186203359, + "learning_rate": 2.223822848328921e-06, + "loss": 0.0601, + "step": 4755 + }, + { + "epoch": 3.4209674518971407, + "grad_norm": 0.8505056629769038, + "learning_rate": 2.223524819859431e-06, + "loss": 0.0015, + "step": 4756 + }, + { + "epoch": 3.4216867469879517, + "grad_norm": 4.657436120293943, + "learning_rate": 2.223226754163318e-06, + "loss": 0.0262, + "step": 4757 + }, + { + "epoch": 3.4224060420787628, + "grad_norm": 2.0413337749692633, + "learning_rate": 2.2229286512559198e-06, + "loss": 0.0589, + "step": 4758 + }, + { + "epoch": 3.423125337169574, + "grad_norm": 3.1019435858687157, + "learning_rate": 2.222630511152573e-06, + "loss": 0.0572, + "step": 4759 + }, + { + "epoch": 3.423844632260385, + "grad_norm": 8.692928929918308, + "learning_rate": 2.222332333868618e-06, + "loss": 0.1414, + "step": 4760 + }, + { + "epoch": 3.424563927351196, + "grad_norm": 2.312905081810137, + "learning_rate": 2.2220341194193968e-06, + "loss": 0.0516, + "step": 4761 + }, + { + "epoch": 3.425283222442007, + "grad_norm": 4.827060618384299, + "learning_rate": 2.221735867820252e-06, + "loss": 0.1791, + "step": 4762 + }, + { + "epoch": 3.426002517532818, + "grad_norm": 1.0573534294508333, + "learning_rate": 2.2214375790865296e-06, + "loss": 0.013, + "step": 4763 + }, + { + "epoch": 3.426721812623629, + "grad_norm": 4.310965306608159, + "learning_rate": 2.221139253233578e-06, + "loss": 0.1026, + "step": 4764 + }, + { + "epoch": 3.42744110771444, + "grad_norm": 3.3568607263317216, + "learning_rate": 2.2208408902767454e-06, + "loss": 0.0919, + "step": 4765 + }, + { + "epoch": 3.428160402805251, + "grad_norm": 1.7000930435793433, + "learning_rate": 2.220542490231384e-06, + "loss": 0.0448, + "step": 4766 + }, + { + "epoch": 3.428879697896062, + "grad_norm": 0.8983436418572041, + "learning_rate": 2.2202440531128468e-06, + "loss": 0.01, + "step": 4767 + }, + { + "epoch": 3.4295989929868727, + "grad_norm": 0.5820506919658236, + "learning_rate": 2.2199455789364886e-06, + "loss": 0.0051, + "step": 4768 + }, + { + "epoch": 3.4303182880776837, + "grad_norm": 1.844664191878256, + "learning_rate": 2.219647067717666e-06, + "loss": 0.0201, + "step": 4769 + }, + { + "epoch": 3.4310375831684947, + "grad_norm": 2.6896506284078194, + "learning_rate": 2.219348519471739e-06, + "loss": 0.0693, + "step": 4770 + }, + { + "epoch": 3.4317568782593058, + "grad_norm": 0.8393228314382527, + "learning_rate": 2.2190499342140676e-06, + "loss": 0.0042, + "step": 4771 + }, + { + "epoch": 3.432476173350117, + "grad_norm": 1.0645244329435726, + "learning_rate": 2.2187513119600148e-06, + "loss": 0.0018, + "step": 4772 + }, + { + "epoch": 3.433195468440928, + "grad_norm": 2.5849355096930937, + "learning_rate": 2.2184526527249456e-06, + "loss": 0.1126, + "step": 4773 + }, + { + "epoch": 3.433914763531739, + "grad_norm": 3.6215160325631706, + "learning_rate": 2.2181539565242258e-06, + "loss": 0.0437, + "step": 4774 + }, + { + "epoch": 3.43463405862255, + "grad_norm": 0.27295305371903983, + "learning_rate": 2.2178552233732246e-06, + "loss": 0.0009, + "step": 4775 + }, + { + "epoch": 3.435353353713361, + "grad_norm": 2.0767110673034637, + "learning_rate": 2.217556453287312e-06, + "loss": 0.0473, + "step": 4776 + }, + { + "epoch": 3.436072648804172, + "grad_norm": 3.623912487586658, + "learning_rate": 2.21725764628186e-06, + "loss": 0.0642, + "step": 4777 + }, + { + "epoch": 3.436791943894983, + "grad_norm": 0.2156963663171608, + "learning_rate": 2.216958802372243e-06, + "loss": 0.0007, + "step": 4778 + }, + { + "epoch": 3.437511238985794, + "grad_norm": 1.407256471520192, + "learning_rate": 2.2166599215738374e-06, + "loss": 0.0029, + "step": 4779 + }, + { + "epoch": 3.438230534076605, + "grad_norm": 2.2997588575231753, + "learning_rate": 2.2163610039020204e-06, + "loss": 0.0072, + "step": 4780 + }, + { + "epoch": 3.438949829167416, + "grad_norm": 4.880387905896978, + "learning_rate": 2.216062049372172e-06, + "loss": 0.2318, + "step": 4781 + }, + { + "epoch": 3.439669124258227, + "grad_norm": 4.383579327297662, + "learning_rate": 2.2157630579996752e-06, + "loss": 0.1023, + "step": 4782 + }, + { + "epoch": 3.4403884193490377, + "grad_norm": 2.269023261897153, + "learning_rate": 2.2154640297999117e-06, + "loss": 0.0372, + "step": 4783 + }, + { + "epoch": 3.441107714439849, + "grad_norm": 5.8591300479251665, + "learning_rate": 2.2151649647882683e-06, + "loss": 0.089, + "step": 4784 + }, + { + "epoch": 3.44182700953066, + "grad_norm": 1.8839948307234988, + "learning_rate": 2.2148658629801317e-06, + "loss": 0.0062, + "step": 4785 + }, + { + "epoch": 3.442546304621471, + "grad_norm": 1.5117026637128765, + "learning_rate": 2.2145667243908917e-06, + "loss": 0.003, + "step": 4786 + }, + { + "epoch": 3.443265599712282, + "grad_norm": 2.0335287134665556, + "learning_rate": 2.2142675490359392e-06, + "loss": 0.0353, + "step": 4787 + }, + { + "epoch": 3.443984894803093, + "grad_norm": 3.9371115547260778, + "learning_rate": 2.213968336930667e-06, + "loss": 0.04, + "step": 4788 + }, + { + "epoch": 3.444704189893904, + "grad_norm": 3.218963408983558, + "learning_rate": 2.2136690880904713e-06, + "loss": 0.099, + "step": 4789 + }, + { + "epoch": 3.445423484984715, + "grad_norm": 2.425191887586878, + "learning_rate": 2.2133698025307485e-06, + "loss": 0.0455, + "step": 4790 + }, + { + "epoch": 3.446142780075526, + "grad_norm": 2.4118511727463092, + "learning_rate": 2.2130704802668964e-06, + "loss": 0.0374, + "step": 4791 + }, + { + "epoch": 3.446862075166337, + "grad_norm": 7.404532667329372, + "learning_rate": 2.2127711213143165e-06, + "loss": 0.1302, + "step": 4792 + }, + { + "epoch": 3.447581370257148, + "grad_norm": 1.497443467854853, + "learning_rate": 2.212471725688411e-06, + "loss": 0.042, + "step": 4793 + }, + { + "epoch": 3.448300665347959, + "grad_norm": 3.2975377265124606, + "learning_rate": 2.212172293404585e-06, + "loss": 0.1353, + "step": 4794 + }, + { + "epoch": 3.44901996043877, + "grad_norm": 0.5311910963585446, + "learning_rate": 2.211872824478244e-06, + "loss": 0.0033, + "step": 4795 + }, + { + "epoch": 3.449739255529581, + "grad_norm": 2.6612211662663072, + "learning_rate": 2.2115733189247966e-06, + "loss": 0.0444, + "step": 4796 + }, + { + "epoch": 3.450458550620392, + "grad_norm": 5.314415983328357, + "learning_rate": 2.2112737767596527e-06, + "loss": 0.1169, + "step": 4797 + }, + { + "epoch": 3.451177845711203, + "grad_norm": 1.1604427158550368, + "learning_rate": 2.2109741979982245e-06, + "loss": 0.0165, + "step": 4798 + }, + { + "epoch": 3.4518971408020143, + "grad_norm": 2.397253116568816, + "learning_rate": 2.2106745826559256e-06, + "loss": 0.0107, + "step": 4799 + }, + { + "epoch": 3.452616435892825, + "grad_norm": 2.995297121283954, + "learning_rate": 2.210374930748172e-06, + "loss": 0.0398, + "step": 4800 + }, + { + "epoch": 3.453335730983636, + "grad_norm": 5.240625330951832, + "learning_rate": 2.210075242290381e-06, + "loss": 0.0503, + "step": 4801 + }, + { + "epoch": 3.454055026074447, + "grad_norm": 2.337034381877142, + "learning_rate": 2.2097755172979722e-06, + "loss": 0.0053, + "step": 4802 + }, + { + "epoch": 3.454774321165258, + "grad_norm": 2.9441691203761993, + "learning_rate": 2.209475755786368e-06, + "loss": 0.0769, + "step": 4803 + }, + { + "epoch": 3.455493616256069, + "grad_norm": 0.33313153810628077, + "learning_rate": 2.20917595777099e-06, + "loss": 0.0008, + "step": 4804 + }, + { + "epoch": 3.45621291134688, + "grad_norm": 4.820762431462426, + "learning_rate": 2.208876123267264e-06, + "loss": 0.051, + "step": 4805 + }, + { + "epoch": 3.456932206437691, + "grad_norm": 3.1215357916527964, + "learning_rate": 2.208576252290617e-06, + "loss": 0.0915, + "step": 4806 + }, + { + "epoch": 3.457651501528502, + "grad_norm": 5.115781000789495, + "learning_rate": 2.208276344856478e-06, + "loss": 0.1369, + "step": 4807 + }, + { + "epoch": 3.458370796619313, + "grad_norm": 4.257051483880606, + "learning_rate": 2.207976400980278e-06, + "loss": 0.0842, + "step": 4808 + }, + { + "epoch": 3.459090091710124, + "grad_norm": 5.732060199041762, + "learning_rate": 2.207676420677449e-06, + "loss": 0.0917, + "step": 4809 + }, + { + "epoch": 3.459809386800935, + "grad_norm": 3.8551372652742573, + "learning_rate": 2.207376403963426e-06, + "loss": 0.0795, + "step": 4810 + }, + { + "epoch": 3.4605286818917462, + "grad_norm": 5.766331900216892, + "learning_rate": 2.2070763508536457e-06, + "loss": 0.1073, + "step": 4811 + }, + { + "epoch": 3.4612479769825573, + "grad_norm": 1.768230523377009, + "learning_rate": 2.2067762613635452e-06, + "loss": 0.0055, + "step": 4812 + }, + { + "epoch": 3.461967272073368, + "grad_norm": 0.3718786956399877, + "learning_rate": 2.2064761355085657e-06, + "loss": 0.0022, + "step": 4813 + }, + { + "epoch": 3.4626865671641793, + "grad_norm": 0.8282951884961748, + "learning_rate": 2.2061759733041485e-06, + "loss": 0.0017, + "step": 4814 + }, + { + "epoch": 3.46340586225499, + "grad_norm": 1.603258333835655, + "learning_rate": 2.2058757747657385e-06, + "loss": 0.0193, + "step": 4815 + }, + { + "epoch": 3.464125157345801, + "grad_norm": 0.34204251628380355, + "learning_rate": 2.2055755399087804e-06, + "loss": 0.0026, + "step": 4816 + }, + { + "epoch": 3.464844452436612, + "grad_norm": 3.8597933053417743, + "learning_rate": 2.205275268748722e-06, + "loss": 0.0734, + "step": 4817 + }, + { + "epoch": 3.465563747527423, + "grad_norm": 6.299047268610459, + "learning_rate": 2.2049749613010132e-06, + "loss": 0.1062, + "step": 4818 + }, + { + "epoch": 3.466283042618234, + "grad_norm": 3.7208291349345775, + "learning_rate": 2.2046746175811052e-06, + "loss": 0.0733, + "step": 4819 + }, + { + "epoch": 3.467002337709045, + "grad_norm": 1.1055361911179837, + "learning_rate": 2.204374237604451e-06, + "loss": 0.0025, + "step": 4820 + }, + { + "epoch": 3.467721632799856, + "grad_norm": 5.443168655841614, + "learning_rate": 2.2040738213865054e-06, + "loss": 0.1361, + "step": 4821 + }, + { + "epoch": 3.468440927890667, + "grad_norm": 5.0597320958529055, + "learning_rate": 2.203773368942726e-06, + "loss": 0.0979, + "step": 4822 + }, + { + "epoch": 3.469160222981478, + "grad_norm": 5.356266334212832, + "learning_rate": 2.2034728802885715e-06, + "loss": 0.0908, + "step": 4823 + }, + { + "epoch": 3.4698795180722892, + "grad_norm": 5.114737848695762, + "learning_rate": 2.2031723554395022e-06, + "loss": 0.1271, + "step": 4824 + }, + { + "epoch": 3.4705988131631003, + "grad_norm": 3.0881994600138327, + "learning_rate": 2.202871794410981e-06, + "loss": 0.0576, + "step": 4825 + }, + { + "epoch": 3.4713181082539113, + "grad_norm": 4.7809283534040485, + "learning_rate": 2.202571197218472e-06, + "loss": 0.1225, + "step": 4826 + }, + { + "epoch": 3.4720374033447223, + "grad_norm": 1.8495115894811514, + "learning_rate": 2.2022705638774416e-06, + "loss": 0.0597, + "step": 4827 + }, + { + "epoch": 3.4727566984355334, + "grad_norm": 4.455856983776783, + "learning_rate": 2.2019698944033576e-06, + "loss": 0.0893, + "step": 4828 + }, + { + "epoch": 3.4734759935263444, + "grad_norm": 7.28770115351474, + "learning_rate": 2.201669188811691e-06, + "loss": 0.0199, + "step": 4829 + }, + { + "epoch": 3.474195288617155, + "grad_norm": 3.4951082900396933, + "learning_rate": 2.2013684471179123e-06, + "loss": 0.0092, + "step": 4830 + }, + { + "epoch": 3.474914583707966, + "grad_norm": 5.740298402600414, + "learning_rate": 2.201067669337496e-06, + "loss": 0.2041, + "step": 4831 + }, + { + "epoch": 3.475633878798777, + "grad_norm": 4.843091644427837, + "learning_rate": 2.2007668554859173e-06, + "loss": 0.1439, + "step": 4832 + }, + { + "epoch": 3.476353173889588, + "grad_norm": 2.814245508710047, + "learning_rate": 2.2004660055786537e-06, + "loss": 0.0458, + "step": 4833 + }, + { + "epoch": 3.477072468980399, + "grad_norm": 4.0456574482737455, + "learning_rate": 2.200165119631185e-06, + "loss": 0.0715, + "step": 4834 + }, + { + "epoch": 3.47779176407121, + "grad_norm": 4.0440437231253314, + "learning_rate": 2.199864197658991e-06, + "loss": 0.0666, + "step": 4835 + }, + { + "epoch": 3.478511059162021, + "grad_norm": 1.902982336394222, + "learning_rate": 2.199563239677556e-06, + "loss": 0.0056, + "step": 4836 + }, + { + "epoch": 3.4792303542528322, + "grad_norm": 6.2582767766186285, + "learning_rate": 2.199262245702364e-06, + "loss": 0.1685, + "step": 4837 + }, + { + "epoch": 3.4799496493436433, + "grad_norm": 3.1754417759809566, + "learning_rate": 2.1989612157489017e-06, + "loss": 0.0316, + "step": 4838 + }, + { + "epoch": 3.4806689444344543, + "grad_norm": 6.567200345995638, + "learning_rate": 2.1986601498326584e-06, + "loss": 0.0496, + "step": 4839 + }, + { + "epoch": 3.4813882395252653, + "grad_norm": 4.70674847353393, + "learning_rate": 2.1983590479691236e-06, + "loss": 0.1979, + "step": 4840 + }, + { + "epoch": 3.4821075346160764, + "grad_norm": 5.48478163915878, + "learning_rate": 2.19805791017379e-06, + "loss": 0.0166, + "step": 4841 + }, + { + "epoch": 3.4828268297068874, + "grad_norm": 0.5247275126491133, + "learning_rate": 2.197756736462152e-06, + "loss": 0.001, + "step": 4842 + }, + { + "epoch": 3.4835461247976984, + "grad_norm": 2.7112718389133836, + "learning_rate": 2.197455526849704e-06, + "loss": 0.0525, + "step": 4843 + }, + { + "epoch": 3.4842654198885095, + "grad_norm": 2.221012068173011, + "learning_rate": 2.1971542813519455e-06, + "loss": 0.0301, + "step": 4844 + }, + { + "epoch": 3.48498471497932, + "grad_norm": 4.475693317034441, + "learning_rate": 2.196852999984375e-06, + "loss": 0.1899, + "step": 4845 + }, + { + "epoch": 3.485704010070131, + "grad_norm": 1.7437235918549527, + "learning_rate": 2.196551682762495e-06, + "loss": 0.0462, + "step": 4846 + }, + { + "epoch": 3.486423305160942, + "grad_norm": 0.16984235785490032, + "learning_rate": 2.1962503297018076e-06, + "loss": 0.0004, + "step": 4847 + }, + { + "epoch": 3.487142600251753, + "grad_norm": 7.079868497042358, + "learning_rate": 2.195948940817819e-06, + "loss": 0.07, + "step": 4848 + }, + { + "epoch": 3.487861895342564, + "grad_norm": 3.7555040124173957, + "learning_rate": 2.1956475161260353e-06, + "loss": 0.1158, + "step": 4849 + }, + { + "epoch": 3.4885811904333752, + "grad_norm": 4.491133366888093, + "learning_rate": 2.1953460556419657e-06, + "loss": 0.0928, + "step": 4850 + }, + { + "epoch": 3.4893004855241863, + "grad_norm": 4.724994717066071, + "learning_rate": 2.195044559381121e-06, + "loss": 0.1828, + "step": 4851 + }, + { + "epoch": 3.4900197806149973, + "grad_norm": 3.9067909934061693, + "learning_rate": 2.1947430273590135e-06, + "loss": 0.098, + "step": 4852 + }, + { + "epoch": 3.4907390757058083, + "grad_norm": 4.2038001135107885, + "learning_rate": 2.1944414595911577e-06, + "loss": 0.0802, + "step": 4853 + }, + { + "epoch": 3.4914583707966194, + "grad_norm": 3.432742509832068, + "learning_rate": 2.19413985609307e-06, + "loss": 0.0218, + "step": 4854 + }, + { + "epoch": 3.4921776658874304, + "grad_norm": 0.7951939351877209, + "learning_rate": 2.193838216880268e-06, + "loss": 0.0021, + "step": 4855 + }, + { + "epoch": 3.4928969609782414, + "grad_norm": 6.390214057799085, + "learning_rate": 2.1935365419682715e-06, + "loss": 0.1049, + "step": 4856 + }, + { + "epoch": 3.4936162560690525, + "grad_norm": 2.1418844754843813, + "learning_rate": 2.193234831372603e-06, + "loss": 0.0399, + "step": 4857 + }, + { + "epoch": 3.4943355511598635, + "grad_norm": 1.6271091114694236, + "learning_rate": 2.192933085108785e-06, + "loss": 0.0308, + "step": 4858 + }, + { + "epoch": 3.4950548462506745, + "grad_norm": 3.817326141273365, + "learning_rate": 2.1926313031923435e-06, + "loss": 0.1025, + "step": 4859 + }, + { + "epoch": 3.495774141341485, + "grad_norm": 4.220084374943218, + "learning_rate": 2.1923294856388054e-06, + "loss": 0.0271, + "step": 4860 + }, + { + "epoch": 3.4964934364322966, + "grad_norm": 3.7373019386303117, + "learning_rate": 2.1920276324637008e-06, + "loss": 0.1036, + "step": 4861 + }, + { + "epoch": 3.497212731523107, + "grad_norm": 1.5690691261621614, + "learning_rate": 2.1917257436825586e-06, + "loss": 0.0101, + "step": 4862 + }, + { + "epoch": 3.497932026613918, + "grad_norm": 3.6064158289155746, + "learning_rate": 2.1914238193109136e-06, + "loss": 0.1491, + "step": 4863 + }, + { + "epoch": 3.4986513217047293, + "grad_norm": 3.743077382579993, + "learning_rate": 2.1911218593642986e-06, + "loss": 0.1132, + "step": 4864 + }, + { + "epoch": 3.4993706167955403, + "grad_norm": 3.187245347990246, + "learning_rate": 2.1908198638582506e-06, + "loss": 0.0114, + "step": 4865 + }, + { + "epoch": 3.5000899118863513, + "grad_norm": 2.1084257743348456, + "learning_rate": 2.1905178328083082e-06, + "loss": 0.0174, + "step": 4866 + }, + { + "epoch": 3.5008092069771624, + "grad_norm": 2.763042267534291, + "learning_rate": 2.190215766230011e-06, + "loss": 0.0686, + "step": 4867 + }, + { + "epoch": 3.5015285020679734, + "grad_norm": 5.438437336498403, + "learning_rate": 2.1899136641389015e-06, + "loss": 0.0419, + "step": 4868 + }, + { + "epoch": 3.5022477971587844, + "grad_norm": 2.6532191757757873, + "learning_rate": 2.1896115265505226e-06, + "loss": 0.0629, + "step": 4869 + }, + { + "epoch": 3.5029670922495955, + "grad_norm": 4.757314616140639, + "learning_rate": 2.1893093534804196e-06, + "loss": 0.1382, + "step": 4870 + }, + { + "epoch": 3.5036863873404065, + "grad_norm": 3.552614355791362, + "learning_rate": 2.1890071449441413e-06, + "loss": 0.0981, + "step": 4871 + }, + { + "epoch": 3.5044056824312175, + "grad_norm": 1.8651862808933177, + "learning_rate": 2.1887049009572348e-06, + "loss": 0.0368, + "step": 4872 + }, + { + "epoch": 3.5051249775220286, + "grad_norm": 1.3662920651085406, + "learning_rate": 2.1884026215352526e-06, + "loss": 0.0251, + "step": 4873 + }, + { + "epoch": 3.5058442726128396, + "grad_norm": 1.1599515501537767, + "learning_rate": 2.1881003066937474e-06, + "loss": 0.0206, + "step": 4874 + }, + { + "epoch": 3.50656356770365, + "grad_norm": 1.7465249054354721, + "learning_rate": 2.1877979564482725e-06, + "loss": 0.004, + "step": 4875 + }, + { + "epoch": 3.5072828627944617, + "grad_norm": 2.6156387384693174, + "learning_rate": 2.1874955708143866e-06, + "loss": 0.0718, + "step": 4876 + }, + { + "epoch": 3.5080021578852723, + "grad_norm": 2.413137535371077, + "learning_rate": 2.1871931498076464e-06, + "loss": 0.024, + "step": 4877 + }, + { + "epoch": 3.5087214529760833, + "grad_norm": 3.629325701602865, + "learning_rate": 2.1868906934436117e-06, + "loss": 0.1032, + "step": 4878 + }, + { + "epoch": 3.5094407480668943, + "grad_norm": 3.2838490134698843, + "learning_rate": 2.186588201737846e-06, + "loss": 0.0643, + "step": 4879 + }, + { + "epoch": 3.5101600431577054, + "grad_norm": 2.212070957134834, + "learning_rate": 2.186285674705911e-06, + "loss": 0.0312, + "step": 4880 + }, + { + "epoch": 3.5108793382485164, + "grad_norm": 3.2144864553877786, + "learning_rate": 2.185983112363374e-06, + "loss": 0.1228, + "step": 4881 + }, + { + "epoch": 3.5115986333393274, + "grad_norm": 3.1966301414430887, + "learning_rate": 2.1856805147258013e-06, + "loss": 0.0033, + "step": 4882 + }, + { + "epoch": 3.5123179284301385, + "grad_norm": 0.04427591360824512, + "learning_rate": 2.1853778818087623e-06, + "loss": 0.0003, + "step": 4883 + }, + { + "epoch": 3.5130372235209495, + "grad_norm": 4.143326777970669, + "learning_rate": 2.1850752136278283e-06, + "loss": 0.0796, + "step": 4884 + }, + { + "epoch": 3.5137565186117605, + "grad_norm": 2.872072409903535, + "learning_rate": 2.1847725101985723e-06, + "loss": 0.0825, + "step": 4885 + }, + { + "epoch": 3.5144758137025716, + "grad_norm": 1.5345612104279909, + "learning_rate": 2.1844697715365684e-06, + "loss": 0.0179, + "step": 4886 + }, + { + "epoch": 3.5151951087933826, + "grad_norm": 0.5790961448218425, + "learning_rate": 2.184166997657393e-06, + "loss": 0.0019, + "step": 4887 + }, + { + "epoch": 3.5159144038841936, + "grad_norm": 2.9251428085817133, + "learning_rate": 2.1838641885766245e-06, + "loss": 0.0693, + "step": 4888 + }, + { + "epoch": 3.5166336989750047, + "grad_norm": 2.985852853842518, + "learning_rate": 2.1835613443098435e-06, + "loss": 0.0822, + "step": 4889 + }, + { + "epoch": 3.5173529940658153, + "grad_norm": 1.2479712757385204, + "learning_rate": 2.1832584648726307e-06, + "loss": 0.0034, + "step": 4890 + }, + { + "epoch": 3.5180722891566267, + "grad_norm": 2.20104165452996, + "learning_rate": 2.1829555502805716e-06, + "loss": 0.0135, + "step": 4891 + }, + { + "epoch": 3.5187915842474373, + "grad_norm": 1.5618470854524142, + "learning_rate": 2.18265260054925e-06, + "loss": 0.0203, + "step": 4892 + }, + { + "epoch": 3.5195108793382484, + "grad_norm": 5.055012622752106, + "learning_rate": 2.182349615694253e-06, + "loss": 0.1166, + "step": 4893 + }, + { + "epoch": 3.5202301744290594, + "grad_norm": 1.2706035843690742, + "learning_rate": 2.182046595731171e-06, + "loss": 0.0422, + "step": 4894 + }, + { + "epoch": 3.5209494695198704, + "grad_norm": 7.4233894590197735, + "learning_rate": 2.1817435406755945e-06, + "loss": 0.1161, + "step": 4895 + }, + { + "epoch": 3.5216687646106815, + "grad_norm": 4.319686697014445, + "learning_rate": 2.181440450543116e-06, + "loss": 0.1826, + "step": 4896 + }, + { + "epoch": 3.5223880597014925, + "grad_norm": 0.6589927516925955, + "learning_rate": 2.1811373253493295e-06, + "loss": 0.0013, + "step": 4897 + }, + { + "epoch": 3.5231073547923035, + "grad_norm": 1.6402064615708674, + "learning_rate": 2.1808341651098327e-06, + "loss": 0.0289, + "step": 4898 + }, + { + "epoch": 3.5238266498831146, + "grad_norm": 3.521533270867982, + "learning_rate": 2.1805309698402225e-06, + "loss": 0.0377, + "step": 4899 + }, + { + "epoch": 3.5245459449739256, + "grad_norm": 2.1475470750666843, + "learning_rate": 2.1802277395561e-06, + "loss": 0.0092, + "step": 4900 + }, + { + "epoch": 3.5252652400647366, + "grad_norm": 1.3894314856570087, + "learning_rate": 2.1799244742730653e-06, + "loss": 0.0269, + "step": 4901 + }, + { + "epoch": 3.5259845351555477, + "grad_norm": 0.1949933081185339, + "learning_rate": 2.179621174006723e-06, + "loss": 0.0005, + "step": 4902 + }, + { + "epoch": 3.5267038302463587, + "grad_norm": 2.9381673242638064, + "learning_rate": 2.1793178387726784e-06, + "loss": 0.0094, + "step": 4903 + }, + { + "epoch": 3.5274231253371697, + "grad_norm": 3.092016306548222, + "learning_rate": 2.1790144685865383e-06, + "loss": 0.1498, + "step": 4904 + }, + { + "epoch": 3.5281424204279803, + "grad_norm": 1.6294322519560263, + "learning_rate": 2.1787110634639123e-06, + "loss": 0.0055, + "step": 4905 + }, + { + "epoch": 3.528861715518792, + "grad_norm": 5.463797396451713, + "learning_rate": 2.1784076234204104e-06, + "loss": 0.1025, + "step": 4906 + }, + { + "epoch": 3.5295810106096024, + "grad_norm": 4.952837610595327, + "learning_rate": 2.178104148471645e-06, + "loss": 0.1765, + "step": 4907 + }, + { + "epoch": 3.530300305700414, + "grad_norm": 6.774843612396574, + "learning_rate": 2.177800638633232e-06, + "loss": 0.1159, + "step": 4908 + }, + { + "epoch": 3.5310196007912245, + "grad_norm": 3.4840700682112224, + "learning_rate": 2.177497093920785e-06, + "loss": 0.1628, + "step": 4909 + }, + { + "epoch": 3.5317388958820355, + "grad_norm": 6.199866654305057, + "learning_rate": 2.1771935143499234e-06, + "loss": 0.1818, + "step": 4910 + }, + { + "epoch": 3.5324581909728465, + "grad_norm": 2.046759083356113, + "learning_rate": 2.1768898999362668e-06, + "loss": 0.0241, + "step": 4911 + }, + { + "epoch": 3.5331774860636576, + "grad_norm": 1.9130583771886223, + "learning_rate": 2.1765862506954364e-06, + "loss": 0.0272, + "step": 4912 + }, + { + "epoch": 3.5338967811544686, + "grad_norm": 3.385717255185832, + "learning_rate": 2.1762825666430563e-06, + "loss": 0.0777, + "step": 4913 + }, + { + "epoch": 3.5346160762452796, + "grad_norm": 7.0984841772019385, + "learning_rate": 2.1759788477947507e-06, + "loss": 0.2091, + "step": 4914 + }, + { + "epoch": 3.5353353713360907, + "grad_norm": 0.5230066272528906, + "learning_rate": 2.175675094166147e-06, + "loss": 0.0027, + "step": 4915 + }, + { + "epoch": 3.5360546664269017, + "grad_norm": 3.0556022336705726, + "learning_rate": 2.175371305772873e-06, + "loss": 0.0162, + "step": 4916 + }, + { + "epoch": 3.5367739615177127, + "grad_norm": 2.514462587495062, + "learning_rate": 2.1750674826305603e-06, + "loss": 0.0474, + "step": 4917 + }, + { + "epoch": 3.5374932566085238, + "grad_norm": 3.219340654847328, + "learning_rate": 2.1747636247548403e-06, + "loss": 0.0701, + "step": 4918 + }, + { + "epoch": 3.538212551699335, + "grad_norm": 1.6700729865669701, + "learning_rate": 2.1744597321613472e-06, + "loss": 0.0433, + "step": 4919 + }, + { + "epoch": 3.5389318467901454, + "grad_norm": 1.2812045924955413, + "learning_rate": 2.1741558048657173e-06, + "loss": 0.0149, + "step": 4920 + }, + { + "epoch": 3.539651141880957, + "grad_norm": 2.997424556815624, + "learning_rate": 2.173851842883588e-06, + "loss": 0.0611, + "step": 4921 + }, + { + "epoch": 3.5403704369717675, + "grad_norm": 2.165117561067377, + "learning_rate": 2.1735478462305973e-06, + "loss": 0.0498, + "step": 4922 + }, + { + "epoch": 3.541089732062579, + "grad_norm": 5.220080729522584, + "learning_rate": 2.1732438149223885e-06, + "loss": 0.0575, + "step": 4923 + }, + { + "epoch": 3.5418090271533895, + "grad_norm": 1.818055944976054, + "learning_rate": 2.172939748974603e-06, + "loss": 0.0445, + "step": 4924 + }, + { + "epoch": 3.5425283222442006, + "grad_norm": 3.934929578568546, + "learning_rate": 2.1726356484028865e-06, + "loss": 0.0846, + "step": 4925 + }, + { + "epoch": 3.5432476173350116, + "grad_norm": 2.7358353990192277, + "learning_rate": 2.172331513222885e-06, + "loss": 0.0046, + "step": 4926 + }, + { + "epoch": 3.5439669124258226, + "grad_norm": 3.1842762373118276, + "learning_rate": 2.1720273434502475e-06, + "loss": 0.1191, + "step": 4927 + }, + { + "epoch": 3.5446862075166337, + "grad_norm": 2.8212096207384234, + "learning_rate": 2.171723139100623e-06, + "loss": 0.0542, + "step": 4928 + }, + { + "epoch": 3.5454055026074447, + "grad_norm": 4.130780804685353, + "learning_rate": 2.171418900189664e-06, + "loss": 0.1254, + "step": 4929 + }, + { + "epoch": 3.5461247976982557, + "grad_norm": 4.766454496182892, + "learning_rate": 2.1711146267330233e-06, + "loss": 0.0377, + "step": 4930 + }, + { + "epoch": 3.5468440927890668, + "grad_norm": 2.2615499055780535, + "learning_rate": 2.170810318746357e-06, + "loss": 0.0511, + "step": 4931 + }, + { + "epoch": 3.547563387879878, + "grad_norm": 3.201066184630596, + "learning_rate": 2.1705059762453226e-06, + "loss": 0.0541, + "step": 4932 + }, + { + "epoch": 3.548282682970689, + "grad_norm": 0.1240405240899677, + "learning_rate": 2.170201599245579e-06, + "loss": 0.0004, + "step": 4933 + }, + { + "epoch": 3.5490019780615, + "grad_norm": 0.4175299488632816, + "learning_rate": 2.169897187762786e-06, + "loss": 0.0031, + "step": 4934 + }, + { + "epoch": 3.549721273152311, + "grad_norm": 2.898932980286494, + "learning_rate": 2.169592741812607e-06, + "loss": 0.088, + "step": 4935 + }, + { + "epoch": 3.550440568243122, + "grad_norm": 2.708537208093538, + "learning_rate": 2.169288261410706e-06, + "loss": 0.0062, + "step": 4936 + }, + { + "epoch": 3.5511598633339325, + "grad_norm": 2.1083417651944876, + "learning_rate": 2.1689837465727493e-06, + "loss": 0.0551, + "step": 4937 + }, + { + "epoch": 3.551879158424744, + "grad_norm": 1.37034943613225, + "learning_rate": 2.168679197314404e-06, + "loss": 0.0343, + "step": 4938 + }, + { + "epoch": 3.5525984535155546, + "grad_norm": 7.961506818593601, + "learning_rate": 2.1683746136513404e-06, + "loss": 0.191, + "step": 4939 + }, + { + "epoch": 3.5533177486063656, + "grad_norm": 4.359091854811136, + "learning_rate": 2.1680699955992294e-06, + "loss": 0.0928, + "step": 4940 + }, + { + "epoch": 3.5540370436971767, + "grad_norm": 3.1692972284042558, + "learning_rate": 2.167765343173745e-06, + "loss": 0.0459, + "step": 4941 + }, + { + "epoch": 3.5547563387879877, + "grad_norm": 2.490250825925234, + "learning_rate": 2.167460656390561e-06, + "loss": 0.0741, + "step": 4942 + }, + { + "epoch": 3.5554756338787987, + "grad_norm": 1.257446423892046, + "learning_rate": 2.1671559352653548e-06, + "loss": 0.0286, + "step": 4943 + }, + { + "epoch": 3.5561949289696098, + "grad_norm": 3.2391543776489216, + "learning_rate": 2.1668511798138043e-06, + "loss": 0.1269, + "step": 4944 + }, + { + "epoch": 3.556914224060421, + "grad_norm": 2.312964952252786, + "learning_rate": 2.1665463900515897e-06, + "loss": 0.0426, + "step": 4945 + }, + { + "epoch": 3.557633519151232, + "grad_norm": 1.4233288164345326, + "learning_rate": 2.166241565994394e-06, + "loss": 0.0039, + "step": 4946 + }, + { + "epoch": 3.558352814242043, + "grad_norm": 3.589166672979967, + "learning_rate": 2.1659367076578995e-06, + "loss": 0.1473, + "step": 4947 + }, + { + "epoch": 3.559072109332854, + "grad_norm": 0.5492191115568714, + "learning_rate": 2.1656318150577926e-06, + "loss": 0.0023, + "step": 4948 + }, + { + "epoch": 3.559791404423665, + "grad_norm": 3.9694371695879047, + "learning_rate": 2.1653268882097605e-06, + "loss": 0.0083, + "step": 4949 + }, + { + "epoch": 3.560510699514476, + "grad_norm": 2.6560628024035977, + "learning_rate": 2.1650219271294923e-06, + "loss": 0.0747, + "step": 4950 + }, + { + "epoch": 3.561229994605287, + "grad_norm": 0.9980010998787516, + "learning_rate": 2.164716931832678e-06, + "loss": 0.0244, + "step": 4951 + }, + { + "epoch": 3.5619492896960976, + "grad_norm": 5.574308064625232, + "learning_rate": 2.164411902335011e-06, + "loss": 0.1874, + "step": 4952 + }, + { + "epoch": 3.562668584786909, + "grad_norm": 0.608369724425219, + "learning_rate": 2.1641068386521846e-06, + "loss": 0.0018, + "step": 4953 + }, + { + "epoch": 3.5633878798777197, + "grad_norm": 6.274084201836541, + "learning_rate": 2.1638017407998963e-06, + "loss": 0.1306, + "step": 4954 + }, + { + "epoch": 3.5641071749685307, + "grad_norm": 2.6271276905631185, + "learning_rate": 2.163496608793843e-06, + "loss": 0.0718, + "step": 4955 + }, + { + "epoch": 3.5648264700593417, + "grad_norm": 2.7970882868535365, + "learning_rate": 2.1631914426497246e-06, + "loss": 0.1004, + "step": 4956 + }, + { + "epoch": 3.5655457651501528, + "grad_norm": 8.269943985785455, + "learning_rate": 2.162886242383242e-06, + "loss": 0.0712, + "step": 4957 + }, + { + "epoch": 3.566265060240964, + "grad_norm": 3.734004450995905, + "learning_rate": 2.162581008010099e-06, + "loss": 0.1203, + "step": 4958 + }, + { + "epoch": 3.566984355331775, + "grad_norm": 1.56706991384246, + "learning_rate": 2.1622757395459994e-06, + "loss": 0.0183, + "step": 4959 + }, + { + "epoch": 3.567703650422586, + "grad_norm": 2.4348180965066692, + "learning_rate": 2.161970437006651e-06, + "loss": 0.0463, + "step": 4960 + }, + { + "epoch": 3.568422945513397, + "grad_norm": 1.5796661406566637, + "learning_rate": 2.161665100407761e-06, + "loss": 0.0039, + "step": 4961 + }, + { + "epoch": 3.569142240604208, + "grad_norm": 3.160879949138244, + "learning_rate": 2.1613597297650407e-06, + "loss": 0.0932, + "step": 4962 + }, + { + "epoch": 3.569861535695019, + "grad_norm": 2.209695827724106, + "learning_rate": 2.161054325094201e-06, + "loss": 0.0591, + "step": 4963 + }, + { + "epoch": 3.57058083078583, + "grad_norm": 3.882947644563503, + "learning_rate": 2.1607488864109554e-06, + "loss": 0.0334, + "step": 4964 + }, + { + "epoch": 3.571300125876641, + "grad_norm": 5.619410526692964, + "learning_rate": 2.1604434137310203e-06, + "loss": 0.2277, + "step": 4965 + }, + { + "epoch": 3.572019420967452, + "grad_norm": 1.0372653389807698, + "learning_rate": 2.1601379070701117e-06, + "loss": 0.0129, + "step": 4966 + }, + { + "epoch": 3.5727387160582627, + "grad_norm": 4.3503687872688666, + "learning_rate": 2.1598323664439494e-06, + "loss": 0.1102, + "step": 4967 + }, + { + "epoch": 3.573458011149074, + "grad_norm": 2.7827456548223606, + "learning_rate": 2.1595267918682534e-06, + "loss": 0.039, + "step": 4968 + }, + { + "epoch": 3.5741773062398847, + "grad_norm": 1.2651431851387254, + "learning_rate": 2.159221183358746e-06, + "loss": 0.0104, + "step": 4969 + }, + { + "epoch": 3.5748966013306958, + "grad_norm": 2.592730312191458, + "learning_rate": 2.1589155409311516e-06, + "loss": 0.0731, + "step": 4970 + }, + { + "epoch": 3.575615896421507, + "grad_norm": 6.741021911897331, + "learning_rate": 2.158609864601196e-06, + "loss": 0.07, + "step": 4971 + }, + { + "epoch": 3.576335191512318, + "grad_norm": 3.6197214238346893, + "learning_rate": 2.158304154384607e-06, + "loss": 0.1352, + "step": 4972 + }, + { + "epoch": 3.577054486603129, + "grad_norm": 2.9266824049036297, + "learning_rate": 2.157998410297113e-06, + "loss": 0.0555, + "step": 4973 + }, + { + "epoch": 3.57777378169394, + "grad_norm": 2.539900770485992, + "learning_rate": 2.157692632354446e-06, + "loss": 0.0594, + "step": 4974 + }, + { + "epoch": 3.578493076784751, + "grad_norm": 5.19123338157212, + "learning_rate": 2.157386820572339e-06, + "loss": 0.1712, + "step": 4975 + }, + { + "epoch": 3.579212371875562, + "grad_norm": 4.18260445525137, + "learning_rate": 2.1570809749665254e-06, + "loss": 0.0818, + "step": 4976 + }, + { + "epoch": 3.579931666966373, + "grad_norm": 5.303174616590207, + "learning_rate": 2.156775095552742e-06, + "loss": 0.2442, + "step": 4977 + }, + { + "epoch": 3.580650962057184, + "grad_norm": 7.615785954899499, + "learning_rate": 2.1564691823467277e-06, + "loss": 0.1377, + "step": 4978 + }, + { + "epoch": 3.581370257147995, + "grad_norm": 5.916285664751357, + "learning_rate": 2.156163235364221e-06, + "loss": 0.095, + "step": 4979 + }, + { + "epoch": 3.582089552238806, + "grad_norm": 4.852688109404041, + "learning_rate": 2.1558572546209646e-06, + "loss": 0.055, + "step": 4980 + }, + { + "epoch": 3.582808847329617, + "grad_norm": 4.047262610326514, + "learning_rate": 2.1555512401327007e-06, + "loss": 0.1325, + "step": 4981 + }, + { + "epoch": 3.5835281424204277, + "grad_norm": 0.16520524597641148, + "learning_rate": 2.1552451919151745e-06, + "loss": 0.0005, + "step": 4982 + }, + { + "epoch": 3.584247437511239, + "grad_norm": 3.8031209495730964, + "learning_rate": 2.1549391099841335e-06, + "loss": 0.0448, + "step": 4983 + }, + { + "epoch": 3.58496673260205, + "grad_norm": 0.5955508347828341, + "learning_rate": 2.1546329943553254e-06, + "loss": 0.0046, + "step": 4984 + }, + { + "epoch": 3.5856860276928613, + "grad_norm": 1.8398059563277926, + "learning_rate": 2.1543268450445003e-06, + "loss": 0.0272, + "step": 4985 + }, + { + "epoch": 3.586405322783672, + "grad_norm": 0.07530066741538657, + "learning_rate": 2.154020662067411e-06, + "loss": 0.0003, + "step": 4986 + }, + { + "epoch": 3.587124617874483, + "grad_norm": 2.715835225097343, + "learning_rate": 2.1537144454398096e-06, + "loss": 0.0883, + "step": 4987 + }, + { + "epoch": 3.587843912965294, + "grad_norm": 2.6481001221907152, + "learning_rate": 2.153408195177453e-06, + "loss": 0.0738, + "step": 4988 + }, + { + "epoch": 3.588563208056105, + "grad_norm": 3.292607766905006, + "learning_rate": 2.153101911296098e-06, + "loss": 0.0175, + "step": 4989 + }, + { + "epoch": 3.589282503146916, + "grad_norm": 1.9499554566716863, + "learning_rate": 2.152795593811503e-06, + "loss": 0.0464, + "step": 4990 + }, + { + "epoch": 3.590001798237727, + "grad_norm": 3.3520262255954476, + "learning_rate": 2.1524892427394283e-06, + "loss": 0.115, + "step": 4991 + }, + { + "epoch": 3.590721093328538, + "grad_norm": 1.9539607069264007, + "learning_rate": 2.152182858095637e-06, + "loss": 0.0027, + "step": 4992 + }, + { + "epoch": 3.591440388419349, + "grad_norm": 2.1188639801532383, + "learning_rate": 2.1518764398958926e-06, + "loss": 0.0526, + "step": 4993 + }, + { + "epoch": 3.59215968351016, + "grad_norm": 3.3221023159324172, + "learning_rate": 2.1515699881559615e-06, + "loss": 0.1019, + "step": 4994 + }, + { + "epoch": 3.592878978600971, + "grad_norm": 2.556026840208387, + "learning_rate": 2.15126350289161e-06, + "loss": 0.0145, + "step": 4995 + }, + { + "epoch": 3.593598273691782, + "grad_norm": 1.9304543270871037, + "learning_rate": 2.1509569841186084e-06, + "loss": 0.004, + "step": 4996 + }, + { + "epoch": 3.594317568782593, + "grad_norm": 3.437969830642564, + "learning_rate": 2.1506504318527274e-06, + "loss": 0.0788, + "step": 4997 + }, + { + "epoch": 3.5950368638734043, + "grad_norm": 0.41351942765961686, + "learning_rate": 2.1503438461097395e-06, + "loss": 0.0028, + "step": 4998 + }, + { + "epoch": 3.595756158964215, + "grad_norm": 2.37687191271257, + "learning_rate": 2.150037226905419e-06, + "loss": 0.062, + "step": 4999 + }, + { + "epoch": 3.5964754540550263, + "grad_norm": 4.995834331651179, + "learning_rate": 2.149730574255542e-06, + "loss": 0.1319, + "step": 5000 + }, + { + "epoch": 3.597194749145837, + "grad_norm": 2.554887937365632, + "learning_rate": 2.1494238881758863e-06, + "loss": 0.0401, + "step": 5001 + }, + { + "epoch": 3.597914044236648, + "grad_norm": 4.5611851486889785, + "learning_rate": 2.1491171686822316e-06, + "loss": 0.1175, + "step": 5002 + }, + { + "epoch": 3.598633339327459, + "grad_norm": 3.4581285533523345, + "learning_rate": 2.1488104157903587e-06, + "loss": 0.0516, + "step": 5003 + }, + { + "epoch": 3.59935263441827, + "grad_norm": 1.684836550269939, + "learning_rate": 2.1485036295160516e-06, + "loss": 0.0532, + "step": 5004 + }, + { + "epoch": 3.600071929509081, + "grad_norm": 1.5195067605752752, + "learning_rate": 2.148196809875094e-06, + "loss": 0.0034, + "step": 5005 + }, + { + "epoch": 3.600791224599892, + "grad_norm": 4.557018678272835, + "learning_rate": 2.1478899568832726e-06, + "loss": 0.1834, + "step": 5006 + }, + { + "epoch": 3.601510519690703, + "grad_norm": 2.6720334225136932, + "learning_rate": 2.147583070556376e-06, + "loss": 0.098, + "step": 5007 + }, + { + "epoch": 3.602229814781514, + "grad_norm": 3.167974297149037, + "learning_rate": 2.1472761509101935e-06, + "loss": 0.1156, + "step": 5008 + }, + { + "epoch": 3.602949109872325, + "grad_norm": 4.87659190456884, + "learning_rate": 2.146969197960516e-06, + "loss": 0.1126, + "step": 5009 + }, + { + "epoch": 3.6036684049631362, + "grad_norm": 0.15945304510316505, + "learning_rate": 2.1466622117231387e-06, + "loss": 0.0002, + "step": 5010 + }, + { + "epoch": 3.6043877000539473, + "grad_norm": 7.785161091488893, + "learning_rate": 2.1463551922138546e-06, + "loss": 0.115, + "step": 5011 + }, + { + "epoch": 3.6051069951447583, + "grad_norm": 2.0182624697484304, + "learning_rate": 2.1460481394484617e-06, + "loss": 0.0455, + "step": 5012 + }, + { + "epoch": 3.6058262902355693, + "grad_norm": 3.1879505210590655, + "learning_rate": 2.145741053442758e-06, + "loss": 0.0817, + "step": 5013 + }, + { + "epoch": 3.60654558532638, + "grad_norm": 2.7135268570706104, + "learning_rate": 2.145433934212543e-06, + "loss": 0.089, + "step": 5014 + }, + { + "epoch": 3.6072648804171914, + "grad_norm": 1.4896050418096392, + "learning_rate": 2.14512678177362e-06, + "loss": 0.0343, + "step": 5015 + }, + { + "epoch": 3.607984175508002, + "grad_norm": 3.227574071633562, + "learning_rate": 2.14481959614179e-06, + "loss": 0.0381, + "step": 5016 + }, + { + "epoch": 3.608703470598813, + "grad_norm": 2.1118166961649107, + "learning_rate": 2.1445123773328614e-06, + "loss": 0.0497, + "step": 5017 + }, + { + "epoch": 3.609422765689624, + "grad_norm": 0.20400755583847602, + "learning_rate": 2.144205125362639e-06, + "loss": 0.0009, + "step": 5018 + }, + { + "epoch": 3.610142060780435, + "grad_norm": 3.7753667619398406, + "learning_rate": 2.143897840246932e-06, + "loss": 0.1096, + "step": 5019 + }, + { + "epoch": 3.610861355871246, + "grad_norm": 6.0750759133159535, + "learning_rate": 2.1435905220015503e-06, + "loss": 0.2467, + "step": 5020 + }, + { + "epoch": 3.611580650962057, + "grad_norm": 2.5581765560618455, + "learning_rate": 2.1432831706423072e-06, + "loss": 0.0769, + "step": 5021 + }, + { + "epoch": 3.612299946052868, + "grad_norm": 1.4776021753761226, + "learning_rate": 2.1429757861850154e-06, + "loss": 0.0262, + "step": 5022 + }, + { + "epoch": 3.6130192411436792, + "grad_norm": 5.250303681239855, + "learning_rate": 2.142668368645491e-06, + "loss": 0.201, + "step": 5023 + }, + { + "epoch": 3.6137385362344903, + "grad_norm": 2.246674209870273, + "learning_rate": 2.1423609180395504e-06, + "loss": 0.0444, + "step": 5024 + }, + { + "epoch": 3.6144578313253013, + "grad_norm": 0.03679294048234328, + "learning_rate": 2.142053434383013e-06, + "loss": 0.0002, + "step": 5025 + }, + { + "epoch": 3.6151771264161123, + "grad_norm": 1.156584350148476, + "learning_rate": 2.1417459176917e-06, + "loss": 0.0117, + "step": 5026 + }, + { + "epoch": 3.6158964215069234, + "grad_norm": 1.7444955310876278, + "learning_rate": 2.141438367981432e-06, + "loss": 0.0215, + "step": 5027 + }, + { + "epoch": 3.6166157165977344, + "grad_norm": 2.1461990112501614, + "learning_rate": 2.1411307852680346e-06, + "loss": 0.0577, + "step": 5028 + }, + { + "epoch": 3.617335011688545, + "grad_norm": 4.269135816401607, + "learning_rate": 2.140823169567333e-06, + "loss": 0.1267, + "step": 5029 + }, + { + "epoch": 3.6180543067793565, + "grad_norm": 3.9535567768827615, + "learning_rate": 2.140515520895154e-06, + "loss": 0.047, + "step": 5030 + }, + { + "epoch": 3.618773601870167, + "grad_norm": 0.3817138035862091, + "learning_rate": 2.1402078392673266e-06, + "loss": 0.0012, + "step": 5031 + }, + { + "epoch": 3.619492896960978, + "grad_norm": 2.771904206476021, + "learning_rate": 2.1399001246996827e-06, + "loss": 0.0523, + "step": 5032 + }, + { + "epoch": 3.620212192051789, + "grad_norm": 2.7772861798739097, + "learning_rate": 2.1395923772080537e-06, + "loss": 0.0316, + "step": 5033 + }, + { + "epoch": 3.6209314871426, + "grad_norm": 3.9682403625127485, + "learning_rate": 2.1392845968082744e-06, + "loss": 0.0849, + "step": 5034 + }, + { + "epoch": 3.621650782233411, + "grad_norm": 2.8699637421656594, + "learning_rate": 2.13897678351618e-06, + "loss": 0.0482, + "step": 5035 + }, + { + "epoch": 3.6223700773242222, + "grad_norm": 6.065286592490418, + "learning_rate": 2.138668937347609e-06, + "loss": 0.1879, + "step": 5036 + }, + { + "epoch": 3.6230893724150333, + "grad_norm": 1.8933844777533373, + "learning_rate": 2.1383610583184e-06, + "loss": 0.0467, + "step": 5037 + }, + { + "epoch": 3.6238086675058443, + "grad_norm": 1.8058415077642145, + "learning_rate": 2.1380531464443933e-06, + "loss": 0.0199, + "step": 5038 + }, + { + "epoch": 3.6245279625966553, + "grad_norm": 3.1702460845153513, + "learning_rate": 2.1377452017414328e-06, + "loss": 0.0454, + "step": 5039 + }, + { + "epoch": 3.6252472576874664, + "grad_norm": 5.536998004233153, + "learning_rate": 2.1374372242253616e-06, + "loss": 0.1595, + "step": 5040 + }, + { + "epoch": 3.6259665527782774, + "grad_norm": 2.1184662940079675, + "learning_rate": 2.1371292139120265e-06, + "loss": 0.0409, + "step": 5041 + }, + { + "epoch": 3.6266858478690884, + "grad_norm": 3.7234732768724768, + "learning_rate": 2.136821170817275e-06, + "loss": 0.1105, + "step": 5042 + }, + { + "epoch": 3.6274051429598995, + "grad_norm": 5.068450924290806, + "learning_rate": 2.136513094956956e-06, + "loss": 0.1947, + "step": 5043 + }, + { + "epoch": 3.62812443805071, + "grad_norm": 5.908121183707499, + "learning_rate": 2.1362049863469215e-06, + "loss": 0.1809, + "step": 5044 + }, + { + "epoch": 3.6288437331415215, + "grad_norm": 2.843669268531057, + "learning_rate": 2.135896845003023e-06, + "loss": 0.0891, + "step": 5045 + }, + { + "epoch": 3.629563028232332, + "grad_norm": 2.878276160251541, + "learning_rate": 2.135588670941116e-06, + "loss": 0.0117, + "step": 5046 + }, + { + "epoch": 3.630282323323143, + "grad_norm": 3.1293782810124378, + "learning_rate": 2.135280464177056e-06, + "loss": 0.079, + "step": 5047 + }, + { + "epoch": 3.631001618413954, + "grad_norm": 3.0355036375292634, + "learning_rate": 2.134972224726701e-06, + "loss": 0.0952, + "step": 5048 + }, + { + "epoch": 3.6317209135047652, + "grad_norm": 2.357408148081353, + "learning_rate": 2.13466395260591e-06, + "loss": 0.0762, + "step": 5049 + }, + { + "epoch": 3.6324402085955763, + "grad_norm": 2.068736019292497, + "learning_rate": 2.134355647830544e-06, + "loss": 0.0365, + "step": 5050 + }, + { + "epoch": 3.6331595036863873, + "grad_norm": 3.2957383956553215, + "learning_rate": 2.134047310416468e-06, + "loss": 0.0514, + "step": 5051 + }, + { + "epoch": 3.6338787987771983, + "grad_norm": 4.292341697178749, + "learning_rate": 2.133738940379544e-06, + "loss": 0.0894, + "step": 5052 + }, + { + "epoch": 3.6345980938680094, + "grad_norm": 2.0874692464735385, + "learning_rate": 2.1334305377356387e-06, + "loss": 0.0513, + "step": 5053 + }, + { + "epoch": 3.6353173889588204, + "grad_norm": 3.0707446252307333, + "learning_rate": 2.1331221025006205e-06, + "loss": 0.1344, + "step": 5054 + }, + { + "epoch": 3.6360366840496314, + "grad_norm": 3.6859882467735883, + "learning_rate": 2.1328136346903592e-06, + "loss": 0.0775, + "step": 5055 + }, + { + "epoch": 3.6367559791404425, + "grad_norm": 1.2391996318531586, + "learning_rate": 2.1325051343207245e-06, + "loss": 0.0144, + "step": 5056 + }, + { + "epoch": 3.6374752742312535, + "grad_norm": 2.267978385283935, + "learning_rate": 2.1321966014075914e-06, + "loss": 0.0585, + "step": 5057 + }, + { + "epoch": 3.6381945693220645, + "grad_norm": 1.9016381758977214, + "learning_rate": 2.1318880359668324e-06, + "loss": 0.0399, + "step": 5058 + }, + { + "epoch": 3.638913864412875, + "grad_norm": 1.8100073275630146, + "learning_rate": 2.1315794380143257e-06, + "loss": 0.0317, + "step": 5059 + }, + { + "epoch": 3.6396331595036866, + "grad_norm": 0.8834142755175648, + "learning_rate": 2.1312708075659475e-06, + "loss": 0.0182, + "step": 5060 + }, + { + "epoch": 3.640352454594497, + "grad_norm": 0.03975554489829409, + "learning_rate": 2.130962144637578e-06, + "loss": 0.0002, + "step": 5061 + }, + { + "epoch": 3.6410717496853087, + "grad_norm": 6.112836054786808, + "learning_rate": 2.130653449245099e-06, + "loss": 0.0854, + "step": 5062 + }, + { + "epoch": 3.6417910447761193, + "grad_norm": 2.62429047113617, + "learning_rate": 2.1303447214043927e-06, + "loss": 0.0629, + "step": 5063 + }, + { + "epoch": 3.6425103398669303, + "grad_norm": 7.453379603412968, + "learning_rate": 2.130035961131344e-06, + "loss": 0.1635, + "step": 5064 + }, + { + "epoch": 3.6432296349577413, + "grad_norm": 1.2230047795699879, + "learning_rate": 2.129727168441839e-06, + "loss": 0.0238, + "step": 5065 + }, + { + "epoch": 3.6439489300485524, + "grad_norm": 2.6156757908389237, + "learning_rate": 2.129418343351766e-06, + "loss": 0.0381, + "step": 5066 + }, + { + "epoch": 3.6446682251393634, + "grad_norm": 2.8925755046417847, + "learning_rate": 2.1291094858770146e-06, + "loss": 0.0809, + "step": 5067 + }, + { + "epoch": 3.6453875202301744, + "grad_norm": 5.346501760660919, + "learning_rate": 2.1288005960334755e-06, + "loss": 0.1329, + "step": 5068 + }, + { + "epoch": 3.6461068153209855, + "grad_norm": 4.564332802938965, + "learning_rate": 2.1284916738370416e-06, + "loss": 0.0675, + "step": 5069 + }, + { + "epoch": 3.6468261104117965, + "grad_norm": 1.800583176708661, + "learning_rate": 2.128182719303608e-06, + "loss": 0.0315, + "step": 5070 + }, + { + "epoch": 3.6475454055026075, + "grad_norm": 2.472898144423607, + "learning_rate": 2.1278737324490706e-06, + "loss": 0.005, + "step": 5071 + }, + { + "epoch": 3.6482647005934186, + "grad_norm": 5.035523360019861, + "learning_rate": 2.1275647132893274e-06, + "loss": 0.1217, + "step": 5072 + }, + { + "epoch": 3.6489839956842296, + "grad_norm": 3.245712214759344, + "learning_rate": 2.1272556618402786e-06, + "loss": 0.0177, + "step": 5073 + }, + { + "epoch": 3.64970329077504, + "grad_norm": 1.474472545499853, + "learning_rate": 2.1269465781178245e-06, + "loss": 0.032, + "step": 5074 + }, + { + "epoch": 3.6504225858658517, + "grad_norm": 2.443654910302049, + "learning_rate": 2.1266374621378684e-06, + "loss": 0.0565, + "step": 5075 + }, + { + "epoch": 3.6511418809566623, + "grad_norm": 5.445272861899874, + "learning_rate": 2.1263283139163147e-06, + "loss": 0.0353, + "step": 5076 + }, + { + "epoch": 3.6518611760474737, + "grad_norm": 18.27791403106358, + "learning_rate": 2.12601913346907e-06, + "loss": 0.2134, + "step": 5077 + }, + { + "epoch": 3.6525804711382843, + "grad_norm": 3.2615628019611145, + "learning_rate": 2.125709920812042e-06, + "loss": 0.0328, + "step": 5078 + }, + { + "epoch": 3.6532997662290954, + "grad_norm": 3.6525931764474633, + "learning_rate": 2.1254006759611393e-06, + "loss": 0.1042, + "step": 5079 + }, + { + "epoch": 3.6540190613199064, + "grad_norm": 7.326861984754964, + "learning_rate": 2.125091398932275e-06, + "loss": 0.2348, + "step": 5080 + }, + { + "epoch": 3.6547383564107174, + "grad_norm": 4.1127065645314715, + "learning_rate": 2.124782089741361e-06, + "loss": 0.0348, + "step": 5081 + }, + { + "epoch": 3.6554576515015285, + "grad_norm": 2.4096022484324613, + "learning_rate": 2.1244727484043113e-06, + "loss": 0.0261, + "step": 5082 + }, + { + "epoch": 3.6561769465923395, + "grad_norm": 2.470003253292898, + "learning_rate": 2.1241633749370423e-06, + "loss": 0.0611, + "step": 5083 + }, + { + "epoch": 3.6568962416831505, + "grad_norm": 1.8297116988176025, + "learning_rate": 2.123853969355472e-06, + "loss": 0.0424, + "step": 5084 + }, + { + "epoch": 3.6576155367739616, + "grad_norm": 2.75004048201188, + "learning_rate": 2.1235445316755206e-06, + "loss": 0.0714, + "step": 5085 + }, + { + "epoch": 3.6583348318647726, + "grad_norm": 3.795015417535723, + "learning_rate": 2.1232350619131077e-06, + "loss": 0.0797, + "step": 5086 + }, + { + "epoch": 3.6590541269555836, + "grad_norm": 5.850310835518141, + "learning_rate": 2.122925560084157e-06, + "loss": 0.0929, + "step": 5087 + }, + { + "epoch": 3.6597734220463947, + "grad_norm": 0.4254216322248615, + "learning_rate": 2.122616026204593e-06, + "loss": 0.001, + "step": 5088 + }, + { + "epoch": 3.6604927171372057, + "grad_norm": 4.928332619754812, + "learning_rate": 2.122306460290341e-06, + "loss": 0.1377, + "step": 5089 + }, + { + "epoch": 3.6612120122280167, + "grad_norm": 2.409714243688249, + "learning_rate": 2.1219968623573293e-06, + "loss": 0.0311, + "step": 5090 + }, + { + "epoch": 3.6619313073188273, + "grad_norm": 2.2013229324870953, + "learning_rate": 2.121687232421487e-06, + "loss": 0.0142, + "step": 5091 + }, + { + "epoch": 3.662650602409639, + "grad_norm": 3.2740210172382644, + "learning_rate": 2.1213775704987456e-06, + "loss": 0.0993, + "step": 5092 + }, + { + "epoch": 3.6633698975004494, + "grad_norm": 1.3067465465613768, + "learning_rate": 2.121067876605037e-06, + "loss": 0.0047, + "step": 5093 + }, + { + "epoch": 3.6640891925912604, + "grad_norm": 7.324741422601724, + "learning_rate": 2.1207581507562963e-06, + "loss": 0.2267, + "step": 5094 + }, + { + "epoch": 3.6648084876820715, + "grad_norm": 2.7937971399736425, + "learning_rate": 2.120448392968459e-06, + "loss": 0.0144, + "step": 5095 + }, + { + "epoch": 3.6655277827728825, + "grad_norm": 3.765742552780067, + "learning_rate": 2.120138603257463e-06, + "loss": 0.0616, + "step": 5096 + }, + { + "epoch": 3.6662470778636935, + "grad_norm": 0.6676401222294482, + "learning_rate": 2.1198287816392466e-06, + "loss": 0.0067, + "step": 5097 + }, + { + "epoch": 3.6669663729545046, + "grad_norm": 3.7461696312060986, + "learning_rate": 2.1195189281297517e-06, + "loss": 0.1244, + "step": 5098 + }, + { + "epoch": 3.6676856680453156, + "grad_norm": 3.625745419407408, + "learning_rate": 2.1192090427449205e-06, + "loss": 0.0905, + "step": 5099 + }, + { + "epoch": 3.6684049631361266, + "grad_norm": 1.0306841248559007, + "learning_rate": 2.1188991255006967e-06, + "loss": 0.0217, + "step": 5100 + }, + { + "epoch": 3.6691242582269377, + "grad_norm": 2.234598926939793, + "learning_rate": 2.1185891764130264e-06, + "loss": 0.0301, + "step": 5101 + }, + { + "epoch": 3.6698435533177487, + "grad_norm": 5.514466092586715, + "learning_rate": 2.1182791954978577e-06, + "loss": 0.225, + "step": 5102 + }, + { + "epoch": 3.6705628484085597, + "grad_norm": 5.3193334605753915, + "learning_rate": 2.1179691827711385e-06, + "loss": 0.1808, + "step": 5103 + }, + { + "epoch": 3.6712821434993708, + "grad_norm": 2.5949429456838917, + "learning_rate": 2.1176591382488204e-06, + "loss": 0.0937, + "step": 5104 + }, + { + "epoch": 3.672001438590182, + "grad_norm": 4.02480002536879, + "learning_rate": 2.117349061946855e-06, + "loss": 0.07, + "step": 5105 + }, + { + "epoch": 3.6727207336809924, + "grad_norm": 2.3478627622851582, + "learning_rate": 2.1170389538811967e-06, + "loss": 0.0678, + "step": 5106 + }, + { + "epoch": 3.673440028771804, + "grad_norm": 2.3359022253614508, + "learning_rate": 2.1167288140678013e-06, + "loss": 0.0045, + "step": 5107 + }, + { + "epoch": 3.6741593238626145, + "grad_norm": 1.828380986283715, + "learning_rate": 2.116418642522625e-06, + "loss": 0.0257, + "step": 5108 + }, + { + "epoch": 3.6748786189534255, + "grad_norm": 2.470067438630471, + "learning_rate": 2.1161084392616283e-06, + "loss": 0.0144, + "step": 5109 + }, + { + "epoch": 3.6755979140442365, + "grad_norm": 4.2639877718137855, + "learning_rate": 2.115798204300771e-06, + "loss": 0.0212, + "step": 5110 + }, + { + "epoch": 3.6763172091350476, + "grad_norm": 3.707357562897044, + "learning_rate": 2.1154879376560144e-06, + "loss": 0.0654, + "step": 5111 + }, + { + "epoch": 3.6770365042258586, + "grad_norm": 3.553986127091118, + "learning_rate": 2.115177639343323e-06, + "loss": 0.0921, + "step": 5112 + }, + { + "epoch": 3.6777557993166696, + "grad_norm": 4.53628412198612, + "learning_rate": 2.114867309378663e-06, + "loss": 0.2018, + "step": 5113 + }, + { + "epoch": 3.6784750944074807, + "grad_norm": 2.666246702829319, + "learning_rate": 2.1145569477779996e-06, + "loss": 0.0693, + "step": 5114 + }, + { + "epoch": 3.6791943894982917, + "grad_norm": 4.243781454028782, + "learning_rate": 2.114246554557303e-06, + "loss": 0.1438, + "step": 5115 + }, + { + "epoch": 3.6799136845891027, + "grad_norm": 5.841872987474929, + "learning_rate": 2.1139361297325423e-06, + "loss": 0.2029, + "step": 5116 + }, + { + "epoch": 3.6806329796799138, + "grad_norm": 2.712189480923726, + "learning_rate": 2.113625673319691e-06, + "loss": 0.0067, + "step": 5117 + }, + { + "epoch": 3.681352274770725, + "grad_norm": 1.91619702060975, + "learning_rate": 2.113315185334721e-06, + "loss": 0.0656, + "step": 5118 + }, + { + "epoch": 3.682071569861536, + "grad_norm": 2.4852180960227606, + "learning_rate": 2.113004665793608e-06, + "loss": 0.0554, + "step": 5119 + }, + { + "epoch": 3.682790864952347, + "grad_norm": 3.1048833995156784, + "learning_rate": 2.1126941147123286e-06, + "loss": 0.0578, + "step": 5120 + }, + { + "epoch": 3.6835101600431575, + "grad_norm": 4.414152549107135, + "learning_rate": 2.1123835321068615e-06, + "loss": 0.0464, + "step": 5121 + }, + { + "epoch": 3.684229455133969, + "grad_norm": 4.033353105913917, + "learning_rate": 2.112072917993187e-06, + "loss": 0.1702, + "step": 5122 + }, + { + "epoch": 3.6849487502247795, + "grad_norm": 2.73414727855693, + "learning_rate": 2.1117622723872865e-06, + "loss": 0.046, + "step": 5123 + }, + { + "epoch": 3.6856680453155906, + "grad_norm": 2.7702349392242978, + "learning_rate": 2.1114515953051435e-06, + "loss": 0.0139, + "step": 5124 + }, + { + "epoch": 3.6863873404064016, + "grad_norm": 0.031200235765421858, + "learning_rate": 2.111140886762742e-06, + "loss": 0.0001, + "step": 5125 + }, + { + "epoch": 3.6871066354972126, + "grad_norm": 3.999043592601153, + "learning_rate": 2.110830146776069e-06, + "loss": 0.0134, + "step": 5126 + }, + { + "epoch": 3.6878259305880237, + "grad_norm": 2.3042475624284964, + "learning_rate": 2.1105193753611126e-06, + "loss": 0.0531, + "step": 5127 + }, + { + "epoch": 3.6885452256788347, + "grad_norm": 2.1452971758321664, + "learning_rate": 2.1102085725338627e-06, + "loss": 0.0025, + "step": 5128 + }, + { + "epoch": 3.6892645207696457, + "grad_norm": 2.6507027925970896, + "learning_rate": 2.1098977383103105e-06, + "loss": 0.056, + "step": 5129 + }, + { + "epoch": 3.6899838158604568, + "grad_norm": 2.595492916842733, + "learning_rate": 2.1095868727064495e-06, + "loss": 0.0685, + "step": 5130 + }, + { + "epoch": 3.690703110951268, + "grad_norm": 4.450288469638079, + "learning_rate": 2.1092759757382736e-06, + "loss": 0.0994, + "step": 5131 + }, + { + "epoch": 3.691422406042079, + "grad_norm": 1.7770632142399136, + "learning_rate": 2.1089650474217788e-06, + "loss": 0.0339, + "step": 5132 + }, + { + "epoch": 3.69214170113289, + "grad_norm": 0.9785526349565402, + "learning_rate": 2.108654087772964e-06, + "loss": 0.0204, + "step": 5133 + }, + { + "epoch": 3.692860996223701, + "grad_norm": 3.3670666654978554, + "learning_rate": 2.108343096807827e-06, + "loss": 0.0752, + "step": 5134 + }, + { + "epoch": 3.693580291314512, + "grad_norm": 4.500811547073495, + "learning_rate": 2.1080320745423705e-06, + "loss": 0.1927, + "step": 5135 + }, + { + "epoch": 3.6942995864053225, + "grad_norm": 2.762228856339184, + "learning_rate": 2.107721020992596e-06, + "loss": 0.0129, + "step": 5136 + }, + { + "epoch": 3.695018881496134, + "grad_norm": 1.4247061858617012, + "learning_rate": 2.1074099361745082e-06, + "loss": 0.0332, + "step": 5137 + }, + { + "epoch": 3.6957381765869446, + "grad_norm": 1.1834887241034813, + "learning_rate": 2.107098820104113e-06, + "loss": 0.0078, + "step": 5138 + }, + { + "epoch": 3.696457471677756, + "grad_norm": 4.763546128639781, + "learning_rate": 2.106787672797418e-06, + "loss": 0.1476, + "step": 5139 + }, + { + "epoch": 3.6971767667685667, + "grad_norm": 3.451807337978096, + "learning_rate": 2.106476494270432e-06, + "loss": 0.1012, + "step": 5140 + }, + { + "epoch": 3.6978960618593777, + "grad_norm": 0.2822145731126379, + "learning_rate": 2.1061652845391654e-06, + "loss": 0.0008, + "step": 5141 + }, + { + "epoch": 3.6986153569501887, + "grad_norm": 3.2102349442426252, + "learning_rate": 2.105854043619631e-06, + "loss": 0.0479, + "step": 5142 + }, + { + "epoch": 3.6993346520409998, + "grad_norm": 3.644311936718463, + "learning_rate": 2.1055427715278424e-06, + "loss": 0.1128, + "step": 5143 + }, + { + "epoch": 3.700053947131811, + "grad_norm": 6.513098915626881, + "learning_rate": 2.1052314682798156e-06, + "loss": 0.1327, + "step": 5144 + }, + { + "epoch": 3.700773242222622, + "grad_norm": 2.389927036170801, + "learning_rate": 2.1049201338915666e-06, + "loss": 0.0439, + "step": 5145 + }, + { + "epoch": 3.701492537313433, + "grad_norm": 4.4598375656633005, + "learning_rate": 2.104608768379116e-06, + "loss": 0.0954, + "step": 5146 + }, + { + "epoch": 3.702211832404244, + "grad_norm": 3.8901237785033347, + "learning_rate": 2.104297371758482e-06, + "loss": 0.172, + "step": 5147 + }, + { + "epoch": 3.702931127495055, + "grad_norm": 2.790083893632643, + "learning_rate": 2.103985944045687e-06, + "loss": 0.0077, + "step": 5148 + }, + { + "epoch": 3.703650422585866, + "grad_norm": 3.451109628738029, + "learning_rate": 2.103674485256756e-06, + "loss": 0.0867, + "step": 5149 + }, + { + "epoch": 3.704369717676677, + "grad_norm": 2.067584370916931, + "learning_rate": 2.1033629954077123e-06, + "loss": 0.022, + "step": 5150 + }, + { + "epoch": 3.7050890127674876, + "grad_norm": 1.2691799661080765, + "learning_rate": 2.1030514745145833e-06, + "loss": 0.0304, + "step": 5151 + }, + { + "epoch": 3.705808307858299, + "grad_norm": 4.187028379906094, + "learning_rate": 2.102739922593398e-06, + "loss": 0.1219, + "step": 5152 + }, + { + "epoch": 3.7065276029491097, + "grad_norm": 5.511889313128873, + "learning_rate": 2.102428339660185e-06, + "loss": 0.1158, + "step": 5153 + }, + { + "epoch": 3.707246898039921, + "grad_norm": 3.1347274468802944, + "learning_rate": 2.1021167257309765e-06, + "loss": 0.0468, + "step": 5154 + }, + { + "epoch": 3.7079661931307317, + "grad_norm": 4.699879995024807, + "learning_rate": 2.1018050808218055e-06, + "loss": 0.1226, + "step": 5155 + }, + { + "epoch": 3.7086854882215428, + "grad_norm": 2.4084748172776593, + "learning_rate": 2.101493404948707e-06, + "loss": 0.0121, + "step": 5156 + }, + { + "epoch": 3.709404783312354, + "grad_norm": 3.175462990620132, + "learning_rate": 2.1011816981277163e-06, + "loss": 0.1201, + "step": 5157 + }, + { + "epoch": 3.710124078403165, + "grad_norm": 2.8395294186860323, + "learning_rate": 2.1008699603748717e-06, + "loss": 0.066, + "step": 5158 + }, + { + "epoch": 3.710843373493976, + "grad_norm": 1.6970417131473838, + "learning_rate": 2.1005581917062136e-06, + "loss": 0.0041, + "step": 5159 + }, + { + "epoch": 3.711562668584787, + "grad_norm": 2.3714744169434216, + "learning_rate": 2.1002463921377817e-06, + "loss": 0.0376, + "step": 5160 + }, + { + "epoch": 3.712281963675598, + "grad_norm": 3.7528971280210985, + "learning_rate": 2.09993456168562e-06, + "loss": 0.0739, + "step": 5161 + }, + { + "epoch": 3.713001258766409, + "grad_norm": 3.8360596449226922, + "learning_rate": 2.0996227003657712e-06, + "loss": 0.1256, + "step": 5162 + }, + { + "epoch": 3.71372055385722, + "grad_norm": 3.0365301706057792, + "learning_rate": 2.099310808194282e-06, + "loss": 0.0673, + "step": 5163 + }, + { + "epoch": 3.714439848948031, + "grad_norm": 0.16575540323375307, + "learning_rate": 2.0989988851872e-06, + "loss": 0.0004, + "step": 5164 + }, + { + "epoch": 3.715159144038842, + "grad_norm": 2.2787357260955146, + "learning_rate": 2.098686931360574e-06, + "loss": 0.0263, + "step": 5165 + }, + { + "epoch": 3.715878439129653, + "grad_norm": 3.845006067713128, + "learning_rate": 2.0983749467304537e-06, + "loss": 0.0283, + "step": 5166 + }, + { + "epoch": 3.716597734220464, + "grad_norm": 4.718651767128302, + "learning_rate": 2.0980629313128925e-06, + "loss": 0.069, + "step": 5167 + }, + { + "epoch": 3.7173170293112747, + "grad_norm": 1.7106295913986218, + "learning_rate": 2.0977508851239435e-06, + "loss": 0.0216, + "step": 5168 + }, + { + "epoch": 3.718036324402086, + "grad_norm": 2.013342105110612, + "learning_rate": 2.0974388081796622e-06, + "loss": 0.0448, + "step": 5169 + }, + { + "epoch": 3.718755619492897, + "grad_norm": 3.2961979886179464, + "learning_rate": 2.097126700496105e-06, + "loss": 0.0992, + "step": 5170 + }, + { + "epoch": 3.719474914583708, + "grad_norm": 1.9951264194499858, + "learning_rate": 2.096814562089331e-06, + "loss": 0.0342, + "step": 5171 + }, + { + "epoch": 3.720194209674519, + "grad_norm": 0.782197580527653, + "learning_rate": 2.0965023929754007e-06, + "loss": 0.0075, + "step": 5172 + }, + { + "epoch": 3.72091350476533, + "grad_norm": 4.582156414153276, + "learning_rate": 2.0961901931703746e-06, + "loss": 0.1843, + "step": 5173 + }, + { + "epoch": 3.721632799856141, + "grad_norm": 3.5663933555041334, + "learning_rate": 2.0958779626903166e-06, + "loss": 0.033, + "step": 5174 + }, + { + "epoch": 3.722352094946952, + "grad_norm": 2.1051478927696987, + "learning_rate": 2.0955657015512917e-06, + "loss": 0.0843, + "step": 5175 + }, + { + "epoch": 3.723071390037763, + "grad_norm": 2.423299283468293, + "learning_rate": 2.0952534097693655e-06, + "loss": 0.0487, + "step": 5176 + }, + { + "epoch": 3.723790685128574, + "grad_norm": 2.9783034641711548, + "learning_rate": 2.094941087360607e-06, + "loss": 0.1047, + "step": 5177 + }, + { + "epoch": 3.724509980219385, + "grad_norm": 2.239848591949911, + "learning_rate": 2.0946287343410846e-06, + "loss": 0.0217, + "step": 5178 + }, + { + "epoch": 3.725229275310196, + "grad_norm": 0.9193386739591862, + "learning_rate": 2.09431635072687e-06, + "loss": 0.0172, + "step": 5179 + }, + { + "epoch": 3.725948570401007, + "grad_norm": 0.6736888221367154, + "learning_rate": 2.094003936534036e-06, + "loss": 0.0044, + "step": 5180 + }, + { + "epoch": 3.726667865491818, + "grad_norm": 4.443654634398247, + "learning_rate": 2.0936914917786568e-06, + "loss": 0.101, + "step": 5181 + }, + { + "epoch": 3.727387160582629, + "grad_norm": 4.824209288793332, + "learning_rate": 2.0933790164768082e-06, + "loss": 0.2048, + "step": 5182 + }, + { + "epoch": 3.72810645567344, + "grad_norm": 3.0848871883178153, + "learning_rate": 2.0930665106445677e-06, + "loss": 0.047, + "step": 5183 + }, + { + "epoch": 3.7288257507642513, + "grad_norm": 3.568180254522311, + "learning_rate": 2.092753974298014e-06, + "loss": 0.0834, + "step": 5184 + }, + { + "epoch": 3.729545045855062, + "grad_norm": 4.776474450663589, + "learning_rate": 2.0924414074532275e-06, + "loss": 0.0635, + "step": 5185 + }, + { + "epoch": 3.730264340945873, + "grad_norm": 2.04642351549561, + "learning_rate": 2.092128810126291e-06, + "loss": 0.0462, + "step": 5186 + }, + { + "epoch": 3.730983636036684, + "grad_norm": 1.81495904949192, + "learning_rate": 2.091816182333287e-06, + "loss": 0.0436, + "step": 5187 + }, + { + "epoch": 3.731702931127495, + "grad_norm": 4.635764694216081, + "learning_rate": 2.0915035240903025e-06, + "loss": 0.1513, + "step": 5188 + }, + { + "epoch": 3.732422226218306, + "grad_norm": 3.6382693666930623, + "learning_rate": 2.091190835413423e-06, + "loss": 0.107, + "step": 5189 + }, + { + "epoch": 3.733141521309117, + "grad_norm": 2.6517675400411695, + "learning_rate": 2.0908781163187372e-06, + "loss": 0.1039, + "step": 5190 + }, + { + "epoch": 3.733860816399928, + "grad_norm": 3.6156606260866155, + "learning_rate": 2.090565366822335e-06, + "loss": 0.1108, + "step": 5191 + }, + { + "epoch": 3.734580111490739, + "grad_norm": 4.405353556171796, + "learning_rate": 2.0902525869403077e-06, + "loss": 0.1956, + "step": 5192 + }, + { + "epoch": 3.73529940658155, + "grad_norm": 1.3933889703772533, + "learning_rate": 2.089939776688749e-06, + "loss": 0.0369, + "step": 5193 + }, + { + "epoch": 3.736018701672361, + "grad_norm": 0.032901824417938916, + "learning_rate": 2.089626936083753e-06, + "loss": 0.0001, + "step": 5194 + }, + { + "epoch": 3.736737996763172, + "grad_norm": 4.167542276692032, + "learning_rate": 2.089314065141416e-06, + "loss": 0.1297, + "step": 5195 + }, + { + "epoch": 3.7374572918539832, + "grad_norm": 2.932182240143979, + "learning_rate": 2.089001163877836e-06, + "loss": 0.0439, + "step": 5196 + }, + { + "epoch": 3.7381765869447943, + "grad_norm": 3.090971964231849, + "learning_rate": 2.0886882323091125e-06, + "loss": 0.0997, + "step": 5197 + }, + { + "epoch": 3.738895882035605, + "grad_norm": 1.7147360095501911, + "learning_rate": 2.088375270451346e-06, + "loss": 0.0304, + "step": 5198 + }, + { + "epoch": 3.7396151771264163, + "grad_norm": 2.9226171996797894, + "learning_rate": 2.0880622783206385e-06, + "loss": 0.0917, + "step": 5199 + }, + { + "epoch": 3.740334472217227, + "grad_norm": 5.88042559531114, + "learning_rate": 2.087749255933094e-06, + "loss": 0.1794, + "step": 5200 + }, + { + "epoch": 3.741053767308038, + "grad_norm": 3.7441550360688627, + "learning_rate": 2.0874362033048195e-06, + "loss": 0.0888, + "step": 5201 + }, + { + "epoch": 3.741773062398849, + "grad_norm": 3.501745767483795, + "learning_rate": 2.0871231204519208e-06, + "loss": 0.0673, + "step": 5202 + }, + { + "epoch": 3.74249235748966, + "grad_norm": 2.942784818551671, + "learning_rate": 2.086810007390507e-06, + "loss": 0.095, + "step": 5203 + }, + { + "epoch": 3.743211652580471, + "grad_norm": 6.171277190281516, + "learning_rate": 2.086496864136688e-06, + "loss": 0.0687, + "step": 5204 + }, + { + "epoch": 3.743930947671282, + "grad_norm": 6.423482651073675, + "learning_rate": 2.086183690706576e-06, + "loss": 0.1129, + "step": 5205 + }, + { + "epoch": 3.744650242762093, + "grad_norm": 3.4018184778555463, + "learning_rate": 2.085870487116284e-06, + "loss": 0.0619, + "step": 5206 + }, + { + "epoch": 3.745369537852904, + "grad_norm": 4.797120068799259, + "learning_rate": 2.085557253381927e-06, + "loss": 0.115, + "step": 5207 + }, + { + "epoch": 3.746088832943715, + "grad_norm": 3.6629677174012247, + "learning_rate": 2.085243989519621e-06, + "loss": 0.1142, + "step": 5208 + }, + { + "epoch": 3.7468081280345262, + "grad_norm": 0.04052049432681643, + "learning_rate": 2.084930695545485e-06, + "loss": 0.0003, + "step": 5209 + }, + { + "epoch": 3.7475274231253373, + "grad_norm": 3.849657883550725, + "learning_rate": 2.0846173714756373e-06, + "loss": 0.091, + "step": 5210 + }, + { + "epoch": 3.7482467182161483, + "grad_norm": 3.0339294390051754, + "learning_rate": 2.0843040173262e-06, + "loss": 0.1007, + "step": 5211 + }, + { + "epoch": 3.7489660133069593, + "grad_norm": 0.6413097824513887, + "learning_rate": 2.0839906331132954e-06, + "loss": 0.0021, + "step": 5212 + }, + { + "epoch": 3.74968530839777, + "grad_norm": 4.023263873772576, + "learning_rate": 2.083677218853047e-06, + "loss": 0.1586, + "step": 5213 + }, + { + "epoch": 3.7504046034885814, + "grad_norm": 2.911260974033768, + "learning_rate": 2.083363774561582e-06, + "loss": 0.0698, + "step": 5214 + }, + { + "epoch": 3.751123898579392, + "grad_norm": 1.2044747503267117, + "learning_rate": 2.0830503002550257e-06, + "loss": 0.0145, + "step": 5215 + }, + { + "epoch": 3.7518431936702035, + "grad_norm": 3.1664745922443225, + "learning_rate": 2.0827367959495084e-06, + "loss": 0.1724, + "step": 5216 + }, + { + "epoch": 3.752562488761014, + "grad_norm": 3.4955323529891187, + "learning_rate": 2.08242326166116e-06, + "loss": 0.1366, + "step": 5217 + }, + { + "epoch": 3.753281783851825, + "grad_norm": 5.58776890742118, + "learning_rate": 2.0821096974061128e-06, + "loss": 0.1341, + "step": 5218 + }, + { + "epoch": 3.754001078942636, + "grad_norm": 11.169440037430569, + "learning_rate": 2.081796103200499e-06, + "loss": 0.1631, + "step": 5219 + }, + { + "epoch": 3.754720374033447, + "grad_norm": 2.8879365226452913, + "learning_rate": 2.0814824790604553e-06, + "loss": 0.0372, + "step": 5220 + }, + { + "epoch": 3.755439669124258, + "grad_norm": 3.330142697290246, + "learning_rate": 2.0811688250021163e-06, + "loss": 0.099, + "step": 5221 + }, + { + "epoch": 3.7561589642150692, + "grad_norm": 1.9990274385949307, + "learning_rate": 2.0808551410416216e-06, + "loss": 0.0567, + "step": 5222 + }, + { + "epoch": 3.7568782593058803, + "grad_norm": 2.5618147904146453, + "learning_rate": 2.0805414271951097e-06, + "loss": 0.0768, + "step": 5223 + }, + { + "epoch": 3.7575975543966913, + "grad_norm": 3.542606890653743, + "learning_rate": 2.080227683478723e-06, + "loss": 0.0799, + "step": 5224 + }, + { + "epoch": 3.7583168494875023, + "grad_norm": 0.6588332004188694, + "learning_rate": 2.079913909908603e-06, + "loss": 0.0059, + "step": 5225 + }, + { + "epoch": 3.7590361445783134, + "grad_norm": 2.0853412016276285, + "learning_rate": 2.0796001065008948e-06, + "loss": 0.0582, + "step": 5226 + }, + { + "epoch": 3.7597554396691244, + "grad_norm": 4.135888249528709, + "learning_rate": 2.0792862732717432e-06, + "loss": 0.1204, + "step": 5227 + }, + { + "epoch": 3.760474734759935, + "grad_norm": 1.5849517880747863, + "learning_rate": 2.0789724102372963e-06, + "loss": 0.0267, + "step": 5228 + }, + { + "epoch": 3.7611940298507465, + "grad_norm": 1.8680775143799908, + "learning_rate": 2.0786585174137018e-06, + "loss": 0.0239, + "step": 5229 + }, + { + "epoch": 3.761913324941557, + "grad_norm": 3.3723608322716476, + "learning_rate": 2.0783445948171114e-06, + "loss": 0.1304, + "step": 5230 + }, + { + "epoch": 3.7626326200323685, + "grad_norm": 1.2009912293871654, + "learning_rate": 2.0780306424636762e-06, + "loss": 0.0024, + "step": 5231 + }, + { + "epoch": 3.763351915123179, + "grad_norm": 5.656541545358141, + "learning_rate": 2.0777166603695498e-06, + "loss": 0.2063, + "step": 5232 + }, + { + "epoch": 3.76407121021399, + "grad_norm": 0.9596567333089107, + "learning_rate": 2.077402648550887e-06, + "loss": 0.0103, + "step": 5233 + }, + { + "epoch": 3.764790505304801, + "grad_norm": 3.80423249572908, + "learning_rate": 2.0770886070238442e-06, + "loss": 0.1748, + "step": 5234 + }, + { + "epoch": 3.7655098003956122, + "grad_norm": 2.1117314850517066, + "learning_rate": 2.07677453580458e-06, + "loss": 0.0706, + "step": 5235 + }, + { + "epoch": 3.7662290954864233, + "grad_norm": 2.60935859203438, + "learning_rate": 2.076460434909253e-06, + "loss": 0.038, + "step": 5236 + }, + { + "epoch": 3.7669483905772343, + "grad_norm": 4.309057290481498, + "learning_rate": 2.0761463043540248e-06, + "loss": 0.0937, + "step": 5237 + }, + { + "epoch": 3.7676676856680453, + "grad_norm": 31.43649416041518, + "learning_rate": 2.0758321441550577e-06, + "loss": 0.0971, + "step": 5238 + }, + { + "epoch": 3.7683869807588564, + "grad_norm": 1.3389849349355012, + "learning_rate": 2.0755179543285158e-06, + "loss": 0.022, + "step": 5239 + }, + { + "epoch": 3.7691062758496674, + "grad_norm": 3.4999800319234056, + "learning_rate": 2.0752037348905656e-06, + "loss": 0.0836, + "step": 5240 + }, + { + "epoch": 3.7698255709404784, + "grad_norm": 1.5182042354175982, + "learning_rate": 2.0748894858573732e-06, + "loss": 0.0331, + "step": 5241 + }, + { + "epoch": 3.7705448660312895, + "grad_norm": 3.5742965318071853, + "learning_rate": 2.074575207245107e-06, + "loss": 0.0696, + "step": 5242 + }, + { + "epoch": 3.7712641611221005, + "grad_norm": 5.052415415055673, + "learning_rate": 2.074260899069938e-06, + "loss": 0.1743, + "step": 5243 + }, + { + "epoch": 3.7719834562129115, + "grad_norm": 1.2993138110278803, + "learning_rate": 2.0739465613480378e-06, + "loss": 0.0257, + "step": 5244 + }, + { + "epoch": 3.772702751303722, + "grad_norm": 3.5646729494132843, + "learning_rate": 2.0736321940955797e-06, + "loss": 0.1475, + "step": 5245 + }, + { + "epoch": 3.7734220463945336, + "grad_norm": 4.072065308053058, + "learning_rate": 2.0733177973287383e-06, + "loss": 0.0821, + "step": 5246 + }, + { + "epoch": 3.774141341485344, + "grad_norm": 4.311065843911253, + "learning_rate": 2.0730033710636896e-06, + "loss": 0.1231, + "step": 5247 + }, + { + "epoch": 3.7748606365761552, + "grad_norm": 3.0450333902683244, + "learning_rate": 2.0726889153166114e-06, + "loss": 0.0162, + "step": 5248 + }, + { + "epoch": 3.7755799316669663, + "grad_norm": 4.946744459512056, + "learning_rate": 2.0723744301036832e-06, + "loss": 0.0773, + "step": 5249 + }, + { + "epoch": 3.7762992267577773, + "grad_norm": 0.2529888062583041, + "learning_rate": 2.072059915441086e-06, + "loss": 0.0006, + "step": 5250 + }, + { + "epoch": 3.7770185218485883, + "grad_norm": 2.722506216204953, + "learning_rate": 2.071745371345002e-06, + "loss": 0.0686, + "step": 5251 + }, + { + "epoch": 3.7777378169393994, + "grad_norm": 3.33485629260362, + "learning_rate": 2.0714307978316144e-06, + "loss": 0.0757, + "step": 5252 + }, + { + "epoch": 3.7784571120302104, + "grad_norm": 0.6694023310202385, + "learning_rate": 2.0711161949171097e-06, + "loss": 0.0016, + "step": 5253 + }, + { + "epoch": 3.7791764071210214, + "grad_norm": 2.6783755217659486, + "learning_rate": 2.070801562617674e-06, + "loss": 0.0677, + "step": 5254 + }, + { + "epoch": 3.7798957022118325, + "grad_norm": 2.6354171571346043, + "learning_rate": 2.0704869009494962e-06, + "loss": 0.0717, + "step": 5255 + }, + { + "epoch": 3.7806149973026435, + "grad_norm": 4.92257818882097, + "learning_rate": 2.0701722099287657e-06, + "loss": 0.2001, + "step": 5256 + }, + { + "epoch": 3.7813342923934545, + "grad_norm": 3.256552349013167, + "learning_rate": 2.0698574895716737e-06, + "loss": 0.1044, + "step": 5257 + }, + { + "epoch": 3.7820535874842656, + "grad_norm": 1.3896606300207637, + "learning_rate": 2.0695427398944142e-06, + "loss": 0.03, + "step": 5258 + }, + { + "epoch": 3.7827728825750766, + "grad_norm": 0.7362214818461137, + "learning_rate": 2.0692279609131805e-06, + "loss": 0.0019, + "step": 5259 + }, + { + "epoch": 3.783492177665887, + "grad_norm": 4.08952445354401, + "learning_rate": 2.0689131526441686e-06, + "loss": 0.1017, + "step": 5260 + }, + { + "epoch": 3.7842114727566987, + "grad_norm": 4.631936122929661, + "learning_rate": 2.0685983151035768e-06, + "loss": 0.169, + "step": 5261 + }, + { + "epoch": 3.7849307678475093, + "grad_norm": 0.07939738765873607, + "learning_rate": 2.0682834483076037e-06, + "loss": 0.0002, + "step": 5262 + }, + { + "epoch": 3.7856500629383203, + "grad_norm": 0.09683627626031414, + "learning_rate": 2.067968552272449e-06, + "loss": 0.0002, + "step": 5263 + }, + { + "epoch": 3.7863693580291313, + "grad_norm": 2.2237578961442233, + "learning_rate": 2.0676536270143163e-06, + "loss": 0.0538, + "step": 5264 + }, + { + "epoch": 3.7870886531199424, + "grad_norm": 4.123619171325498, + "learning_rate": 2.067338672549407e-06, + "loss": 0.1262, + "step": 5265 + }, + { + "epoch": 3.7878079482107534, + "grad_norm": 3.6939943363724614, + "learning_rate": 2.0670236888939275e-06, + "loss": 0.0792, + "step": 5266 + }, + { + "epoch": 3.7885272433015644, + "grad_norm": 3.45674453419902, + "learning_rate": 2.0667086760640837e-06, + "loss": 0.0526, + "step": 5267 + }, + { + "epoch": 3.7892465383923755, + "grad_norm": 0.29651674477262363, + "learning_rate": 2.0663936340760837e-06, + "loss": 0.0006, + "step": 5268 + }, + { + "epoch": 3.7899658334831865, + "grad_norm": 1.563908682081413, + "learning_rate": 2.066078562946137e-06, + "loss": 0.0364, + "step": 5269 + }, + { + "epoch": 3.7906851285739975, + "grad_norm": 3.1238726741275316, + "learning_rate": 2.0657634626904547e-06, + "loss": 0.1126, + "step": 5270 + }, + { + "epoch": 3.7914044236648086, + "grad_norm": 6.618193746795537, + "learning_rate": 2.065448333325249e-06, + "loss": 0.1242, + "step": 5271 + }, + { + "epoch": 3.7921237187556196, + "grad_norm": 2.6710508111122304, + "learning_rate": 2.065133174866734e-06, + "loss": 0.0897, + "step": 5272 + }, + { + "epoch": 3.7928430138464306, + "grad_norm": 3.4614775141742444, + "learning_rate": 2.0648179873311245e-06, + "loss": 0.0979, + "step": 5273 + }, + { + "epoch": 3.7935623089372417, + "grad_norm": 0.5528722134289475, + "learning_rate": 2.0645027707346384e-06, + "loss": 0.0008, + "step": 5274 + }, + { + "epoch": 3.7942816040280523, + "grad_norm": 3.6310319925898136, + "learning_rate": 2.0641875250934943e-06, + "loss": 0.1235, + "step": 5275 + }, + { + "epoch": 3.7950008991188637, + "grad_norm": 4.087068370024193, + "learning_rate": 2.0638722504239105e-06, + "loss": 0.0791, + "step": 5276 + }, + { + "epoch": 3.7957201942096743, + "grad_norm": 2.8161353237288402, + "learning_rate": 2.0635569467421105e-06, + "loss": 0.0703, + "step": 5277 + }, + { + "epoch": 3.7964394893004854, + "grad_norm": 1.0655856315136323, + "learning_rate": 2.0632416140643156e-06, + "loss": 0.0134, + "step": 5278 + }, + { + "epoch": 3.7971587843912964, + "grad_norm": 2.075236258960517, + "learning_rate": 2.0629262524067517e-06, + "loss": 0.0476, + "step": 5279 + }, + { + "epoch": 3.7978780794821074, + "grad_norm": 2.322056433439602, + "learning_rate": 2.0626108617856437e-06, + "loss": 0.0267, + "step": 5280 + }, + { + "epoch": 3.7985973745729185, + "grad_norm": 3.9031999485354785, + "learning_rate": 2.062295442217219e-06, + "loss": 0.1229, + "step": 5281 + }, + { + "epoch": 3.7993166696637295, + "grad_norm": 3.8496289866328275, + "learning_rate": 2.0619799937177068e-06, + "loss": 0.0639, + "step": 5282 + }, + { + "epoch": 3.8000359647545405, + "grad_norm": 4.768590175223172, + "learning_rate": 2.0616645163033375e-06, + "loss": 0.1613, + "step": 5283 + }, + { + "epoch": 3.8007552598453516, + "grad_norm": 4.034859323153518, + "learning_rate": 2.0613490099903427e-06, + "loss": 0.1489, + "step": 5284 + }, + { + "epoch": 3.8014745549361626, + "grad_norm": 3.2439829850848754, + "learning_rate": 2.061033474794956e-06, + "loss": 0.0618, + "step": 5285 + }, + { + "epoch": 3.8021938500269736, + "grad_norm": 1.2786236391097863, + "learning_rate": 2.0607179107334124e-06, + "loss": 0.0021, + "step": 5286 + }, + { + "epoch": 3.8029131451177847, + "grad_norm": 2.581824931694143, + "learning_rate": 2.0604023178219477e-06, + "loss": 0.0724, + "step": 5287 + }, + { + "epoch": 3.8036324402085957, + "grad_norm": 6.449707920466521, + "learning_rate": 2.0600866960768e-06, + "loss": 0.0775, + "step": 5288 + }, + { + "epoch": 3.8043517352994067, + "grad_norm": 3.156072535648808, + "learning_rate": 2.0597710455142088e-06, + "loss": 0.0835, + "step": 5289 + }, + { + "epoch": 3.8050710303902173, + "grad_norm": 3.617693275891581, + "learning_rate": 2.0594553661504144e-06, + "loss": 0.0666, + "step": 5290 + }, + { + "epoch": 3.805790325481029, + "grad_norm": 5.945310649830234, + "learning_rate": 2.0591396580016593e-06, + "loss": 0.2293, + "step": 5291 + }, + { + "epoch": 3.8065096205718394, + "grad_norm": 2.6163968426491295, + "learning_rate": 2.0588239210841874e-06, + "loss": 0.0365, + "step": 5292 + }, + { + "epoch": 3.807228915662651, + "grad_norm": 2.6800580433565, + "learning_rate": 2.058508155414244e-06, + "loss": 0.0241, + "step": 5293 + }, + { + "epoch": 3.8079482107534615, + "grad_norm": 3.872893852636392, + "learning_rate": 2.058192361008075e-06, + "loss": 0.0732, + "step": 5294 + }, + { + "epoch": 3.8086675058442725, + "grad_norm": 2.0145220755387117, + "learning_rate": 2.0578765378819296e-06, + "loss": 0.0412, + "step": 5295 + }, + { + "epoch": 3.8093868009350835, + "grad_norm": 6.20416410089323, + "learning_rate": 2.057560686052057e-06, + "loss": 0.1759, + "step": 5296 + }, + { + "epoch": 3.8101060960258946, + "grad_norm": 2.2023415267114084, + "learning_rate": 2.0572448055347075e-06, + "loss": 0.0563, + "step": 5297 + }, + { + "epoch": 3.8108253911167056, + "grad_norm": 3.215803986108004, + "learning_rate": 2.0569288963461357e-06, + "loss": 0.0192, + "step": 5298 + }, + { + "epoch": 3.8115446862075166, + "grad_norm": 5.045189272336017, + "learning_rate": 2.0566129585025943e-06, + "loss": 0.1667, + "step": 5299 + }, + { + "epoch": 3.8122639812983277, + "grad_norm": 6.446309256397397, + "learning_rate": 2.0562969920203387e-06, + "loss": 0.1152, + "step": 5300 + }, + { + "epoch": 3.8129832763891387, + "grad_norm": 2.3047637018262987, + "learning_rate": 2.0559809969156265e-06, + "loss": 0.0287, + "step": 5301 + }, + { + "epoch": 3.8137025714799497, + "grad_norm": 5.48712702346315, + "learning_rate": 2.055664973204716e-06, + "loss": 0.2237, + "step": 5302 + }, + { + "epoch": 3.8144218665707608, + "grad_norm": 2.7091601967663324, + "learning_rate": 2.055348920903867e-06, + "loss": 0.0646, + "step": 5303 + }, + { + "epoch": 3.815141161661572, + "grad_norm": 3.1022093439121607, + "learning_rate": 2.0550328400293416e-06, + "loss": 0.0949, + "step": 5304 + }, + { + "epoch": 3.8158604567523824, + "grad_norm": 7.74848227139838, + "learning_rate": 2.0547167305974017e-06, + "loss": 0.1664, + "step": 5305 + }, + { + "epoch": 3.816579751843194, + "grad_norm": 2.2005970130257495, + "learning_rate": 2.054400592624313e-06, + "loss": 0.0112, + "step": 5306 + }, + { + "epoch": 3.8172990469340045, + "grad_norm": 3.4768104683866747, + "learning_rate": 2.0540844261263396e-06, + "loss": 0.0525, + "step": 5307 + }, + { + "epoch": 3.818018342024816, + "grad_norm": 3.9439958723722954, + "learning_rate": 2.0537682311197503e-06, + "loss": 0.1658, + "step": 5308 + }, + { + "epoch": 3.8187376371156265, + "grad_norm": 3.1713066144492164, + "learning_rate": 2.0534520076208137e-06, + "loss": 0.0744, + "step": 5309 + }, + { + "epoch": 3.8194569322064376, + "grad_norm": 4.102089118613079, + "learning_rate": 2.053135755645799e-06, + "loss": 0.0278, + "step": 5310 + }, + { + "epoch": 3.8201762272972486, + "grad_norm": 3.4340301519885856, + "learning_rate": 2.0528194752109795e-06, + "loss": 0.0953, + "step": 5311 + }, + { + "epoch": 3.8208955223880596, + "grad_norm": 1.0544050615983573, + "learning_rate": 2.052503166332627e-06, + "loss": 0.003, + "step": 5312 + }, + { + "epoch": 3.8216148174788707, + "grad_norm": 2.86117919214093, + "learning_rate": 2.0521868290270174e-06, + "loss": 0.1093, + "step": 5313 + }, + { + "epoch": 3.8223341125696817, + "grad_norm": 2.370189032064602, + "learning_rate": 2.051870463310425e-06, + "loss": 0.0884, + "step": 5314 + }, + { + "epoch": 3.8230534076604927, + "grad_norm": 2.004715951197122, + "learning_rate": 2.051554069199129e-06, + "loss": 0.0257, + "step": 5315 + }, + { + "epoch": 3.8237727027513038, + "grad_norm": 5.144166935336616, + "learning_rate": 2.0512376467094086e-06, + "loss": 0.163, + "step": 5316 + }, + { + "epoch": 3.824491997842115, + "grad_norm": 2.409050470813284, + "learning_rate": 2.050921195857543e-06, + "loss": 0.0885, + "step": 5317 + }, + { + "epoch": 3.825211292932926, + "grad_norm": 1.590612848765668, + "learning_rate": 2.0506047166598146e-06, + "loss": 0.0187, + "step": 5318 + }, + { + "epoch": 3.825930588023737, + "grad_norm": 1.515203674002266, + "learning_rate": 2.050288209132507e-06, + "loss": 0.021, + "step": 5319 + }, + { + "epoch": 3.826649883114548, + "grad_norm": 2.939673396268763, + "learning_rate": 2.0499716732919052e-06, + "loss": 0.0417, + "step": 5320 + }, + { + "epoch": 3.827369178205359, + "grad_norm": 4.484721189126171, + "learning_rate": 2.0496551091542952e-06, + "loss": 0.2198, + "step": 5321 + }, + { + "epoch": 3.8280884732961695, + "grad_norm": 6.356281828349507, + "learning_rate": 2.0493385167359653e-06, + "loss": 0.1253, + "step": 5322 + }, + { + "epoch": 3.828807768386981, + "grad_norm": 0.8594442891300129, + "learning_rate": 2.0490218960532043e-06, + "loss": 0.0012, + "step": 5323 + }, + { + "epoch": 3.8295270634777916, + "grad_norm": 2.00974800915284, + "learning_rate": 2.0487052471223027e-06, + "loss": 0.0483, + "step": 5324 + }, + { + "epoch": 3.8302463585686026, + "grad_norm": 3.2068779946632087, + "learning_rate": 2.0483885699595533e-06, + "loss": 0.1353, + "step": 5325 + }, + { + "epoch": 3.8309656536594137, + "grad_norm": 7.210172891677194, + "learning_rate": 2.0480718645812488e-06, + "loss": 0.1683, + "step": 5326 + }, + { + "epoch": 3.8316849487502247, + "grad_norm": 3.353685442426929, + "learning_rate": 2.047755131003685e-06, + "loss": 0.0428, + "step": 5327 + }, + { + "epoch": 3.8324042438410357, + "grad_norm": 1.6323318681733163, + "learning_rate": 2.0474383692431584e-06, + "loss": 0.0054, + "step": 5328 + }, + { + "epoch": 3.8331235389318468, + "grad_norm": 3.1070545481201117, + "learning_rate": 2.0471215793159665e-06, + "loss": 0.1367, + "step": 5329 + }, + { + "epoch": 3.833842834022658, + "grad_norm": 4.007331426422302, + "learning_rate": 2.046804761238409e-06, + "loss": 0.0574, + "step": 5330 + }, + { + "epoch": 3.834562129113469, + "grad_norm": 3.8423157934447607, + "learning_rate": 2.0464879150267863e-06, + "loss": 0.0621, + "step": 5331 + }, + { + "epoch": 3.83528142420428, + "grad_norm": 3.0098604196883048, + "learning_rate": 2.046171040697401e-06, + "loss": 0.0249, + "step": 5332 + }, + { + "epoch": 3.836000719295091, + "grad_norm": 3.7126767922149213, + "learning_rate": 2.0458541382665575e-06, + "loss": 0.1176, + "step": 5333 + }, + { + "epoch": 3.836720014385902, + "grad_norm": 4.893344186425532, + "learning_rate": 2.04553720775056e-06, + "loss": 0.1333, + "step": 5334 + }, + { + "epoch": 3.837439309476713, + "grad_norm": 2.8654242716946046, + "learning_rate": 2.0452202491657156e-06, + "loss": 0.0592, + "step": 5335 + }, + { + "epoch": 3.838158604567524, + "grad_norm": 3.215886457238917, + "learning_rate": 2.044903262528332e-06, + "loss": 0.0647, + "step": 5336 + }, + { + "epoch": 3.8388778996583346, + "grad_norm": 1.6346872203531968, + "learning_rate": 2.0445862478547194e-06, + "loss": 0.0407, + "step": 5337 + }, + { + "epoch": 3.839597194749146, + "grad_norm": 0.6153967178747611, + "learning_rate": 2.0442692051611883e-06, + "loss": 0.0016, + "step": 5338 + }, + { + "epoch": 3.8403164898399567, + "grad_norm": 4.914572195472663, + "learning_rate": 2.0439521344640504e-06, + "loss": 0.1313, + "step": 5339 + }, + { + "epoch": 3.8410357849307677, + "grad_norm": 3.5070591586091706, + "learning_rate": 2.0436350357796213e-06, + "loss": 0.0622, + "step": 5340 + }, + { + "epoch": 3.8417550800215787, + "grad_norm": 4.336613180480845, + "learning_rate": 2.043317909124215e-06, + "loss": 0.007, + "step": 5341 + }, + { + "epoch": 3.8424743751123898, + "grad_norm": 2.3273715273355937, + "learning_rate": 2.0430007545141487e-06, + "loss": 0.0107, + "step": 5342 + }, + { + "epoch": 3.843193670203201, + "grad_norm": 6.069869622927708, + "learning_rate": 2.04268357196574e-06, + "loss": 0.1676, + "step": 5343 + }, + { + "epoch": 3.843912965294012, + "grad_norm": 2.9534903012390408, + "learning_rate": 2.042366361495309e-06, + "loss": 0.0854, + "step": 5344 + }, + { + "epoch": 3.844632260384823, + "grad_norm": 2.4277347856566136, + "learning_rate": 2.0420491231191764e-06, + "loss": 0.056, + "step": 5345 + }, + { + "epoch": 3.845351555475634, + "grad_norm": 4.4797805710145795, + "learning_rate": 2.0417318568536646e-06, + "loss": 0.0814, + "step": 5346 + }, + { + "epoch": 3.846070850566445, + "grad_norm": 3.2364647734698315, + "learning_rate": 2.0414145627150984e-06, + "loss": 0.1081, + "step": 5347 + }, + { + "epoch": 3.846790145657256, + "grad_norm": 4.038025541928996, + "learning_rate": 2.0410972407198026e-06, + "loss": 0.0142, + "step": 5348 + }, + { + "epoch": 3.847509440748067, + "grad_norm": 3.7079634535873005, + "learning_rate": 2.0407798908841035e-06, + "loss": 0.0484, + "step": 5349 + }, + { + "epoch": 3.848228735838878, + "grad_norm": 1.7686605058901939, + "learning_rate": 2.0404625132243304e-06, + "loss": 0.0042, + "step": 5350 + }, + { + "epoch": 3.848948030929689, + "grad_norm": 3.04233567571756, + "learning_rate": 2.0401451077568117e-06, + "loss": 0.0615, + "step": 5351 + }, + { + "epoch": 3.8496673260204997, + "grad_norm": 4.178662080960521, + "learning_rate": 2.0398276744978785e-06, + "loss": 0.1876, + "step": 5352 + }, + { + "epoch": 3.850386621111311, + "grad_norm": 1.6708100649875677, + "learning_rate": 2.0395102134638647e-06, + "loss": 0.0378, + "step": 5353 + }, + { + "epoch": 3.8511059162021217, + "grad_norm": 2.362808163149021, + "learning_rate": 2.0391927246711027e-06, + "loss": 0.1023, + "step": 5354 + }, + { + "epoch": 3.8518252112929328, + "grad_norm": 2.70943358821561, + "learning_rate": 2.038875208135929e-06, + "loss": 0.0753, + "step": 5355 + }, + { + "epoch": 3.852544506383744, + "grad_norm": 0.02980492695360001, + "learning_rate": 2.0385576638746794e-06, + "loss": 0.0001, + "step": 5356 + }, + { + "epoch": 3.853263801474555, + "grad_norm": 5.11878187050395, + "learning_rate": 2.038240091903693e-06, + "loss": 0.1178, + "step": 5357 + }, + { + "epoch": 3.853983096565366, + "grad_norm": 2.854284041801305, + "learning_rate": 2.0379224922393087e-06, + "loss": 0.123, + "step": 5358 + }, + { + "epoch": 3.854702391656177, + "grad_norm": 2.410060462809048, + "learning_rate": 2.0376048648978685e-06, + "loss": 0.041, + "step": 5359 + }, + { + "epoch": 3.855421686746988, + "grad_norm": 1.240053051965953, + "learning_rate": 2.0372872098957132e-06, + "loss": 0.0175, + "step": 5360 + }, + { + "epoch": 3.856140981837799, + "grad_norm": 1.1276464584675983, + "learning_rate": 2.0369695272491885e-06, + "loss": 0.0044, + "step": 5361 + }, + { + "epoch": 3.85686027692861, + "grad_norm": 5.175724001070538, + "learning_rate": 2.0366518169746385e-06, + "loss": 0.2011, + "step": 5362 + }, + { + "epoch": 3.857579572019421, + "grad_norm": 2.0785680952976024, + "learning_rate": 2.036334079088411e-06, + "loss": 0.0277, + "step": 5363 + }, + { + "epoch": 3.858298867110232, + "grad_norm": 3.224710518064959, + "learning_rate": 2.0360163136068536e-06, + "loss": 0.0893, + "step": 5364 + }, + { + "epoch": 3.859018162201043, + "grad_norm": 7.416492353752845, + "learning_rate": 2.035698520546316e-06, + "loss": 0.1562, + "step": 5365 + }, + { + "epoch": 3.859737457291854, + "grad_norm": 3.2865413107826793, + "learning_rate": 2.035380699923149e-06, + "loss": 0.0694, + "step": 5366 + }, + { + "epoch": 3.8604567523826647, + "grad_norm": 2.437087678407865, + "learning_rate": 2.0350628517537046e-06, + "loss": 0.0099, + "step": 5367 + }, + { + "epoch": 3.861176047473476, + "grad_norm": 4.168206056596029, + "learning_rate": 2.034744976054338e-06, + "loss": 0.0681, + "step": 5368 + }, + { + "epoch": 3.861895342564287, + "grad_norm": 0.6940493410287935, + "learning_rate": 2.034427072841403e-06, + "loss": 0.0063, + "step": 5369 + }, + { + "epoch": 3.8626146376550983, + "grad_norm": 3.8933123145464728, + "learning_rate": 2.034109142131257e-06, + "loss": 0.1809, + "step": 5370 + }, + { + "epoch": 3.863333932745909, + "grad_norm": 2.116231072471638, + "learning_rate": 2.0337911839402585e-06, + "loss": 0.0444, + "step": 5371 + }, + { + "epoch": 3.86405322783672, + "grad_norm": 2.9969554609304483, + "learning_rate": 2.0334731982847666e-06, + "loss": 0.0491, + "step": 5372 + }, + { + "epoch": 3.864772522927531, + "grad_norm": 5.545603502396158, + "learning_rate": 2.033155185181142e-06, + "loss": 0.1338, + "step": 5373 + }, + { + "epoch": 3.865491818018342, + "grad_norm": 3.773188442837339, + "learning_rate": 2.0328371446457473e-06, + "loss": 0.0972, + "step": 5374 + }, + { + "epoch": 3.866211113109153, + "grad_norm": 3.643086353251341, + "learning_rate": 2.032519076694946e-06, + "loss": 0.0654, + "step": 5375 + }, + { + "epoch": 3.866930408199964, + "grad_norm": 1.6172355840912942, + "learning_rate": 2.032200981345103e-06, + "loss": 0.0451, + "step": 5376 + }, + { + "epoch": 3.867649703290775, + "grad_norm": 4.660063784246082, + "learning_rate": 2.0318828586125855e-06, + "loss": 0.1108, + "step": 5377 + }, + { + "epoch": 3.868368998381586, + "grad_norm": 5.366688038613864, + "learning_rate": 2.0315647085137615e-06, + "loss": 0.2536, + "step": 5378 + }, + { + "epoch": 3.869088293472397, + "grad_norm": 5.107614443716177, + "learning_rate": 2.031246531065e-06, + "loss": 0.1932, + "step": 5379 + }, + { + "epoch": 3.869807588563208, + "grad_norm": 0.07619259861668488, + "learning_rate": 2.030928326282672e-06, + "loss": 0.0005, + "step": 5380 + }, + { + "epoch": 3.870526883654019, + "grad_norm": 0.9362200166562602, + "learning_rate": 2.0306100941831493e-06, + "loss": 0.0101, + "step": 5381 + }, + { + "epoch": 3.87124617874483, + "grad_norm": 1.8469638858128437, + "learning_rate": 2.030291834782806e-06, + "loss": 0.0236, + "step": 5382 + }, + { + "epoch": 3.8719654738356413, + "grad_norm": 2.748775780042448, + "learning_rate": 2.0299735480980168e-06, + "loss": 0.0513, + "step": 5383 + }, + { + "epoch": 3.872684768926452, + "grad_norm": 3.4352723472167037, + "learning_rate": 2.0296552341451584e-06, + "loss": 0.0153, + "step": 5384 + }, + { + "epoch": 3.8734040640172633, + "grad_norm": 3.2822351483197387, + "learning_rate": 2.0293368929406083e-06, + "loss": 0.0695, + "step": 5385 + }, + { + "epoch": 3.874123359108074, + "grad_norm": 2.6309115444838302, + "learning_rate": 2.029018524500746e-06, + "loss": 0.0412, + "step": 5386 + }, + { + "epoch": 3.874842654198885, + "grad_norm": 5.879034105684064, + "learning_rate": 2.0287001288419514e-06, + "loss": 0.277, + "step": 5387 + }, + { + "epoch": 3.875561949289696, + "grad_norm": 2.488435280650013, + "learning_rate": 2.0283817059806074e-06, + "loss": 0.0216, + "step": 5388 + }, + { + "epoch": 3.876281244380507, + "grad_norm": 0.131253161211255, + "learning_rate": 2.0280632559330973e-06, + "loss": 0.0003, + "step": 5389 + }, + { + "epoch": 3.877000539471318, + "grad_norm": 2.4827163865491593, + "learning_rate": 2.0277447787158056e-06, + "loss": 0.0604, + "step": 5390 + }, + { + "epoch": 3.877719834562129, + "grad_norm": 4.2162803040585235, + "learning_rate": 2.027426274345118e-06, + "loss": 0.1507, + "step": 5391 + }, + { + "epoch": 3.87843912965294, + "grad_norm": 3.5941211610079615, + "learning_rate": 2.0271077428374237e-06, + "loss": 0.0113, + "step": 5392 + }, + { + "epoch": 3.879158424743751, + "grad_norm": 0.23357902073084164, + "learning_rate": 2.0267891842091104e-06, + "loss": 0.0007, + "step": 5393 + }, + { + "epoch": 3.879877719834562, + "grad_norm": 6.789857140195157, + "learning_rate": 2.026470598476569e-06, + "loss": 0.1061, + "step": 5394 + }, + { + "epoch": 3.8805970149253732, + "grad_norm": 1.7571152177173903, + "learning_rate": 2.0261519856561905e-06, + "loss": 0.036, + "step": 5395 + }, + { + "epoch": 3.8813163100161843, + "grad_norm": 1.5819050232586616, + "learning_rate": 2.0258333457643688e-06, + "loss": 0.0268, + "step": 5396 + }, + { + "epoch": 3.8820356051069953, + "grad_norm": 4.223149175580265, + "learning_rate": 2.025514678817499e-06, + "loss": 0.0805, + "step": 5397 + }, + { + "epoch": 3.8827549001978063, + "grad_norm": 2.0836797033612084, + "learning_rate": 2.025195984831976e-06, + "loss": 0.028, + "step": 5398 + }, + { + "epoch": 3.883474195288617, + "grad_norm": 3.086264057600444, + "learning_rate": 2.0248772638241975e-06, + "loss": 0.0613, + "step": 5399 + }, + { + "epoch": 3.8841934903794284, + "grad_norm": 2.6899952810079135, + "learning_rate": 2.0245585158105627e-06, + "loss": 0.0635, + "step": 5400 + }, + { + "epoch": 3.884912785470239, + "grad_norm": 3.5959572855437743, + "learning_rate": 2.0242397408074716e-06, + "loss": 0.0832, + "step": 5401 + }, + { + "epoch": 3.88563208056105, + "grad_norm": 4.229664070073748, + "learning_rate": 2.0239209388313253e-06, + "loss": 0.1572, + "step": 5402 + }, + { + "epoch": 3.886351375651861, + "grad_norm": 2.3841954250459496, + "learning_rate": 2.023602109898527e-06, + "loss": 0.055, + "step": 5403 + }, + { + "epoch": 3.887070670742672, + "grad_norm": 2.690863002653772, + "learning_rate": 2.0232832540254807e-06, + "loss": 0.0831, + "step": 5404 + }, + { + "epoch": 3.887789965833483, + "grad_norm": 3.0823479163741916, + "learning_rate": 2.0229643712285926e-06, + "loss": 0.1378, + "step": 5405 + }, + { + "epoch": 3.888509260924294, + "grad_norm": 2.891209727852462, + "learning_rate": 2.0226454615242697e-06, + "loss": 0.0846, + "step": 5406 + }, + { + "epoch": 3.889228556015105, + "grad_norm": 2.0985792789865876, + "learning_rate": 2.0223265249289196e-06, + "loss": 0.0516, + "step": 5407 + }, + { + "epoch": 3.8899478511059162, + "grad_norm": 1.629908614659466, + "learning_rate": 2.0220075614589543e-06, + "loss": 0.0304, + "step": 5408 + }, + { + "epoch": 3.8906671461967273, + "grad_norm": 4.711270084268678, + "learning_rate": 2.0216885711307825e-06, + "loss": 0.1255, + "step": 5409 + }, + { + "epoch": 3.8913864412875383, + "grad_norm": 3.7087466962416107, + "learning_rate": 2.021369553960818e-06, + "loss": 0.1125, + "step": 5410 + }, + { + "epoch": 3.8921057363783493, + "grad_norm": 3.749600235027917, + "learning_rate": 2.0210505099654756e-06, + "loss": 0.0949, + "step": 5411 + }, + { + "epoch": 3.8928250314691604, + "grad_norm": 2.8888906210380934, + "learning_rate": 2.0207314391611686e-06, + "loss": 0.0553, + "step": 5412 + }, + { + "epoch": 3.8935443265599714, + "grad_norm": 1.7885984685549063, + "learning_rate": 2.0204123415643157e-06, + "loss": 0.0209, + "step": 5413 + }, + { + "epoch": 3.894263621650782, + "grad_norm": 2.963662733268407, + "learning_rate": 2.020093217191334e-06, + "loss": 0.0248, + "step": 5414 + }, + { + "epoch": 3.8949829167415935, + "grad_norm": 0.5301917938308697, + "learning_rate": 2.0197740660586433e-06, + "loss": 0.0047, + "step": 5415 + }, + { + "epoch": 3.895702211832404, + "grad_norm": 0.6946249851000219, + "learning_rate": 2.019454888182665e-06, + "loss": 0.0017, + "step": 5416 + }, + { + "epoch": 3.896421506923215, + "grad_norm": 4.469477313517812, + "learning_rate": 2.0191356835798197e-06, + "loss": 0.1239, + "step": 5417 + }, + { + "epoch": 3.897140802014026, + "grad_norm": 5.170834161867368, + "learning_rate": 2.018816452266533e-06, + "loss": 0.1706, + "step": 5418 + }, + { + "epoch": 3.897860097104837, + "grad_norm": 3.136477333545605, + "learning_rate": 2.018497194259229e-06, + "loss": 0.1184, + "step": 5419 + }, + { + "epoch": 3.898579392195648, + "grad_norm": 0.9109973975248912, + "learning_rate": 2.0181779095743335e-06, + "loss": 0.002, + "step": 5420 + }, + { + "epoch": 3.8992986872864592, + "grad_norm": 4.5804159906991595, + "learning_rate": 2.0178585982282755e-06, + "loss": 0.0936, + "step": 5421 + }, + { + "epoch": 3.9000179823772703, + "grad_norm": 2.140266187228584, + "learning_rate": 2.017539260237484e-06, + "loss": 0.0518, + "step": 5422 + }, + { + "epoch": 3.9007372774680813, + "grad_norm": 3.708328670535963, + "learning_rate": 2.0172198956183885e-06, + "loss": 0.1687, + "step": 5423 + }, + { + "epoch": 3.9014565725588923, + "grad_norm": 2.6955542244102877, + "learning_rate": 2.0169005043874217e-06, + "loss": 0.0118, + "step": 5424 + }, + { + "epoch": 3.9021758676497034, + "grad_norm": 2.146017558094474, + "learning_rate": 2.016581086561016e-06, + "loss": 0.0448, + "step": 5425 + }, + { + "epoch": 3.9028951627405144, + "grad_norm": 0.7094886433524717, + "learning_rate": 2.016261642155607e-06, + "loss": 0.0017, + "step": 5426 + }, + { + "epoch": 3.9036144578313254, + "grad_norm": 2.878902004035189, + "learning_rate": 2.0159421711876303e-06, + "loss": 0.1103, + "step": 5427 + }, + { + "epoch": 3.9043337529221365, + "grad_norm": 3.506635339058208, + "learning_rate": 2.0156226736735227e-06, + "loss": 0.0246, + "step": 5428 + }, + { + "epoch": 3.905053048012947, + "grad_norm": 2.644780270533339, + "learning_rate": 2.0153031496297236e-06, + "loss": 0.0633, + "step": 5429 + }, + { + "epoch": 3.9057723431037585, + "grad_norm": 3.3675917014597037, + "learning_rate": 2.014983599072673e-06, + "loss": 0.1005, + "step": 5430 + }, + { + "epoch": 3.906491638194569, + "grad_norm": 3.1225200825553623, + "learning_rate": 2.0146640220188123e-06, + "loss": 0.1515, + "step": 5431 + }, + { + "epoch": 3.90721093328538, + "grad_norm": 3.561796674764048, + "learning_rate": 2.014344418484584e-06, + "loss": 0.1382, + "step": 5432 + }, + { + "epoch": 3.907930228376191, + "grad_norm": 2.5874344337232937, + "learning_rate": 2.0140247884864326e-06, + "loss": 0.0716, + "step": 5433 + }, + { + "epoch": 3.9086495234670022, + "grad_norm": 4.50645624526715, + "learning_rate": 2.0137051320408036e-06, + "loss": 0.2128, + "step": 5434 + }, + { + "epoch": 3.9093688185578133, + "grad_norm": 1.6666239219940133, + "learning_rate": 2.0133854491641437e-06, + "loss": 0.0133, + "step": 5435 + }, + { + "epoch": 3.9100881136486243, + "grad_norm": 1.7657625822370628, + "learning_rate": 2.013065739872901e-06, + "loss": 0.0364, + "step": 5436 + }, + { + "epoch": 3.9108074087394353, + "grad_norm": 3.728566597900546, + "learning_rate": 2.0127460041835256e-06, + "loss": 0.1265, + "step": 5437 + }, + { + "epoch": 3.9115267038302464, + "grad_norm": 0.07434633976800208, + "learning_rate": 2.012426242112468e-06, + "loss": 0.0005, + "step": 5438 + }, + { + "epoch": 3.9122459989210574, + "grad_norm": 3.438058216528022, + "learning_rate": 2.0121064536761807e-06, + "loss": 0.1224, + "step": 5439 + }, + { + "epoch": 3.9129652940118684, + "grad_norm": 2.3862290052891417, + "learning_rate": 2.0117866388911176e-06, + "loss": 0.0052, + "step": 5440 + }, + { + "epoch": 3.9136845891026795, + "grad_norm": 1.7310645391329496, + "learning_rate": 2.011466797773733e-06, + "loss": 0.0258, + "step": 5441 + }, + { + "epoch": 3.9144038841934905, + "grad_norm": 3.635721249184957, + "learning_rate": 2.0111469303404843e-06, + "loss": 0.0971, + "step": 5442 + }, + { + "epoch": 3.9151231792843015, + "grad_norm": 2.089355641913553, + "learning_rate": 2.0108270366078284e-06, + "loss": 0.0446, + "step": 5443 + }, + { + "epoch": 3.915842474375112, + "grad_norm": 5.810586444615338, + "learning_rate": 2.010507116592225e-06, + "loss": 0.2797, + "step": 5444 + }, + { + "epoch": 3.9165617694659236, + "grad_norm": 2.5669560736652226, + "learning_rate": 2.010187170310135e-06, + "loss": 0.0615, + "step": 5445 + }, + { + "epoch": 3.917281064556734, + "grad_norm": 6.383491661095621, + "learning_rate": 2.0098671977780183e-06, + "loss": 0.2065, + "step": 5446 + }, + { + "epoch": 3.9180003596475457, + "grad_norm": 6.066848837493063, + "learning_rate": 2.0095471990123394e-06, + "loss": 0.1633, + "step": 5447 + }, + { + "epoch": 3.9187196547383563, + "grad_norm": 0.32165780030920316, + "learning_rate": 2.009227174029563e-06, + "loss": 0.0007, + "step": 5448 + }, + { + "epoch": 3.9194389498291673, + "grad_norm": 4.525658245710453, + "learning_rate": 2.0089071228461543e-06, + "loss": 0.034, + "step": 5449 + }, + { + "epoch": 3.9201582449199783, + "grad_norm": 3.2309924990266037, + "learning_rate": 2.0085870454785813e-06, + "loss": 0.0913, + "step": 5450 + }, + { + "epoch": 3.9208775400107894, + "grad_norm": 1.5202404567270105, + "learning_rate": 2.008266941943311e-06, + "loss": 0.0067, + "step": 5451 + }, + { + "epoch": 3.9215968351016004, + "grad_norm": 0.4970364770299013, + "learning_rate": 2.007946812256815e-06, + "loss": 0.0029, + "step": 5452 + }, + { + "epoch": 3.9223161301924114, + "grad_norm": 0.6285758605507731, + "learning_rate": 2.0076266564355644e-06, + "loss": 0.0014, + "step": 5453 + }, + { + "epoch": 3.9230354252832225, + "grad_norm": 1.540101777286954, + "learning_rate": 2.00730647449603e-06, + "loss": 0.0256, + "step": 5454 + }, + { + "epoch": 3.9237547203740335, + "grad_norm": 3.5753951278668263, + "learning_rate": 2.0069862664546877e-06, + "loss": 0.0942, + "step": 5455 + }, + { + "epoch": 3.9244740154648445, + "grad_norm": 7.846983062724066, + "learning_rate": 2.006666032328012e-06, + "loss": 0.171, + "step": 5456 + }, + { + "epoch": 3.9251933105556556, + "grad_norm": 0.20121854363860073, + "learning_rate": 2.006345772132479e-06, + "loss": 0.0007, + "step": 5457 + }, + { + "epoch": 3.9259126056464666, + "grad_norm": 2.941069914129251, + "learning_rate": 2.006025485884568e-06, + "loss": 0.0485, + "step": 5458 + }, + { + "epoch": 3.926631900737277, + "grad_norm": 4.627913804416625, + "learning_rate": 2.0057051736007566e-06, + "loss": 0.0059, + "step": 5459 + }, + { + "epoch": 3.9273511958280887, + "grad_norm": 0.1283859467657036, + "learning_rate": 2.0053848352975267e-06, + "loss": 0.0004, + "step": 5460 + }, + { + "epoch": 3.9280704909188993, + "grad_norm": 3.504431995539154, + "learning_rate": 2.00506447099136e-06, + "loss": 0.0704, + "step": 5461 + }, + { + "epoch": 3.9287897860097107, + "grad_norm": 2.0663005734746736, + "learning_rate": 2.0047440806987394e-06, + "loss": 0.0457, + "step": 5462 + }, + { + "epoch": 3.9295090811005213, + "grad_norm": 2.5561306227371463, + "learning_rate": 2.0044236644361497e-06, + "loss": 0.0323, + "step": 5463 + }, + { + "epoch": 3.9302283761913324, + "grad_norm": 3.3837184661243134, + "learning_rate": 2.004103222220077e-06, + "loss": 0.0766, + "step": 5464 + }, + { + "epoch": 3.9309476712821434, + "grad_norm": 2.892789692513667, + "learning_rate": 2.003782754067009e-06, + "loss": 0.021, + "step": 5465 + }, + { + "epoch": 3.9316669663729544, + "grad_norm": 1.3132205120975196, + "learning_rate": 2.0034622599934332e-06, + "loss": 0.0095, + "step": 5466 + }, + { + "epoch": 3.9323862614637655, + "grad_norm": 0.027661753741693666, + "learning_rate": 2.003141740015841e-06, + "loss": 0.0001, + "step": 5467 + }, + { + "epoch": 3.9331055565545765, + "grad_norm": 1.9403790312730997, + "learning_rate": 2.002821194150723e-06, + "loss": 0.0578, + "step": 5468 + }, + { + "epoch": 3.9338248516453875, + "grad_norm": 1.2313024742852494, + "learning_rate": 2.002500622414572e-06, + "loss": 0.0219, + "step": 5469 + }, + { + "epoch": 3.9345441467361986, + "grad_norm": 0.7946361631626475, + "learning_rate": 2.002180024823881e-06, + "loss": 0.003, + "step": 5470 + }, + { + "epoch": 3.9352634418270096, + "grad_norm": 2.1669427837771673, + "learning_rate": 2.0018594013951466e-06, + "loss": 0.0456, + "step": 5471 + }, + { + "epoch": 3.9359827369178206, + "grad_norm": 3.306902842771055, + "learning_rate": 2.0015387521448653e-06, + "loss": 0.113, + "step": 5472 + }, + { + "epoch": 3.9367020320086317, + "grad_norm": 0.3959839544780945, + "learning_rate": 2.0012180770895346e-06, + "loss": 0.0022, + "step": 5473 + }, + { + "epoch": 3.9374213270994427, + "grad_norm": 5.093578844220308, + "learning_rate": 2.0008973762456544e-06, + "loss": 0.0193, + "step": 5474 + }, + { + "epoch": 3.9381406221902537, + "grad_norm": 1.2886832960645653, + "learning_rate": 2.0005766496297244e-06, + "loss": 0.0096, + "step": 5475 + }, + { + "epoch": 3.9388599172810643, + "grad_norm": 3.5490485615909932, + "learning_rate": 2.000255897258247e-06, + "loss": 0.0679, + "step": 5476 + }, + { + "epoch": 3.939579212371876, + "grad_norm": 0.47776609850885055, + "learning_rate": 1.999935119147726e-06, + "loss": 0.0017, + "step": 5477 + }, + { + "epoch": 3.9402985074626864, + "grad_norm": 3.3494088608091452, + "learning_rate": 1.9996143153146645e-06, + "loss": 0.0523, + "step": 5478 + }, + { + "epoch": 3.9410178025534974, + "grad_norm": 0.5077973521240079, + "learning_rate": 1.9992934857755705e-06, + "loss": 0.005, + "step": 5479 + }, + { + "epoch": 3.9417370976443085, + "grad_norm": 4.3607839664385395, + "learning_rate": 1.99897263054695e-06, + "loss": 0.1008, + "step": 5480 + }, + { + "epoch": 3.9424563927351195, + "grad_norm": 0.7945462241567474, + "learning_rate": 1.9986517496453114e-06, + "loss": 0.01, + "step": 5481 + }, + { + "epoch": 3.9431756878259305, + "grad_norm": 5.262571814119534, + "learning_rate": 1.998330843087165e-06, + "loss": 0.0981, + "step": 5482 + }, + { + "epoch": 3.9438949829167416, + "grad_norm": 5.621074627468403, + "learning_rate": 1.9980099108890216e-06, + "loss": 0.0919, + "step": 5483 + }, + { + "epoch": 3.9446142780075526, + "grad_norm": 4.401237162729927, + "learning_rate": 1.9976889530673946e-06, + "loss": 0.066, + "step": 5484 + }, + { + "epoch": 3.9453335730983636, + "grad_norm": 3.8585422741225344, + "learning_rate": 1.997367969638797e-06, + "loss": 0.1439, + "step": 5485 + }, + { + "epoch": 3.9460528681891747, + "grad_norm": 2.2391085825499557, + "learning_rate": 1.9970469606197446e-06, + "loss": 0.0133, + "step": 5486 + }, + { + "epoch": 3.9467721632799857, + "grad_norm": 0.08128072822391273, + "learning_rate": 1.9967259260267532e-06, + "loss": 0.0002, + "step": 5487 + }, + { + "epoch": 3.9474914583707967, + "grad_norm": 1.548318716920402, + "learning_rate": 1.9964048658763414e-06, + "loss": 0.0079, + "step": 5488 + }, + { + "epoch": 3.9482107534616078, + "grad_norm": 3.1243780555299563, + "learning_rate": 1.9960837801850274e-06, + "loss": 0.0195, + "step": 5489 + }, + { + "epoch": 3.948930048552419, + "grad_norm": 4.361417957070066, + "learning_rate": 1.995762668969332e-06, + "loss": 0.0507, + "step": 5490 + }, + { + "epoch": 3.9496493436432294, + "grad_norm": 1.8488301441392883, + "learning_rate": 1.995441532245777e-06, + "loss": 0.035, + "step": 5491 + }, + { + "epoch": 3.950368638734041, + "grad_norm": 4.224606395312267, + "learning_rate": 1.9951203700308855e-06, + "loss": 0.0281, + "step": 5492 + }, + { + "epoch": 3.9510879338248515, + "grad_norm": 1.4278211111296704, + "learning_rate": 1.994799182341182e-06, + "loss": 0.0054, + "step": 5493 + }, + { + "epoch": 3.9518072289156625, + "grad_norm": 4.40202335276364, + "learning_rate": 1.994477969193192e-06, + "loss": 0.184, + "step": 5494 + }, + { + "epoch": 3.9525265240064735, + "grad_norm": 1.9442711734265485, + "learning_rate": 1.9941567306034422e-06, + "loss": 0.0394, + "step": 5495 + }, + { + "epoch": 3.9532458190972846, + "grad_norm": 2.88894365735405, + "learning_rate": 1.9938354665884613e-06, + "loss": 0.0649, + "step": 5496 + }, + { + "epoch": 3.9539651141880956, + "grad_norm": 0.22364668376683655, + "learning_rate": 1.993514177164779e-06, + "loss": 0.0014, + "step": 5497 + }, + { + "epoch": 3.9546844092789066, + "grad_norm": 3.923497944226381, + "learning_rate": 1.993192862348925e-06, + "loss": 0.1114, + "step": 5498 + }, + { + "epoch": 3.9554037043697177, + "grad_norm": 3.6806706959489675, + "learning_rate": 1.9928715221574336e-06, + "loss": 0.0616, + "step": 5499 + }, + { + "epoch": 3.9561229994605287, + "grad_norm": 4.485348509179861, + "learning_rate": 1.9925501566068367e-06, + "loss": 0.077, + "step": 5500 + }, + { + "epoch": 3.9568422945513397, + "grad_norm": 1.9879807211513736, + "learning_rate": 1.992228765713669e-06, + "loss": 0.037, + "step": 5501 + }, + { + "epoch": 3.9575615896421508, + "grad_norm": 2.567672510361923, + "learning_rate": 1.991907349494468e-06, + "loss": 0.041, + "step": 5502 + }, + { + "epoch": 3.958280884732962, + "grad_norm": 5.160169423425565, + "learning_rate": 1.991585907965771e-06, + "loss": 0.0597, + "step": 5503 + }, + { + "epoch": 3.959000179823773, + "grad_norm": 3.0741943023498504, + "learning_rate": 1.9912644411441146e-06, + "loss": 0.0648, + "step": 5504 + }, + { + "epoch": 3.959719474914584, + "grad_norm": 1.0606487179727198, + "learning_rate": 1.9909429490460414e-06, + "loss": 0.0294, + "step": 5505 + }, + { + "epoch": 3.9604387700053945, + "grad_norm": 0.21482288911255237, + "learning_rate": 1.990621431688091e-06, + "loss": 0.0004, + "step": 5506 + }, + { + "epoch": 3.961158065096206, + "grad_norm": 2.6878978994050318, + "learning_rate": 1.9902998890868066e-06, + "loss": 0.0524, + "step": 5507 + }, + { + "epoch": 3.9618773601870165, + "grad_norm": 5.3643029727119345, + "learning_rate": 1.989978321258733e-06, + "loss": 0.2607, + "step": 5508 + }, + { + "epoch": 3.962596655277828, + "grad_norm": 3.5872097912166754, + "learning_rate": 1.989656728220414e-06, + "loss": 0.0612, + "step": 5509 + }, + { + "epoch": 3.9633159503686386, + "grad_norm": 6.1013589873449945, + "learning_rate": 1.989335109988397e-06, + "loss": 0.0473, + "step": 5510 + }, + { + "epoch": 3.9640352454594496, + "grad_norm": 1.1986790965095284, + "learning_rate": 1.9890134665792294e-06, + "loss": 0.0181, + "step": 5511 + }, + { + "epoch": 3.9647545405502607, + "grad_norm": 3.4842959354560112, + "learning_rate": 1.9886917980094607e-06, + "loss": 0.0457, + "step": 5512 + }, + { + "epoch": 3.9654738356410717, + "grad_norm": 2.309978795442391, + "learning_rate": 1.988370104295641e-06, + "loss": 0.042, + "step": 5513 + }, + { + "epoch": 3.9661931307318827, + "grad_norm": 5.692286387141629, + "learning_rate": 1.988048385454322e-06, + "loss": 0.1224, + "step": 5514 + }, + { + "epoch": 3.9669124258226938, + "grad_norm": 6.1578796705460785, + "learning_rate": 1.987726641502057e-06, + "loss": 0.0631, + "step": 5515 + }, + { + "epoch": 3.967631720913505, + "grad_norm": 0.4964453530546975, + "learning_rate": 1.9874048724554e-06, + "loss": 0.0009, + "step": 5516 + }, + { + "epoch": 3.968351016004316, + "grad_norm": 2.485940184374736, + "learning_rate": 1.987083078330907e-06, + "loss": 0.0277, + "step": 5517 + }, + { + "epoch": 3.969070311095127, + "grad_norm": 4.1396874580368355, + "learning_rate": 1.986761259145135e-06, + "loss": 0.0716, + "step": 5518 + }, + { + "epoch": 3.969789606185938, + "grad_norm": 3.5940088953448215, + "learning_rate": 1.9864394149146403e-06, + "loss": 0.1265, + "step": 5519 + }, + { + "epoch": 3.970508901276749, + "grad_norm": 3.7208447070777737, + "learning_rate": 1.986117545655985e-06, + "loss": 0.1207, + "step": 5520 + }, + { + "epoch": 3.9712281963675595, + "grad_norm": 3.8469311617217623, + "learning_rate": 1.9857956513857283e-06, + "loss": 0.0475, + "step": 5521 + }, + { + "epoch": 3.971947491458371, + "grad_norm": 4.166960932584081, + "learning_rate": 1.985473732120432e-06, + "loss": 0.1098, + "step": 5522 + }, + { + "epoch": 3.9726667865491816, + "grad_norm": 4.356264374419339, + "learning_rate": 1.9851517878766607e-06, + "loss": 0.0562, + "step": 5523 + }, + { + "epoch": 3.973386081639993, + "grad_norm": 6.165201691289361, + "learning_rate": 1.984829818670978e-06, + "loss": 0.1165, + "step": 5524 + }, + { + "epoch": 3.9741053767308037, + "grad_norm": 1.3392505645822996, + "learning_rate": 1.98450782451995e-06, + "loss": 0.0247, + "step": 5525 + }, + { + "epoch": 3.9748246718216147, + "grad_norm": 0.22224519621060682, + "learning_rate": 1.9841858054401442e-06, + "loss": 0.0004, + "step": 5526 + }, + { + "epoch": 3.9755439669124257, + "grad_norm": 1.8102600041321733, + "learning_rate": 1.983863761448128e-06, + "loss": 0.0093, + "step": 5527 + }, + { + "epoch": 3.9762632620032368, + "grad_norm": 3.3998105507313645, + "learning_rate": 1.983541692560473e-06, + "loss": 0.1575, + "step": 5528 + }, + { + "epoch": 3.976982557094048, + "grad_norm": 0.1187107860899517, + "learning_rate": 1.983219598793748e-06, + "loss": 0.0003, + "step": 5529 + }, + { + "epoch": 3.977701852184859, + "grad_norm": 3.8954253182107856, + "learning_rate": 1.9828974801645267e-06, + "loss": 0.0753, + "step": 5530 + }, + { + "epoch": 3.97842114727567, + "grad_norm": 2.8724925557565526, + "learning_rate": 1.982575336689383e-06, + "loss": 0.0048, + "step": 5531 + }, + { + "epoch": 3.979140442366481, + "grad_norm": 1.2656307996880056, + "learning_rate": 1.9822531683848905e-06, + "loss": 0.0318, + "step": 5532 + }, + { + "epoch": 3.979859737457292, + "grad_norm": 3.651805533113786, + "learning_rate": 1.981930975267626e-06, + "loss": 0.2099, + "step": 5533 + }, + { + "epoch": 3.980579032548103, + "grad_norm": 3.3151630758223485, + "learning_rate": 1.9816087573541666e-06, + "loss": 0.0897, + "step": 5534 + }, + { + "epoch": 3.981298327638914, + "grad_norm": 1.4104633709024037, + "learning_rate": 1.9812865146610914e-06, + "loss": 0.017, + "step": 5535 + }, + { + "epoch": 3.9820176227297246, + "grad_norm": 1.6821698223744563, + "learning_rate": 1.98096424720498e-06, + "loss": 0.0307, + "step": 5536 + }, + { + "epoch": 3.982736917820536, + "grad_norm": 2.686006522946958, + "learning_rate": 1.980641955002414e-06, + "loss": 0.0824, + "step": 5537 + }, + { + "epoch": 3.9834562129113467, + "grad_norm": 4.747025221513913, + "learning_rate": 1.980319638069976e-06, + "loss": 0.103, + "step": 5538 + }, + { + "epoch": 3.984175508002158, + "grad_norm": 4.919956651355761, + "learning_rate": 1.9799972964242488e-06, + "loss": 0.0613, + "step": 5539 + }, + { + "epoch": 3.9848948030929687, + "grad_norm": 1.7178862299974536, + "learning_rate": 1.9796749300818185e-06, + "loss": 0.0512, + "step": 5540 + }, + { + "epoch": 3.9856140981837798, + "grad_norm": 2.4362320721141986, + "learning_rate": 1.9793525390592707e-06, + "loss": 0.07, + "step": 5541 + }, + { + "epoch": 3.986333393274591, + "grad_norm": 6.6175357797327, + "learning_rate": 1.9790301233731935e-06, + "loss": 0.0262, + "step": 5542 + }, + { + "epoch": 3.987052688365402, + "grad_norm": 3.2068082300256338, + "learning_rate": 1.9787076830401754e-06, + "loss": 0.08, + "step": 5543 + }, + { + "epoch": 3.987771983456213, + "grad_norm": 3.2601021243582715, + "learning_rate": 1.9783852180768063e-06, + "loss": 0.0954, + "step": 5544 + }, + { + "epoch": 3.988491278547024, + "grad_norm": 4.322350908093845, + "learning_rate": 1.9780627284996788e-06, + "loss": 0.1198, + "step": 5545 + }, + { + "epoch": 3.989210573637835, + "grad_norm": 3.7629266223351703, + "learning_rate": 1.9777402143253844e-06, + "loss": 0.0986, + "step": 5546 + }, + { + "epoch": 3.989929868728646, + "grad_norm": 2.6575151080998656, + "learning_rate": 1.9774176755705175e-06, + "loss": 0.0057, + "step": 5547 + }, + { + "epoch": 3.990649163819457, + "grad_norm": 1.7785944523180859, + "learning_rate": 1.9770951122516723e-06, + "loss": 0.0497, + "step": 5548 + }, + { + "epoch": 3.991368458910268, + "grad_norm": 2.574852126934876, + "learning_rate": 1.9767725243854467e-06, + "loss": 0.0589, + "step": 5549 + }, + { + "epoch": 3.992087754001079, + "grad_norm": 3.4478338141270917, + "learning_rate": 1.9764499119884376e-06, + "loss": 0.0443, + "step": 5550 + }, + { + "epoch": 3.99280704909189, + "grad_norm": 4.4453764871797015, + "learning_rate": 1.9761272750772443e-06, + "loss": 0.134, + "step": 5551 + }, + { + "epoch": 3.993526344182701, + "grad_norm": 4.547120628209831, + "learning_rate": 1.975804613668467e-06, + "loss": 0.0804, + "step": 5552 + }, + { + "epoch": 3.9942456392735117, + "grad_norm": 0.2542132292298775, + "learning_rate": 1.9754819277787068e-06, + "loss": 0.0007, + "step": 5553 + }, + { + "epoch": 3.994964934364323, + "grad_norm": 0.10799028775402214, + "learning_rate": 1.9751592174245667e-06, + "loss": 0.0002, + "step": 5554 + }, + { + "epoch": 3.995684229455134, + "grad_norm": 2.232774432659303, + "learning_rate": 1.974836482622651e-06, + "loss": 0.0595, + "step": 5555 + }, + { + "epoch": 3.996403524545945, + "grad_norm": 1.2822246337843277, + "learning_rate": 1.9745137233895647e-06, + "loss": 0.019, + "step": 5556 + }, + { + "epoch": 3.997122819636756, + "grad_norm": 0.9181648293468428, + "learning_rate": 1.974190939741914e-06, + "loss": 0.0025, + "step": 5557 + }, + { + "epoch": 3.997842114727567, + "grad_norm": 5.192714592596067, + "learning_rate": 1.9738681316963072e-06, + "loss": 0.1684, + "step": 5558 + }, + { + "epoch": 3.998561409818378, + "grad_norm": 2.462070020802091, + "learning_rate": 1.973545299269353e-06, + "loss": 0.0732, + "step": 5559 + }, + { + "epoch": 3.999280704909189, + "grad_norm": 3.446112710138286, + "learning_rate": 1.973222442477662e-06, + "loss": 0.1721, + "step": 5560 + }, + { + "epoch": 4.0, + "grad_norm": 3.125009306920019, + "learning_rate": 1.972899561337846e-06, + "loss": 0.0753, + "step": 5561 + }, + { + "epoch": 4.000719295090811, + "grad_norm": 2.6378329080687677, + "learning_rate": 1.972576655866517e-06, + "loss": 0.055, + "step": 5562 + }, + { + "epoch": 4.001438590181622, + "grad_norm": 1.3124484700191574, + "learning_rate": 1.97225372608029e-06, + "loss": 0.0188, + "step": 5563 + }, + { + "epoch": 4.002157885272433, + "grad_norm": 5.749959310615003, + "learning_rate": 1.9719307719957788e-06, + "loss": 0.0276, + "step": 5564 + }, + { + "epoch": 4.002877180363244, + "grad_norm": 0.8413148471638774, + "learning_rate": 1.971607793629602e-06, + "loss": 0.0071, + "step": 5565 + }, + { + "epoch": 4.003596475454055, + "grad_norm": 3.041657502103039, + "learning_rate": 1.971284790998376e-06, + "loss": 0.0933, + "step": 5566 + }, + { + "epoch": 4.004315770544866, + "grad_norm": 2.4179210471451675, + "learning_rate": 1.9709617641187198e-06, + "loss": 0.0798, + "step": 5567 + }, + { + "epoch": 4.005035065635677, + "grad_norm": 2.4335881403936965, + "learning_rate": 1.970638713007255e-06, + "loss": 0.0293, + "step": 5568 + }, + { + "epoch": 4.005754360726488, + "grad_norm": 2.145058812012006, + "learning_rate": 1.970315637680601e-06, + "loss": 0.0401, + "step": 5569 + }, + { + "epoch": 4.006473655817299, + "grad_norm": 4.200999163498747, + "learning_rate": 1.9699925381553823e-06, + "loss": 0.1309, + "step": 5570 + }, + { + "epoch": 4.00719295090811, + "grad_norm": 4.5144827620981545, + "learning_rate": 1.9696694144482225e-06, + "loss": 0.0954, + "step": 5571 + }, + { + "epoch": 4.007912245998921, + "grad_norm": 4.724415125781821, + "learning_rate": 1.9693462665757468e-06, + "loss": 0.2418, + "step": 5572 + }, + { + "epoch": 4.008631541089732, + "grad_norm": 2.6591619743355612, + "learning_rate": 1.969023094554582e-06, + "loss": 0.0589, + "step": 5573 + }, + { + "epoch": 4.009350836180543, + "grad_norm": 1.0779182576278026, + "learning_rate": 1.9686998984013557e-06, + "loss": 0.0123, + "step": 5574 + }, + { + "epoch": 4.0100701312713545, + "grad_norm": 1.6729116541856488, + "learning_rate": 1.9683766781326966e-06, + "loss": 0.005, + "step": 5575 + }, + { + "epoch": 4.010789426362165, + "grad_norm": 4.595174106434808, + "learning_rate": 1.9680534337652356e-06, + "loss": 0.1569, + "step": 5576 + }, + { + "epoch": 4.011508721452976, + "grad_norm": 1.842382582600284, + "learning_rate": 1.967730165315603e-06, + "loss": 0.0058, + "step": 5577 + }, + { + "epoch": 4.012228016543787, + "grad_norm": 1.6959115599835204, + "learning_rate": 1.967406872800433e-06, + "loss": 0.0312, + "step": 5578 + }, + { + "epoch": 4.012947311634598, + "grad_norm": 2.873988296230702, + "learning_rate": 1.967083556236359e-06, + "loss": 0.0584, + "step": 5579 + }, + { + "epoch": 4.013666606725409, + "grad_norm": 0.12158542518542123, + "learning_rate": 1.9667602156400153e-06, + "loss": 0.0005, + "step": 5580 + }, + { + "epoch": 4.01438590181622, + "grad_norm": 1.1527853623371864, + "learning_rate": 1.9664368510280396e-06, + "loss": 0.0205, + "step": 5581 + }, + { + "epoch": 4.015105196907031, + "grad_norm": 0.05348946704594314, + "learning_rate": 1.9661134624170693e-06, + "loss": 0.0002, + "step": 5582 + }, + { + "epoch": 4.015824491997842, + "grad_norm": 1.491862473431524, + "learning_rate": 1.965790049823743e-06, + "loss": 0.0146, + "step": 5583 + }, + { + "epoch": 4.016543787088653, + "grad_norm": 2.305600855218038, + "learning_rate": 1.965466613264701e-06, + "loss": 0.0497, + "step": 5584 + }, + { + "epoch": 4.017263082179464, + "grad_norm": 1.9988928413818747, + "learning_rate": 1.965143152756584e-06, + "loss": 0.0454, + "step": 5585 + }, + { + "epoch": 4.017982377270275, + "grad_norm": 2.343120503460134, + "learning_rate": 1.964819668316036e-06, + "loss": 0.017, + "step": 5586 + }, + { + "epoch": 4.018701672361086, + "grad_norm": 2.3935095498709207, + "learning_rate": 1.9644961599597e-06, + "loss": 0.0543, + "step": 5587 + }, + { + "epoch": 4.0194209674518975, + "grad_norm": 2.68691427681927, + "learning_rate": 1.9641726277042207e-06, + "loss": 0.0335, + "step": 5588 + }, + { + "epoch": 4.020140262542708, + "grad_norm": 8.696255935310184, + "learning_rate": 1.9638490715662457e-06, + "loss": 0.0656, + "step": 5589 + }, + { + "epoch": 4.0208595576335195, + "grad_norm": 6.538491580371057, + "learning_rate": 1.963525491562421e-06, + "loss": 0.1177, + "step": 5590 + }, + { + "epoch": 4.02157885272433, + "grad_norm": 2.220776289030348, + "learning_rate": 1.9632018877093963e-06, + "loss": 0.0452, + "step": 5591 + }, + { + "epoch": 4.022298147815141, + "grad_norm": 5.138264696916617, + "learning_rate": 1.9628782600238217e-06, + "loss": 0.1107, + "step": 5592 + }, + { + "epoch": 4.023017442905952, + "grad_norm": 2.5046389175062376, + "learning_rate": 1.962554608522348e-06, + "loss": 0.0501, + "step": 5593 + }, + { + "epoch": 4.023736737996763, + "grad_norm": 4.2118768547335454, + "learning_rate": 1.9622309332216273e-06, + "loss": 0.1112, + "step": 5594 + }, + { + "epoch": 4.024456033087574, + "grad_norm": 3.72786561692059, + "learning_rate": 1.9619072341383137e-06, + "loss": 0.0599, + "step": 5595 + }, + { + "epoch": 4.025175328178385, + "grad_norm": 5.391119538834329, + "learning_rate": 1.961583511289062e-06, + "loss": 0.0761, + "step": 5596 + }, + { + "epoch": 4.025894623269196, + "grad_norm": 7.639126852141957, + "learning_rate": 1.961259764690529e-06, + "loss": 0.1354, + "step": 5597 + }, + { + "epoch": 4.026613918360007, + "grad_norm": 0.06409968785285923, + "learning_rate": 1.9609359943593707e-06, + "loss": 0.0003, + "step": 5598 + }, + { + "epoch": 4.027333213450818, + "grad_norm": 3.583892351621556, + "learning_rate": 1.9606122003122465e-06, + "loss": 0.0941, + "step": 5599 + }, + { + "epoch": 4.028052508541629, + "grad_norm": 2.7854228264044107, + "learning_rate": 1.960288382565816e-06, + "loss": 0.0463, + "step": 5600 + }, + { + "epoch": 4.0287718036324405, + "grad_norm": 2.083117794050754, + "learning_rate": 1.9599645411367402e-06, + "loss": 0.0579, + "step": 5601 + }, + { + "epoch": 4.029491098723251, + "grad_norm": 2.7501864308626143, + "learning_rate": 1.9596406760416817e-06, + "loss": 0.0293, + "step": 5602 + }, + { + "epoch": 4.0302103938140625, + "grad_norm": 1.1640114124703, + "learning_rate": 1.9593167872973027e-06, + "loss": 0.0083, + "step": 5603 + }, + { + "epoch": 4.030929688904873, + "grad_norm": 1.878776693285718, + "learning_rate": 1.95899287492027e-06, + "loss": 0.0297, + "step": 5604 + }, + { + "epoch": 4.031648983995685, + "grad_norm": 1.3986451602613668, + "learning_rate": 1.958668938927247e-06, + "loss": 0.0236, + "step": 5605 + }, + { + "epoch": 4.032368279086495, + "grad_norm": 0.29249987724659526, + "learning_rate": 1.958344979334902e-06, + "loss": 0.0006, + "step": 5606 + }, + { + "epoch": 4.033087574177307, + "grad_norm": 1.860735340719085, + "learning_rate": 1.958020996159904e-06, + "loss": 0.0719, + "step": 5607 + }, + { + "epoch": 4.033806869268117, + "grad_norm": 1.688248562133601, + "learning_rate": 1.957696989418921e-06, + "loss": 0.0214, + "step": 5608 + }, + { + "epoch": 4.034526164358928, + "grad_norm": 2.3284785064262774, + "learning_rate": 1.9573729591286244e-06, + "loss": 0.043, + "step": 5609 + }, + { + "epoch": 4.035245459449739, + "grad_norm": 0.8663610535166623, + "learning_rate": 1.957048905305687e-06, + "loss": 0.0027, + "step": 5610 + }, + { + "epoch": 4.03596475454055, + "grad_norm": 3.16079870202343, + "learning_rate": 1.9567248279667803e-06, + "loss": 0.02, + "step": 5611 + }, + { + "epoch": 4.036684049631361, + "grad_norm": 5.902110947687226, + "learning_rate": 1.95640072712858e-06, + "loss": 0.0741, + "step": 5612 + }, + { + "epoch": 4.037403344722172, + "grad_norm": 2.169710407079322, + "learning_rate": 1.9560766028077605e-06, + "loss": 0.0411, + "step": 5613 + }, + { + "epoch": 4.0381226398129835, + "grad_norm": 3.712037426629766, + "learning_rate": 1.955752455020999e-06, + "loss": 0.1177, + "step": 5614 + }, + { + "epoch": 4.038841934903794, + "grad_norm": 4.0868971910803396, + "learning_rate": 1.955428283784975e-06, + "loss": 0.0967, + "step": 5615 + }, + { + "epoch": 4.0395612299946055, + "grad_norm": 2.365990311307959, + "learning_rate": 1.9551040891163646e-06, + "loss": 0.0626, + "step": 5616 + }, + { + "epoch": 4.040280525085416, + "grad_norm": 3.742750749092191, + "learning_rate": 1.954779871031851e-06, + "loss": 0.0922, + "step": 5617 + }, + { + "epoch": 4.040999820176228, + "grad_norm": 3.386251878317679, + "learning_rate": 1.954455629548114e-06, + "loss": 0.0556, + "step": 5618 + }, + { + "epoch": 4.041719115267038, + "grad_norm": 2.6895062259425755, + "learning_rate": 1.9541313646818377e-06, + "loss": 0.0076, + "step": 5619 + }, + { + "epoch": 4.04243841035785, + "grad_norm": 4.243463543536009, + "learning_rate": 1.9538070764497055e-06, + "loss": 0.1038, + "step": 5620 + }, + { + "epoch": 4.04315770544866, + "grad_norm": 2.5273240952313425, + "learning_rate": 1.9534827648684025e-06, + "loss": 0.0915, + "step": 5621 + }, + { + "epoch": 4.043877000539472, + "grad_norm": 3.363189608904149, + "learning_rate": 1.9531584299546147e-06, + "loss": 0.1004, + "step": 5622 + }, + { + "epoch": 4.044596295630282, + "grad_norm": 2.46851735037571, + "learning_rate": 1.9528340717250308e-06, + "loss": 0.0209, + "step": 5623 + }, + { + "epoch": 4.045315590721093, + "grad_norm": 5.802639854812304, + "learning_rate": 1.952509690196338e-06, + "loss": 0.0939, + "step": 5624 + }, + { + "epoch": 4.046034885811904, + "grad_norm": 2.463264582425902, + "learning_rate": 1.952185285385228e-06, + "loss": 0.0839, + "step": 5625 + }, + { + "epoch": 4.046754180902715, + "grad_norm": 1.252750745089526, + "learning_rate": 1.9518608573083917e-06, + "loss": 0.0206, + "step": 5626 + }, + { + "epoch": 4.0474734759935265, + "grad_norm": 2.7013417513386786, + "learning_rate": 1.951536405982521e-06, + "loss": 0.0821, + "step": 5627 + }, + { + "epoch": 4.048192771084337, + "grad_norm": 2.225886095945009, + "learning_rate": 1.9512119314243087e-06, + "loss": 0.039, + "step": 5628 + }, + { + "epoch": 4.0489120661751485, + "grad_norm": 4.393532538789844, + "learning_rate": 1.9508874336504513e-06, + "loss": 0.12, + "step": 5629 + }, + { + "epoch": 4.049631361265959, + "grad_norm": 3.148495664952311, + "learning_rate": 1.9505629126776434e-06, + "loss": 0.1035, + "step": 5630 + }, + { + "epoch": 4.050350656356771, + "grad_norm": 3.7889871180372903, + "learning_rate": 1.9502383685225827e-06, + "loss": 0.108, + "step": 5631 + }, + { + "epoch": 4.051069951447581, + "grad_norm": 3.6837635385824687, + "learning_rate": 1.9499138012019675e-06, + "loss": 0.0724, + "step": 5632 + }, + { + "epoch": 4.051789246538393, + "grad_norm": 7.344244243443534, + "learning_rate": 1.9495892107324984e-06, + "loss": 0.0467, + "step": 5633 + }, + { + "epoch": 4.052508541629203, + "grad_norm": 0.3736425703145218, + "learning_rate": 1.9492645971308746e-06, + "loss": 0.0012, + "step": 5634 + }, + { + "epoch": 4.053227836720015, + "grad_norm": 0.6428946257915615, + "learning_rate": 1.9489399604137982e-06, + "loss": 0.0051, + "step": 5635 + }, + { + "epoch": 4.053947131810825, + "grad_norm": 4.885741985291091, + "learning_rate": 1.9486153005979734e-06, + "loss": 0.0515, + "step": 5636 + }, + { + "epoch": 4.054666426901637, + "grad_norm": 4.502329247918408, + "learning_rate": 1.948290617700103e-06, + "loss": 0.0467, + "step": 5637 + }, + { + "epoch": 4.055385721992447, + "grad_norm": 2.4420964393489992, + "learning_rate": 1.947965911736894e-06, + "loss": 0.0589, + "step": 5638 + }, + { + "epoch": 4.056105017083258, + "grad_norm": 2.90201432128634, + "learning_rate": 1.9476411827250525e-06, + "loss": 0.0433, + "step": 5639 + }, + { + "epoch": 4.0568243121740695, + "grad_norm": 3.0729440647269755, + "learning_rate": 1.9473164306812865e-06, + "loss": 0.0613, + "step": 5640 + }, + { + "epoch": 4.05754360726488, + "grad_norm": 1.2744657231365868, + "learning_rate": 1.946991655622304e-06, + "loss": 0.0186, + "step": 5641 + }, + { + "epoch": 4.0582629023556915, + "grad_norm": 4.358182949560074, + "learning_rate": 1.946666857564817e-06, + "loss": 0.1559, + "step": 5642 + }, + { + "epoch": 4.058982197446502, + "grad_norm": 0.06422802928990409, + "learning_rate": 1.9463420365255356e-06, + "loss": 0.0002, + "step": 5643 + }, + { + "epoch": 4.059701492537314, + "grad_norm": 2.066484177909009, + "learning_rate": 1.9460171925211732e-06, + "loss": 0.0175, + "step": 5644 + }, + { + "epoch": 4.060420787628124, + "grad_norm": 2.494214092731577, + "learning_rate": 1.945692325568443e-06, + "loss": 0.0812, + "step": 5645 + }, + { + "epoch": 4.061140082718936, + "grad_norm": 4.988892662843045, + "learning_rate": 1.9453674356840606e-06, + "loss": 0.163, + "step": 5646 + }, + { + "epoch": 4.061859377809746, + "grad_norm": 2.0396114713602085, + "learning_rate": 1.945042522884742e-06, + "loss": 0.0308, + "step": 5647 + }, + { + "epoch": 4.062578672900558, + "grad_norm": 5.301394066579486, + "learning_rate": 1.9447175871872038e-06, + "loss": 0.0528, + "step": 5648 + }, + { + "epoch": 4.063297967991368, + "grad_norm": 1.191959399670513, + "learning_rate": 1.9443926286081653e-06, + "loss": 0.0085, + "step": 5649 + }, + { + "epoch": 4.06401726308218, + "grad_norm": 4.3484991399552735, + "learning_rate": 1.944067647164346e-06, + "loss": 0.0543, + "step": 5650 + }, + { + "epoch": 4.06473655817299, + "grad_norm": 2.0216968272426, + "learning_rate": 1.943742642872467e-06, + "loss": 0.0036, + "step": 5651 + }, + { + "epoch": 4.065455853263802, + "grad_norm": 1.622868540748743, + "learning_rate": 1.9434176157492493e-06, + "loss": 0.0314, + "step": 5652 + }, + { + "epoch": 4.0661751483546125, + "grad_norm": 2.6696127483776064, + "learning_rate": 1.943092565811417e-06, + "loss": 0.0448, + "step": 5653 + }, + { + "epoch": 4.066894443445423, + "grad_norm": 4.418356814156999, + "learning_rate": 1.942767493075695e-06, + "loss": 0.1277, + "step": 5654 + }, + { + "epoch": 4.0676137385362345, + "grad_norm": 1.8996965975735507, + "learning_rate": 1.9424423975588077e-06, + "loss": 0.0497, + "step": 5655 + }, + { + "epoch": 4.068333033627045, + "grad_norm": 5.176788770343264, + "learning_rate": 1.942117279277483e-06, + "loss": 0.0981, + "step": 5656 + }, + { + "epoch": 4.069052328717857, + "grad_norm": 0.82918157616306, + "learning_rate": 1.9417921382484477e-06, + "loss": 0.0081, + "step": 5657 + }, + { + "epoch": 4.069771623808667, + "grad_norm": 1.3135293122632894, + "learning_rate": 1.9414669744884308e-06, + "loss": 0.0139, + "step": 5658 + }, + { + "epoch": 4.070490918899479, + "grad_norm": 0.21854786545779445, + "learning_rate": 1.9411417880141636e-06, + "loss": 0.001, + "step": 5659 + }, + { + "epoch": 4.071210213990289, + "grad_norm": 0.4243846743794653, + "learning_rate": 1.9408165788423776e-06, + "loss": 0.0008, + "step": 5660 + }, + { + "epoch": 4.071929509081101, + "grad_norm": 2.591837679663566, + "learning_rate": 1.9404913469898038e-06, + "loss": 0.066, + "step": 5661 + }, + { + "epoch": 4.072648804171911, + "grad_norm": 2.5820776624268245, + "learning_rate": 1.940166092473177e-06, + "loss": 0.0618, + "step": 5662 + }, + { + "epoch": 4.073368099262723, + "grad_norm": 0.602460428979046, + "learning_rate": 1.9398408153092335e-06, + "loss": 0.0009, + "step": 5663 + }, + { + "epoch": 4.074087394353533, + "grad_norm": 2.3803415860441675, + "learning_rate": 1.9395155155147062e-06, + "loss": 0.0339, + "step": 5664 + }, + { + "epoch": 4.074806689444345, + "grad_norm": 1.9659281112207505, + "learning_rate": 1.939190193106335e-06, + "loss": 0.0203, + "step": 5665 + }, + { + "epoch": 4.0755259845351555, + "grad_norm": 2.9168498236125378, + "learning_rate": 1.938864848100857e-06, + "loss": 0.0536, + "step": 5666 + }, + { + "epoch": 4.076245279625967, + "grad_norm": 0.10630146339074485, + "learning_rate": 1.9385394805150123e-06, + "loss": 0.0004, + "step": 5667 + }, + { + "epoch": 4.0769645747167775, + "grad_norm": 3.705895058280914, + "learning_rate": 1.9382140903655413e-06, + "loss": 0.0826, + "step": 5668 + }, + { + "epoch": 4.077683869807588, + "grad_norm": 1.7981800147154507, + "learning_rate": 1.9378886776691865e-06, + "loss": 0.0202, + "step": 5669 + }, + { + "epoch": 4.0784031648984, + "grad_norm": 2.6291159673191924, + "learning_rate": 1.9375632424426905e-06, + "loss": 0.074, + "step": 5670 + }, + { + "epoch": 4.07912245998921, + "grad_norm": 2.3021008339901425, + "learning_rate": 1.937237784702797e-06, + "loss": 0.0072, + "step": 5671 + }, + { + "epoch": 4.079841755080022, + "grad_norm": 4.321209741329555, + "learning_rate": 1.9369123044662525e-06, + "loss": 0.125, + "step": 5672 + }, + { + "epoch": 4.080561050170832, + "grad_norm": 2.7554739237603756, + "learning_rate": 1.936586801749803e-06, + "loss": 0.0175, + "step": 5673 + }, + { + "epoch": 4.081280345261644, + "grad_norm": 3.3073925265832327, + "learning_rate": 1.9362612765701956e-06, + "loss": 0.0892, + "step": 5674 + }, + { + "epoch": 4.081999640352454, + "grad_norm": 4.476785908571246, + "learning_rate": 1.93593572894418e-06, + "loss": 0.1484, + "step": 5675 + }, + { + "epoch": 4.082718935443266, + "grad_norm": 1.2719691137816835, + "learning_rate": 1.935610158888506e-06, + "loss": 0.0311, + "step": 5676 + }, + { + "epoch": 4.083438230534076, + "grad_norm": 3.137004101216999, + "learning_rate": 1.9352845664199247e-06, + "loss": 0.0616, + "step": 5677 + }, + { + "epoch": 4.084157525624888, + "grad_norm": 0.2326699076235698, + "learning_rate": 1.934958951555188e-06, + "loss": 0.001, + "step": 5678 + }, + { + "epoch": 4.0848768207156985, + "grad_norm": 3.917584111006002, + "learning_rate": 1.93463331431105e-06, + "loss": 0.1312, + "step": 5679 + }, + { + "epoch": 4.08559611580651, + "grad_norm": 2.21227847155993, + "learning_rate": 1.934307654704265e-06, + "loss": 0.0436, + "step": 5680 + }, + { + "epoch": 4.0863154108973205, + "grad_norm": 5.431917649920485, + "learning_rate": 1.9339819727515887e-06, + "loss": 0.104, + "step": 5681 + }, + { + "epoch": 4.087034705988132, + "grad_norm": 1.421417514885958, + "learning_rate": 1.933656268469778e-06, + "loss": 0.0252, + "step": 5682 + }, + { + "epoch": 4.087754001078943, + "grad_norm": 1.9702361740038556, + "learning_rate": 1.933330541875591e-06, + "loss": 0.0534, + "step": 5683 + }, + { + "epoch": 4.088473296169754, + "grad_norm": 2.6788091609012468, + "learning_rate": 1.933004792985787e-06, + "loss": 0.0787, + "step": 5684 + }, + { + "epoch": 4.089192591260565, + "grad_norm": 3.0010032305261767, + "learning_rate": 1.932679021817126e-06, + "loss": 0.0851, + "step": 5685 + }, + { + "epoch": 4.089911886351375, + "grad_norm": 3.2357560370662592, + "learning_rate": 1.9323532283863703e-06, + "loss": 0.1042, + "step": 5686 + }, + { + "epoch": 4.090631181442187, + "grad_norm": 3.389784680269663, + "learning_rate": 1.9320274127102814e-06, + "loss": 0.028, + "step": 5687 + }, + { + "epoch": 4.091350476532997, + "grad_norm": 2.329511942048084, + "learning_rate": 1.9317015748056245e-06, + "loss": 0.0223, + "step": 5688 + }, + { + "epoch": 4.092069771623809, + "grad_norm": 2.199674307479924, + "learning_rate": 1.931375714689163e-06, + "loss": 0.0052, + "step": 5689 + }, + { + "epoch": 4.092789066714619, + "grad_norm": 6.643773508790654, + "learning_rate": 1.931049832377664e-06, + "loss": 0.1457, + "step": 5690 + }, + { + "epoch": 4.093508361805431, + "grad_norm": 4.0315699815198265, + "learning_rate": 1.9307239278878947e-06, + "loss": 0.0667, + "step": 5691 + }, + { + "epoch": 4.0942276568962415, + "grad_norm": 3.218335873877121, + "learning_rate": 1.9303980012366233e-06, + "loss": 0.0293, + "step": 5692 + }, + { + "epoch": 4.094946951987053, + "grad_norm": 1.635739859368482, + "learning_rate": 1.9300720524406187e-06, + "loss": 0.0033, + "step": 5693 + }, + { + "epoch": 4.0956662470778635, + "grad_norm": 1.23582640770648, + "learning_rate": 1.929746081516652e-06, + "loss": 0.0126, + "step": 5694 + }, + { + "epoch": 4.096385542168675, + "grad_norm": 2.074867088725708, + "learning_rate": 1.929420088481495e-06, + "loss": 0.0333, + "step": 5695 + }, + { + "epoch": 4.097104837259486, + "grad_norm": 5.638993583096589, + "learning_rate": 1.9290940733519213e-06, + "loss": 0.1638, + "step": 5696 + }, + { + "epoch": 4.097824132350297, + "grad_norm": 1.2256146513871569, + "learning_rate": 1.928768036144704e-06, + "loss": 0.0135, + "step": 5697 + }, + { + "epoch": 4.098543427441108, + "grad_norm": 4.11357022039547, + "learning_rate": 1.928441976876618e-06, + "loss": 0.0287, + "step": 5698 + }, + { + "epoch": 4.099262722531919, + "grad_norm": 0.04131071736499514, + "learning_rate": 1.9281158955644407e-06, + "loss": 0.0001, + "step": 5699 + }, + { + "epoch": 4.09998201762273, + "grad_norm": 3.368450421827848, + "learning_rate": 1.927789792224949e-06, + "loss": 0.0546, + "step": 5700 + }, + { + "epoch": 4.10070131271354, + "grad_norm": 3.3013483282626472, + "learning_rate": 1.9274636668749214e-06, + "loss": 0.0849, + "step": 5701 + }, + { + "epoch": 4.101420607804352, + "grad_norm": 3.262689128940304, + "learning_rate": 1.927137519531138e-06, + "loss": 0.0803, + "step": 5702 + }, + { + "epoch": 4.102139902895162, + "grad_norm": 3.784835000847464, + "learning_rate": 1.926811350210379e-06, + "loss": 0.0825, + "step": 5703 + }, + { + "epoch": 4.102859197985974, + "grad_norm": 3.1038381341836696, + "learning_rate": 1.926485158929427e-06, + "loss": 0.0747, + "step": 5704 + }, + { + "epoch": 4.1035784930767845, + "grad_norm": 3.4392822328910233, + "learning_rate": 1.9261589457050646e-06, + "loss": 0.0693, + "step": 5705 + }, + { + "epoch": 4.104297788167596, + "grad_norm": 0.4117083014027817, + "learning_rate": 1.925832710554077e-06, + "loss": 0.0013, + "step": 5706 + }, + { + "epoch": 4.1050170832584065, + "grad_norm": 3.2929148839140354, + "learning_rate": 1.9255064534932485e-06, + "loss": 0.0625, + "step": 5707 + }, + { + "epoch": 4.105736378349218, + "grad_norm": 2.5743850171932516, + "learning_rate": 1.925180174539366e-06, + "loss": 0.0301, + "step": 5708 + }, + { + "epoch": 4.106455673440029, + "grad_norm": 4.184311637130611, + "learning_rate": 1.924853873709217e-06, + "loss": 0.0471, + "step": 5709 + }, + { + "epoch": 4.10717496853084, + "grad_norm": 0.18877509491201283, + "learning_rate": 1.924527551019591e-06, + "loss": 0.0011, + "step": 5710 + }, + { + "epoch": 4.107894263621651, + "grad_norm": 0.6935606346084154, + "learning_rate": 1.9242012064872765e-06, + "loss": 0.007, + "step": 5711 + }, + { + "epoch": 4.108613558712462, + "grad_norm": 5.037396147543799, + "learning_rate": 1.9238748401290655e-06, + "loss": 0.065, + "step": 5712 + }, + { + "epoch": 4.109332853803273, + "grad_norm": 4.179762704144928, + "learning_rate": 1.92354845196175e-06, + "loss": 0.0905, + "step": 5713 + }, + { + "epoch": 4.110052148894084, + "grad_norm": 3.3334325117631627, + "learning_rate": 1.9232220420021237e-06, + "loss": 0.0064, + "step": 5714 + }, + { + "epoch": 4.110771443984895, + "grad_norm": 2.537348670167864, + "learning_rate": 1.9228956102669795e-06, + "loss": 0.0636, + "step": 5715 + }, + { + "epoch": 4.111490739075705, + "grad_norm": 7.133358547703262, + "learning_rate": 1.9225691567731144e-06, + "loss": 0.1233, + "step": 5716 + }, + { + "epoch": 4.112210034166517, + "grad_norm": 7.2006027618131085, + "learning_rate": 1.922242681537324e-06, + "loss": 0.158, + "step": 5717 + }, + { + "epoch": 4.1129293292573275, + "grad_norm": 7.8498605224873685, + "learning_rate": 1.921916184576407e-06, + "loss": 0.1387, + "step": 5718 + }, + { + "epoch": 4.113648624348139, + "grad_norm": 0.18298045334854587, + "learning_rate": 1.9215896659071605e-06, + "loss": 0.0002, + "step": 5719 + }, + { + "epoch": 4.1143679194389495, + "grad_norm": 2.0065236673455535, + "learning_rate": 1.9212631255463863e-06, + "loss": 0.0148, + "step": 5720 + }, + { + "epoch": 4.115087214529761, + "grad_norm": 0.17850466987159372, + "learning_rate": 1.920936563510885e-06, + "loss": 0.0005, + "step": 5721 + }, + { + "epoch": 4.115806509620572, + "grad_norm": 1.0367893465812918, + "learning_rate": 1.920609979817458e-06, + "loss": 0.013, + "step": 5722 + }, + { + "epoch": 4.116525804711383, + "grad_norm": 3.9852943048876694, + "learning_rate": 1.92028337448291e-06, + "loss": 0.0854, + "step": 5723 + }, + { + "epoch": 4.117245099802194, + "grad_norm": 1.90017075813126, + "learning_rate": 1.9199567475240438e-06, + "loss": 0.0321, + "step": 5724 + }, + { + "epoch": 4.117964394893005, + "grad_norm": 4.331468621034906, + "learning_rate": 1.919630098957666e-06, + "loss": 0.0756, + "step": 5725 + }, + { + "epoch": 4.118683689983816, + "grad_norm": 2.3326702374836334, + "learning_rate": 1.9193034288005835e-06, + "loss": 0.0327, + "step": 5726 + }, + { + "epoch": 4.119402985074627, + "grad_norm": 3.2480675535863095, + "learning_rate": 1.9189767370696024e-06, + "loss": 0.0099, + "step": 5727 + }, + { + "epoch": 4.120122280165438, + "grad_norm": 1.769670056186677, + "learning_rate": 1.918650023781534e-06, + "loss": 0.0315, + "step": 5728 + }, + { + "epoch": 4.120841575256249, + "grad_norm": 7.538363905925796, + "learning_rate": 1.9183232889531858e-06, + "loss": 0.1714, + "step": 5729 + }, + { + "epoch": 4.12156087034706, + "grad_norm": 2.9432531095076366, + "learning_rate": 1.9179965326013707e-06, + "loss": 0.0472, + "step": 5730 + }, + { + "epoch": 4.1222801654378705, + "grad_norm": 0.07088180399235038, + "learning_rate": 1.9176697547429e-06, + "loss": 0.0003, + "step": 5731 + }, + { + "epoch": 4.122999460528682, + "grad_norm": 1.1241375101194537, + "learning_rate": 1.9173429553945872e-06, + "loss": 0.0044, + "step": 5732 + }, + { + "epoch": 4.1237187556194925, + "grad_norm": 5.106520940121957, + "learning_rate": 1.917016134573247e-06, + "loss": 0.132, + "step": 5733 + }, + { + "epoch": 4.124438050710304, + "grad_norm": 4.907589544477937, + "learning_rate": 1.916689292295694e-06, + "loss": 0.0787, + "step": 5734 + }, + { + "epoch": 4.125157345801115, + "grad_norm": 2.9636857249696362, + "learning_rate": 1.9163624285787463e-06, + "loss": 0.0659, + "step": 5735 + }, + { + "epoch": 4.125876640891926, + "grad_norm": 2.5711356098861686, + "learning_rate": 1.9160355434392207e-06, + "loss": 0.0295, + "step": 5736 + }, + { + "epoch": 4.126595935982737, + "grad_norm": 3.076514978360546, + "learning_rate": 1.9157086368939354e-06, + "loss": 0.0821, + "step": 5737 + }, + { + "epoch": 4.127315231073548, + "grad_norm": 0.01007448110292288, + "learning_rate": 1.9153817089597112e-06, + "loss": 0.0001, + "step": 5738 + }, + { + "epoch": 4.128034526164359, + "grad_norm": 0.7631736990912048, + "learning_rate": 1.915054759653369e-06, + "loss": 0.0023, + "step": 5739 + }, + { + "epoch": 4.12875382125517, + "grad_norm": 3.3255090637484583, + "learning_rate": 1.914727788991731e-06, + "loss": 0.039, + "step": 5740 + }, + { + "epoch": 4.129473116345981, + "grad_norm": 3.997873654300003, + "learning_rate": 1.9144007969916195e-06, + "loss": 0.0609, + "step": 5741 + }, + { + "epoch": 4.130192411436792, + "grad_norm": 1.330952232204275, + "learning_rate": 1.9140737836698602e-06, + "loss": 0.0029, + "step": 5742 + }, + { + "epoch": 4.130911706527603, + "grad_norm": 8.463264423109544, + "learning_rate": 1.913746749043278e-06, + "loss": 0.0147, + "step": 5743 + }, + { + "epoch": 4.131631001618414, + "grad_norm": 3.7504634060275497, + "learning_rate": 1.913419693128699e-06, + "loss": 0.0696, + "step": 5744 + }, + { + "epoch": 4.132350296709225, + "grad_norm": 0.7937706890853117, + "learning_rate": 1.913092615942951e-06, + "loss": 0.0135, + "step": 5745 + }, + { + "epoch": 4.1330695918000355, + "grad_norm": 4.055137934897723, + "learning_rate": 1.912765517502862e-06, + "loss": 0.1069, + "step": 5746 + }, + { + "epoch": 4.133788886890847, + "grad_norm": 3.8763779035471178, + "learning_rate": 1.912438397825264e-06, + "loss": 0.0747, + "step": 5747 + }, + { + "epoch": 4.134508181981658, + "grad_norm": 5.3533595084666645, + "learning_rate": 1.9121112569269853e-06, + "loss": 0.1169, + "step": 5748 + }, + { + "epoch": 4.135227477072469, + "grad_norm": 0.036475825017675566, + "learning_rate": 1.9117840948248595e-06, + "loss": 0.0002, + "step": 5749 + }, + { + "epoch": 4.13594677216328, + "grad_norm": 5.473958445411924, + "learning_rate": 1.911456911535719e-06, + "loss": 0.13, + "step": 5750 + }, + { + "epoch": 4.136666067254091, + "grad_norm": 2.423564956203651, + "learning_rate": 1.9111297070763982e-06, + "loss": 0.0457, + "step": 5751 + }, + { + "epoch": 4.137385362344902, + "grad_norm": 2.595427684046276, + "learning_rate": 1.9108024814637323e-06, + "loss": 0.0208, + "step": 5752 + }, + { + "epoch": 4.138104657435713, + "grad_norm": 1.7769957239712937, + "learning_rate": 1.910475234714557e-06, + "loss": 0.0216, + "step": 5753 + }, + { + "epoch": 4.138823952526524, + "grad_norm": 1.9788640637880044, + "learning_rate": 1.9101479668457114e-06, + "loss": 0.0568, + "step": 5754 + }, + { + "epoch": 4.139543247617335, + "grad_norm": 1.1212027177133155, + "learning_rate": 1.909820677874032e-06, + "loss": 0.0134, + "step": 5755 + }, + { + "epoch": 4.140262542708146, + "grad_norm": 1.9857453706365673, + "learning_rate": 1.9094933678163596e-06, + "loss": 0.0382, + "step": 5756 + }, + { + "epoch": 4.140981837798957, + "grad_norm": 3.114014620615098, + "learning_rate": 1.9091660366895347e-06, + "loss": 0.0759, + "step": 5757 + }, + { + "epoch": 4.141701132889768, + "grad_norm": 0.09972055536699499, + "learning_rate": 1.9088386845103987e-06, + "loss": 0.0005, + "step": 5758 + }, + { + "epoch": 4.142420427980579, + "grad_norm": 4.890353676999161, + "learning_rate": 1.908511311295795e-06, + "loss": 0.0197, + "step": 5759 + }, + { + "epoch": 4.14313972307139, + "grad_norm": 3.516461408750449, + "learning_rate": 1.908183917062567e-06, + "loss": 0.0858, + "step": 5760 + }, + { + "epoch": 4.1438590181622015, + "grad_norm": 3.170449586205371, + "learning_rate": 1.9078565018275597e-06, + "loss": 0.0658, + "step": 5761 + }, + { + "epoch": 4.144578313253012, + "grad_norm": 2.6889901968826773, + "learning_rate": 1.9075290656076198e-06, + "loss": 0.0236, + "step": 5762 + }, + { + "epoch": 4.145297608343823, + "grad_norm": 0.036971276535441605, + "learning_rate": 1.9072016084195941e-06, + "loss": 0.0001, + "step": 5763 + }, + { + "epoch": 4.146016903434634, + "grad_norm": 1.6408124329273948, + "learning_rate": 1.9068741302803307e-06, + "loss": 0.0128, + "step": 5764 + }, + { + "epoch": 4.146736198525445, + "grad_norm": 0.47286451801824736, + "learning_rate": 1.9065466312066794e-06, + "loss": 0.001, + "step": 5765 + }, + { + "epoch": 4.147455493616256, + "grad_norm": 2.383708330005948, + "learning_rate": 1.9062191112154895e-06, + "loss": 0.0432, + "step": 5766 + }, + { + "epoch": 4.148174788707067, + "grad_norm": 2.524267738584006, + "learning_rate": 1.9058915703236138e-06, + "loss": 0.0648, + "step": 5767 + }, + { + "epoch": 4.148894083797878, + "grad_norm": 1.6119247926309686, + "learning_rate": 1.905564008547904e-06, + "loss": 0.0157, + "step": 5768 + }, + { + "epoch": 4.149613378888689, + "grad_norm": 3.589480421996863, + "learning_rate": 1.9052364259052142e-06, + "loss": 0.1253, + "step": 5769 + }, + { + "epoch": 4.1503326739795, + "grad_norm": 3.0843119702416217, + "learning_rate": 1.9049088224123987e-06, + "loss": 0.085, + "step": 5770 + }, + { + "epoch": 4.151051969070311, + "grad_norm": 1.6600056034156323, + "learning_rate": 1.9045811980863139e-06, + "loss": 0.0105, + "step": 5771 + }, + { + "epoch": 4.151771264161122, + "grad_norm": 3.48869019972954, + "learning_rate": 1.904253552943816e-06, + "loss": 0.0404, + "step": 5772 + }, + { + "epoch": 4.152490559251933, + "grad_norm": 4.26626934618449, + "learning_rate": 1.903925887001763e-06, + "loss": 0.0281, + "step": 5773 + }, + { + "epoch": 4.1532098543427445, + "grad_norm": 2.306421585229354, + "learning_rate": 1.9035982002770142e-06, + "loss": 0.0123, + "step": 5774 + }, + { + "epoch": 4.153929149433555, + "grad_norm": 3.623595342085852, + "learning_rate": 1.9032704927864294e-06, + "loss": 0.0518, + "step": 5775 + }, + { + "epoch": 4.1546484445243665, + "grad_norm": 0.9709260979998818, + "learning_rate": 1.9029427645468696e-06, + "loss": 0.0079, + "step": 5776 + }, + { + "epoch": 4.155367739615177, + "grad_norm": 2.8125965419524683, + "learning_rate": 1.902615015575198e-06, + "loss": 0.0574, + "step": 5777 + }, + { + "epoch": 4.156087034705988, + "grad_norm": 1.4766688078141321, + "learning_rate": 1.9022872458882762e-06, + "loss": 0.005, + "step": 5778 + }, + { + "epoch": 4.156806329796799, + "grad_norm": 5.4351842438709355, + "learning_rate": 1.90195945550297e-06, + "loss": 0.209, + "step": 5779 + }, + { + "epoch": 4.15752562488761, + "grad_norm": 3.1445047329719302, + "learning_rate": 1.9016316444361442e-06, + "loss": 0.0192, + "step": 5780 + }, + { + "epoch": 4.158244919978421, + "grad_norm": 2.4781853647458982, + "learning_rate": 1.901303812704665e-06, + "loss": 0.0496, + "step": 5781 + }, + { + "epoch": 4.158964215069232, + "grad_norm": 2.116496252777341, + "learning_rate": 1.9009759603254001e-06, + "loss": 0.0451, + "step": 5782 + }, + { + "epoch": 4.159683510160043, + "grad_norm": 6.270144550982344, + "learning_rate": 1.9006480873152183e-06, + "loss": 0.0304, + "step": 5783 + }, + { + "epoch": 4.160402805250854, + "grad_norm": 3.3902958020506957, + "learning_rate": 1.900320193690989e-06, + "loss": 0.0961, + "step": 5784 + }, + { + "epoch": 4.161122100341665, + "grad_norm": 0.8102162561300114, + "learning_rate": 1.899992279469583e-06, + "loss": 0.0016, + "step": 5785 + }, + { + "epoch": 4.161841395432476, + "grad_norm": 3.1928791797563996, + "learning_rate": 1.8996643446678726e-06, + "loss": 0.0933, + "step": 5786 + }, + { + "epoch": 4.1625606905232875, + "grad_norm": 4.25379461117837, + "learning_rate": 1.8993363893027295e-06, + "loss": 0.1626, + "step": 5787 + }, + { + "epoch": 4.163279985614098, + "grad_norm": 0.930861104758418, + "learning_rate": 1.8990084133910282e-06, + "loss": 0.014, + "step": 5788 + }, + { + "epoch": 4.1639992807049095, + "grad_norm": 1.8826449397337839, + "learning_rate": 1.8986804169496434e-06, + "loss": 0.029, + "step": 5789 + }, + { + "epoch": 4.16471857579572, + "grad_norm": 1.2891820621279988, + "learning_rate": 1.8983523999954519e-06, + "loss": 0.0083, + "step": 5790 + }, + { + "epoch": 4.165437870886532, + "grad_norm": 3.974385775948942, + "learning_rate": 1.8980243625453297e-06, + "loss": 0.1245, + "step": 5791 + }, + { + "epoch": 4.166157165977342, + "grad_norm": 0.4912990922845634, + "learning_rate": 1.8976963046161553e-06, + "loss": 0.0004, + "step": 5792 + }, + { + "epoch": 4.166876461068153, + "grad_norm": 3.690645002955706, + "learning_rate": 1.897368226224808e-06, + "loss": 0.0642, + "step": 5793 + }, + { + "epoch": 4.167595756158964, + "grad_norm": 2.769770308178185, + "learning_rate": 1.8970401273881683e-06, + "loss": 0.0594, + "step": 5794 + }, + { + "epoch": 4.168315051249775, + "grad_norm": 3.2113057093588537, + "learning_rate": 1.8967120081231166e-06, + "loss": 0.0666, + "step": 5795 + }, + { + "epoch": 4.169034346340586, + "grad_norm": 4.056167657611307, + "learning_rate": 1.8963838684465358e-06, + "loss": 0.0828, + "step": 5796 + }, + { + "epoch": 4.169753641431397, + "grad_norm": 6.944010524991526, + "learning_rate": 1.8960557083753084e-06, + "loss": 0.0602, + "step": 5797 + }, + { + "epoch": 4.170472936522208, + "grad_norm": 1.1320164462256035, + "learning_rate": 1.8957275279263204e-06, + "loss": 0.0051, + "step": 5798 + }, + { + "epoch": 4.171192231613019, + "grad_norm": 0.411178589554054, + "learning_rate": 1.8953993271164565e-06, + "loss": 0.001, + "step": 5799 + }, + { + "epoch": 4.1719115267038305, + "grad_norm": 0.14130655923721003, + "learning_rate": 1.8950711059626032e-06, + "loss": 0.0004, + "step": 5800 + }, + { + "epoch": 4.172630821794641, + "grad_norm": 0.8734141844504232, + "learning_rate": 1.8947428644816474e-06, + "loss": 0.002, + "step": 5801 + }, + { + "epoch": 4.1733501168854525, + "grad_norm": 2.7702609095547284, + "learning_rate": 1.8944146026904787e-06, + "loss": 0.0466, + "step": 5802 + }, + { + "epoch": 4.174069411976263, + "grad_norm": 2.95616042881728, + "learning_rate": 1.8940863206059857e-06, + "loss": 0.0573, + "step": 5803 + }, + { + "epoch": 4.174788707067075, + "grad_norm": 1.75491999514677, + "learning_rate": 1.8937580182450606e-06, + "loss": 0.0248, + "step": 5804 + }, + { + "epoch": 4.175508002157885, + "grad_norm": 2.58997774510672, + "learning_rate": 1.8934296956245934e-06, + "loss": 0.0102, + "step": 5805 + }, + { + "epoch": 4.176227297248697, + "grad_norm": 2.6396945577987236, + "learning_rate": 1.8931013527614781e-06, + "loss": 0.0606, + "step": 5806 + }, + { + "epoch": 4.176946592339507, + "grad_norm": 3.9452208816044356, + "learning_rate": 1.8927729896726084e-06, + "loss": 0.1285, + "step": 5807 + }, + { + "epoch": 4.177665887430318, + "grad_norm": 3.49414760330343, + "learning_rate": 1.8924446063748787e-06, + "loss": 0.1091, + "step": 5808 + }, + { + "epoch": 4.178385182521129, + "grad_norm": 0.014010169834209732, + "learning_rate": 1.892116202885185e-06, + "loss": 0.0001, + "step": 5809 + }, + { + "epoch": 4.17910447761194, + "grad_norm": 3.49949304103724, + "learning_rate": 1.8917877792204239e-06, + "loss": 0.1036, + "step": 5810 + }, + { + "epoch": 4.179823772702751, + "grad_norm": 4.872635589140993, + "learning_rate": 1.8914593353974943e-06, + "loss": 0.0205, + "step": 5811 + }, + { + "epoch": 4.180543067793562, + "grad_norm": 2.049418031178065, + "learning_rate": 1.8911308714332949e-06, + "loss": 0.0305, + "step": 5812 + }, + { + "epoch": 4.1812623628843735, + "grad_norm": 1.029690270464678, + "learning_rate": 1.8908023873447249e-06, + "loss": 0.0102, + "step": 5813 + }, + { + "epoch": 4.181981657975184, + "grad_norm": 4.158214959083454, + "learning_rate": 1.8904738831486866e-06, + "loss": 0.1606, + "step": 5814 + }, + { + "epoch": 4.1827009530659955, + "grad_norm": 3.9784793213067045, + "learning_rate": 1.8901453588620815e-06, + "loss": 0.0841, + "step": 5815 + }, + { + "epoch": 4.183420248156806, + "grad_norm": 3.389362132236326, + "learning_rate": 1.8898168145018127e-06, + "loss": 0.1125, + "step": 5816 + }, + { + "epoch": 4.184139543247618, + "grad_norm": 2.6974953943164968, + "learning_rate": 1.8894882500847846e-06, + "loss": 0.0413, + "step": 5817 + }, + { + "epoch": 4.184858838338428, + "grad_norm": 1.7920817806249878, + "learning_rate": 1.8891596656279023e-06, + "loss": 0.0069, + "step": 5818 + }, + { + "epoch": 4.18557813342924, + "grad_norm": 4.692882547588171, + "learning_rate": 1.8888310611480719e-06, + "loss": 0.1354, + "step": 5819 + }, + { + "epoch": 4.18629742852005, + "grad_norm": 0.08825633687724205, + "learning_rate": 1.888502436662201e-06, + "loss": 0.0003, + "step": 5820 + }, + { + "epoch": 4.187016723610862, + "grad_norm": 1.2705597798690191, + "learning_rate": 1.8881737921871972e-06, + "loss": 0.0345, + "step": 5821 + }, + { + "epoch": 4.187736018701672, + "grad_norm": 1.6541408399956077, + "learning_rate": 1.8878451277399708e-06, + "loss": 0.0283, + "step": 5822 + }, + { + "epoch": 4.188455313792483, + "grad_norm": 3.249451201991626, + "learning_rate": 1.8875164433374323e-06, + "loss": 0.0496, + "step": 5823 + }, + { + "epoch": 4.189174608883294, + "grad_norm": 4.82677512996356, + "learning_rate": 1.8871877389964916e-06, + "loss": 0.0683, + "step": 5824 + }, + { + "epoch": 4.189893903974105, + "grad_norm": 1.2299801448342416, + "learning_rate": 1.8868590147340625e-06, + "loss": 0.0113, + "step": 5825 + }, + { + "epoch": 4.1906131990649165, + "grad_norm": 2.2226091511290678, + "learning_rate": 1.8865302705670578e-06, + "loss": 0.008, + "step": 5826 + }, + { + "epoch": 4.191332494155727, + "grad_norm": 0.5467715948718692, + "learning_rate": 1.8862015065123923e-06, + "loss": 0.0007, + "step": 5827 + }, + { + "epoch": 4.1920517892465385, + "grad_norm": 2.82455693388467, + "learning_rate": 1.885872722586981e-06, + "loss": 0.0692, + "step": 5828 + }, + { + "epoch": 4.192771084337349, + "grad_norm": 1.2803323217742164, + "learning_rate": 1.8855439188077415e-06, + "loss": 0.0036, + "step": 5829 + }, + { + "epoch": 4.193490379428161, + "grad_norm": 3.358229586143644, + "learning_rate": 1.88521509519159e-06, + "loss": 0.005, + "step": 5830 + }, + { + "epoch": 4.194209674518971, + "grad_norm": 2.648898693548551, + "learning_rate": 1.8848862517554455e-06, + "loss": 0.0533, + "step": 5831 + }, + { + "epoch": 4.194928969609783, + "grad_norm": 3.3655808224049353, + "learning_rate": 1.8845573885162282e-06, + "loss": 0.082, + "step": 5832 + }, + { + "epoch": 4.195648264700593, + "grad_norm": 3.4945885275606945, + "learning_rate": 1.8842285054908578e-06, + "loss": 0.0825, + "step": 5833 + }, + { + "epoch": 4.196367559791405, + "grad_norm": 8.504554917793012, + "learning_rate": 1.8838996026962561e-06, + "loss": 0.1192, + "step": 5834 + }, + { + "epoch": 4.197086854882215, + "grad_norm": 0.004031159194195304, + "learning_rate": 1.8835706801493463e-06, + "loss": 0.0, + "step": 5835 + }, + { + "epoch": 4.197806149973027, + "grad_norm": 1.045044851110039, + "learning_rate": 1.8832417378670516e-06, + "loss": 0.0027, + "step": 5836 + }, + { + "epoch": 4.198525445063837, + "grad_norm": 2.0584204775627675, + "learning_rate": 1.882912775866297e-06, + "loss": 0.0494, + "step": 5837 + }, + { + "epoch": 4.199244740154649, + "grad_norm": 3.537161862012772, + "learning_rate": 1.8825837941640077e-06, + "loss": 0.0707, + "step": 5838 + }, + { + "epoch": 4.1999640352454595, + "grad_norm": 3.5712675659349067, + "learning_rate": 1.8822547927771105e-06, + "loss": 0.1291, + "step": 5839 + }, + { + "epoch": 4.20068333033627, + "grad_norm": 1.862633146828256, + "learning_rate": 1.881925771722533e-06, + "loss": 0.0093, + "step": 5840 + }, + { + "epoch": 4.2014026254270815, + "grad_norm": 2.7771843135641796, + "learning_rate": 1.8815967310172045e-06, + "loss": 0.0424, + "step": 5841 + }, + { + "epoch": 4.202121920517892, + "grad_norm": 3.25360022028823, + "learning_rate": 1.8812676706780538e-06, + "loss": 0.0725, + "step": 5842 + }, + { + "epoch": 4.202841215608704, + "grad_norm": 0.04127061186070438, + "learning_rate": 1.8809385907220127e-06, + "loss": 0.0001, + "step": 5843 + }, + { + "epoch": 4.203560510699514, + "grad_norm": 0.8930751241959299, + "learning_rate": 1.8806094911660122e-06, + "loss": 0.0016, + "step": 5844 + }, + { + "epoch": 4.204279805790326, + "grad_norm": 3.2490621063289447, + "learning_rate": 1.8802803720269852e-06, + "loss": 0.0364, + "step": 5845 + }, + { + "epoch": 4.204999100881136, + "grad_norm": 4.766043594658544, + "learning_rate": 1.8799512333218658e-06, + "loss": 0.1198, + "step": 5846 + }, + { + "epoch": 4.205718395971948, + "grad_norm": 2.7814063832779827, + "learning_rate": 1.8796220750675878e-06, + "loss": 0.102, + "step": 5847 + }, + { + "epoch": 4.206437691062758, + "grad_norm": 1.8370619488034219, + "learning_rate": 1.8792928972810878e-06, + "loss": 0.0361, + "step": 5848 + }, + { + "epoch": 4.20715698615357, + "grad_norm": 1.8601667561197879, + "learning_rate": 1.878963699979303e-06, + "loss": 0.0293, + "step": 5849 + }, + { + "epoch": 4.20787628124438, + "grad_norm": 0.06678646711756969, + "learning_rate": 1.8786344831791699e-06, + "loss": 0.0003, + "step": 5850 + }, + { + "epoch": 4.208595576335192, + "grad_norm": 1.65262893562749, + "learning_rate": 1.878305246897628e-06, + "loss": 0.0098, + "step": 5851 + }, + { + "epoch": 4.2093148714260025, + "grad_norm": 3.4603183835149425, + "learning_rate": 1.8779759911516177e-06, + "loss": 0.0835, + "step": 5852 + }, + { + "epoch": 4.210034166516814, + "grad_norm": 1.9235748854834687, + "learning_rate": 1.8776467159580787e-06, + "loss": 0.0373, + "step": 5853 + }, + { + "epoch": 4.2107534616076245, + "grad_norm": 3.733667925593394, + "learning_rate": 1.8773174213339533e-06, + "loss": 0.117, + "step": 5854 + }, + { + "epoch": 4.211472756698435, + "grad_norm": 2.309125943642675, + "learning_rate": 1.876988107296184e-06, + "loss": 0.049, + "step": 5855 + }, + { + "epoch": 4.212192051789247, + "grad_norm": 0.015154384890002204, + "learning_rate": 1.876658773861715e-06, + "loss": 0.0001, + "step": 5856 + }, + { + "epoch": 4.212911346880057, + "grad_norm": 1.7312710737042403, + "learning_rate": 1.8763294210474913e-06, + "loss": 0.0317, + "step": 5857 + }, + { + "epoch": 4.213630641970869, + "grad_norm": 4.468124772147032, + "learning_rate": 1.8760000488704573e-06, + "loss": 0.1248, + "step": 5858 + }, + { + "epoch": 4.214349937061679, + "grad_norm": 3.7833563846027105, + "learning_rate": 1.8756706573475618e-06, + "loss": 0.0567, + "step": 5859 + }, + { + "epoch": 4.215069232152491, + "grad_norm": 3.3560000676703985, + "learning_rate": 1.8753412464957506e-06, + "loss": 0.1047, + "step": 5860 + }, + { + "epoch": 4.215788527243301, + "grad_norm": 0.3524999107866227, + "learning_rate": 1.875011816331974e-06, + "loss": 0.0019, + "step": 5861 + }, + { + "epoch": 4.216507822334113, + "grad_norm": 3.197371783227172, + "learning_rate": 1.8746823668731807e-06, + "loss": 0.0516, + "step": 5862 + }, + { + "epoch": 4.217227117424923, + "grad_norm": 1.4717284599707876, + "learning_rate": 1.8743528981363221e-06, + "loss": 0.0083, + "step": 5863 + }, + { + "epoch": 4.217946412515735, + "grad_norm": 2.982139880903086, + "learning_rate": 1.8740234101383497e-06, + "loss": 0.0844, + "step": 5864 + }, + { + "epoch": 4.2186657076065455, + "grad_norm": 2.447398251985535, + "learning_rate": 1.8736939028962164e-06, + "loss": 0.0464, + "step": 5865 + }, + { + "epoch": 4.219385002697357, + "grad_norm": 2.2219479048759223, + "learning_rate": 1.8733643764268759e-06, + "loss": 0.0077, + "step": 5866 + }, + { + "epoch": 4.2201042977881675, + "grad_norm": 6.374355359391193, + "learning_rate": 1.8730348307472826e-06, + "loss": 0.0851, + "step": 5867 + }, + { + "epoch": 4.220823592878979, + "grad_norm": 6.396304342548548, + "learning_rate": 1.8727052658743921e-06, + "loss": 0.1469, + "step": 5868 + }, + { + "epoch": 4.22154288796979, + "grad_norm": 0.3257897792405374, + "learning_rate": 1.8723756818251618e-06, + "loss": 0.0003, + "step": 5869 + }, + { + "epoch": 4.2222621830606, + "grad_norm": 5.368853724884114, + "learning_rate": 1.872046078616549e-06, + "loss": 0.1368, + "step": 5870 + }, + { + "epoch": 4.222981478151412, + "grad_norm": 1.5154694475466703, + "learning_rate": 1.871716456265512e-06, + "loss": 0.0318, + "step": 5871 + }, + { + "epoch": 4.223700773242222, + "grad_norm": 2.7963764665022026, + "learning_rate": 1.871386814789011e-06, + "loss": 0.0527, + "step": 5872 + }, + { + "epoch": 4.224420068333034, + "grad_norm": 0.44228936397304863, + "learning_rate": 1.8710571542040066e-06, + "loss": 0.0011, + "step": 5873 + }, + { + "epoch": 4.225139363423844, + "grad_norm": 0.36316969575848845, + "learning_rate": 1.8707274745274602e-06, + "loss": 0.0038, + "step": 5874 + }, + { + "epoch": 4.225858658514656, + "grad_norm": 4.080020288225842, + "learning_rate": 1.8703977757763347e-06, + "loss": 0.0812, + "step": 5875 + }, + { + "epoch": 4.226577953605466, + "grad_norm": 0.7876272724385197, + "learning_rate": 1.8700680579675928e-06, + "loss": 0.0011, + "step": 5876 + }, + { + "epoch": 4.227297248696278, + "grad_norm": 3.0979831758105845, + "learning_rate": 1.8697383211182e-06, + "loss": 0.0434, + "step": 5877 + }, + { + "epoch": 4.2280165437870885, + "grad_norm": 1.8495942584723701, + "learning_rate": 1.8694085652451215e-06, + "loss": 0.037, + "step": 5878 + }, + { + "epoch": 4.2287358388779, + "grad_norm": 7.332297330499419, + "learning_rate": 1.8690787903653236e-06, + "loss": 0.0192, + "step": 5879 + }, + { + "epoch": 4.2294551339687105, + "grad_norm": 4.596710372970895, + "learning_rate": 1.8687489964957744e-06, + "loss": 0.0906, + "step": 5880 + }, + { + "epoch": 4.230174429059522, + "grad_norm": 3.170970191678553, + "learning_rate": 1.8684191836534419e-06, + "loss": 0.0593, + "step": 5881 + }, + { + "epoch": 4.230893724150333, + "grad_norm": 2.6106000644772345, + "learning_rate": 1.868089351855296e-06, + "loss": 0.0489, + "step": 5882 + }, + { + "epoch": 4.231613019241144, + "grad_norm": 2.473199122836248, + "learning_rate": 1.8677595011183062e-06, + "loss": 0.021, + "step": 5883 + }, + { + "epoch": 4.232332314331955, + "grad_norm": 0.09789603831506533, + "learning_rate": 1.8674296314594448e-06, + "loss": 0.0006, + "step": 5884 + }, + { + "epoch": 4.233051609422765, + "grad_norm": 2.066506966795153, + "learning_rate": 1.8670997428956838e-06, + "loss": 0.0589, + "step": 5885 + }, + { + "epoch": 4.233770904513577, + "grad_norm": 5.059624413130225, + "learning_rate": 1.8667698354439963e-06, + "loss": 0.1442, + "step": 5886 + }, + { + "epoch": 4.234490199604387, + "grad_norm": 2.56604183774849, + "learning_rate": 1.8664399091213572e-06, + "loss": 0.0446, + "step": 5887 + }, + { + "epoch": 4.235209494695199, + "grad_norm": 3.5055602939738324, + "learning_rate": 1.866109963944742e-06, + "loss": 0.0734, + "step": 5888 + }, + { + "epoch": 4.235928789786009, + "grad_norm": 1.7880852712683308, + "learning_rate": 1.8657799999311257e-06, + "loss": 0.0208, + "step": 5889 + }, + { + "epoch": 4.236648084876821, + "grad_norm": 6.088501720061008, + "learning_rate": 1.8654500170974865e-06, + "loss": 0.1036, + "step": 5890 + }, + { + "epoch": 4.2373673799676315, + "grad_norm": 0.11984516287083895, + "learning_rate": 1.8651200154608027e-06, + "loss": 0.0003, + "step": 5891 + }, + { + "epoch": 4.238086675058443, + "grad_norm": 4.62055890538338, + "learning_rate": 1.8647899950380527e-06, + "loss": 0.174, + "step": 5892 + }, + { + "epoch": 4.2388059701492535, + "grad_norm": 4.0930345376421, + "learning_rate": 1.8644599558462173e-06, + "loss": 0.1247, + "step": 5893 + }, + { + "epoch": 4.239525265240065, + "grad_norm": 1.1658417524406146, + "learning_rate": 1.864129897902277e-06, + "loss": 0.0025, + "step": 5894 + }, + { + "epoch": 4.240244560330876, + "grad_norm": 1.5589410821745784, + "learning_rate": 1.8637998212232151e-06, + "loss": 0.02, + "step": 5895 + }, + { + "epoch": 4.240963855421687, + "grad_norm": 1.189773461128153, + "learning_rate": 1.8634697258260136e-06, + "loss": 0.0034, + "step": 5896 + }, + { + "epoch": 4.241683150512498, + "grad_norm": 1.7620092528277491, + "learning_rate": 1.863139611727656e-06, + "loss": 0.0215, + "step": 5897 + }, + { + "epoch": 4.242402445603309, + "grad_norm": 3.966145326277798, + "learning_rate": 1.8628094789451285e-06, + "loss": 0.1079, + "step": 5898 + }, + { + "epoch": 4.24312174069412, + "grad_norm": 8.574081463336737, + "learning_rate": 1.8624793274954163e-06, + "loss": 0.0242, + "step": 5899 + }, + { + "epoch": 4.24384103578493, + "grad_norm": 3.340915560071435, + "learning_rate": 1.862149157395506e-06, + "loss": 0.0351, + "step": 5900 + }, + { + "epoch": 4.244560330875742, + "grad_norm": 0.8495320609386708, + "learning_rate": 1.8618189686623862e-06, + "loss": 0.0012, + "step": 5901 + }, + { + "epoch": 4.245279625966552, + "grad_norm": 2.820093087267449, + "learning_rate": 1.861488761313045e-06, + "loss": 0.0255, + "step": 5902 + }, + { + "epoch": 4.245998921057364, + "grad_norm": 4.382744446646367, + "learning_rate": 1.861158535364473e-06, + "loss": 0.1202, + "step": 5903 + }, + { + "epoch": 4.2467182161481745, + "grad_norm": 1.6694581527214025, + "learning_rate": 1.86082829083366e-06, + "loss": 0.0181, + "step": 5904 + }, + { + "epoch": 4.247437511238986, + "grad_norm": 2.4666991277095844, + "learning_rate": 1.860498027737598e-06, + "loss": 0.0042, + "step": 5905 + }, + { + "epoch": 4.2481568063297965, + "grad_norm": 3.0446000012271655, + "learning_rate": 1.8601677460932798e-06, + "loss": 0.0498, + "step": 5906 + }, + { + "epoch": 4.248876101420608, + "grad_norm": 2.4431406259454804, + "learning_rate": 1.8598374459176983e-06, + "loss": 0.0124, + "step": 5907 + }, + { + "epoch": 4.249595396511419, + "grad_norm": 3.800070646470269, + "learning_rate": 1.859507127227849e-06, + "loss": 0.0909, + "step": 5908 + }, + { + "epoch": 4.25031469160223, + "grad_norm": 2.421468155590113, + "learning_rate": 1.8591767900407268e-06, + "loss": 0.0441, + "step": 5909 + }, + { + "epoch": 4.251033986693041, + "grad_norm": 0.6632424365974716, + "learning_rate": 1.8588464343733285e-06, + "loss": 0.0015, + "step": 5910 + }, + { + "epoch": 4.251753281783852, + "grad_norm": 3.789378577544711, + "learning_rate": 1.8585160602426513e-06, + "loss": 0.098, + "step": 5911 + }, + { + "epoch": 4.252472576874663, + "grad_norm": 3.0708831123741183, + "learning_rate": 1.8581856676656932e-06, + "loss": 0.054, + "step": 5912 + }, + { + "epoch": 4.253191871965474, + "grad_norm": 5.348343120393847, + "learning_rate": 1.8578552566594534e-06, + "loss": 0.1034, + "step": 5913 + }, + { + "epoch": 4.253911167056285, + "grad_norm": 2.117913939365991, + "learning_rate": 1.8575248272409328e-06, + "loss": 0.0369, + "step": 5914 + }, + { + "epoch": 4.254630462147096, + "grad_norm": 3.0106151536993067, + "learning_rate": 1.8571943794271318e-06, + "loss": 0.0327, + "step": 5915 + }, + { + "epoch": 4.255349757237907, + "grad_norm": 2.6650093696981205, + "learning_rate": 1.8568639132350534e-06, + "loss": 0.0591, + "step": 5916 + }, + { + "epoch": 4.2560690523287175, + "grad_norm": 0.029628981727282486, + "learning_rate": 1.8565334286817002e-06, + "loss": 0.0002, + "step": 5917 + }, + { + "epoch": 4.256788347419529, + "grad_norm": 2.6689963136708847, + "learning_rate": 1.8562029257840762e-06, + "loss": 0.0583, + "step": 5918 + }, + { + "epoch": 4.2575076425103395, + "grad_norm": 2.109592289804135, + "learning_rate": 1.8558724045591863e-06, + "loss": 0.0492, + "step": 5919 + }, + { + "epoch": 4.258226937601151, + "grad_norm": 5.793288385997074, + "learning_rate": 1.8555418650240362e-06, + "loss": 0.1825, + "step": 5920 + }, + { + "epoch": 4.258946232691962, + "grad_norm": 2.54439290499571, + "learning_rate": 1.8552113071956334e-06, + "loss": 0.0407, + "step": 5921 + }, + { + "epoch": 4.259665527782773, + "grad_norm": 2.0790628092205745, + "learning_rate": 1.8548807310909852e-06, + "loss": 0.0251, + "step": 5922 + }, + { + "epoch": 4.260384822873584, + "grad_norm": 1.9822665018532393, + "learning_rate": 1.8545501367271002e-06, + "loss": 0.0193, + "step": 5923 + }, + { + "epoch": 4.261104117964395, + "grad_norm": 3.7620818226604738, + "learning_rate": 1.854219524120989e-06, + "loss": 0.1059, + "step": 5924 + }, + { + "epoch": 4.261823413055206, + "grad_norm": 1.724665407219222, + "learning_rate": 1.853888893289661e-06, + "loss": 0.0239, + "step": 5925 + }, + { + "epoch": 4.262542708146017, + "grad_norm": 3.4892246719920554, + "learning_rate": 1.8535582442501276e-06, + "loss": 0.0493, + "step": 5926 + }, + { + "epoch": 4.263262003236828, + "grad_norm": 0.004883306361694121, + "learning_rate": 1.8532275770194024e-06, + "loss": 0.0, + "step": 5927 + }, + { + "epoch": 4.263981298327639, + "grad_norm": 1.5143205045891344, + "learning_rate": 1.8528968916144982e-06, + "loss": 0.0243, + "step": 5928 + }, + { + "epoch": 4.26470059341845, + "grad_norm": 1.4621053779602087, + "learning_rate": 1.8525661880524297e-06, + "loss": 0.0496, + "step": 5929 + }, + { + "epoch": 4.2654198885092605, + "grad_norm": 3.826356025732097, + "learning_rate": 1.8522354663502118e-06, + "loss": 0.0695, + "step": 5930 + }, + { + "epoch": 4.266139183600072, + "grad_norm": 0.4021148193844498, + "learning_rate": 1.8519047265248608e-06, + "loss": 0.0013, + "step": 5931 + }, + { + "epoch": 4.2668584786908825, + "grad_norm": 2.4810729237426283, + "learning_rate": 1.851573968593394e-06, + "loss": 0.0852, + "step": 5932 + }, + { + "epoch": 4.267577773781694, + "grad_norm": 3.0129186319855523, + "learning_rate": 1.851243192572829e-06, + "loss": 0.0362, + "step": 5933 + }, + { + "epoch": 4.268297068872505, + "grad_norm": 4.588402126750859, + "learning_rate": 1.8509123984801853e-06, + "loss": 0.1104, + "step": 5934 + }, + { + "epoch": 4.269016363963316, + "grad_norm": 0.02730327864304445, + "learning_rate": 1.8505815863324827e-06, + "loss": 0.0001, + "step": 5935 + }, + { + "epoch": 4.269735659054127, + "grad_norm": 3.1541518237774224, + "learning_rate": 1.850250756146742e-06, + "loss": 0.0141, + "step": 5936 + }, + { + "epoch": 4.270454954144938, + "grad_norm": 2.5854167823494385, + "learning_rate": 1.8499199079399852e-06, + "loss": 0.0144, + "step": 5937 + }, + { + "epoch": 4.271174249235749, + "grad_norm": 2.338314473524037, + "learning_rate": 1.8495890417292348e-06, + "loss": 0.0561, + "step": 5938 + }, + { + "epoch": 4.27189354432656, + "grad_norm": 2.107504559815359, + "learning_rate": 1.8492581575315148e-06, + "loss": 0.06, + "step": 5939 + }, + { + "epoch": 4.272612839417371, + "grad_norm": 2.1596265860836255, + "learning_rate": 1.848927255363849e-06, + "loss": 0.0166, + "step": 5940 + }, + { + "epoch": 4.273332134508182, + "grad_norm": 6.8946750320470915, + "learning_rate": 1.8485963352432635e-06, + "loss": 0.1014, + "step": 5941 + }, + { + "epoch": 4.274051429598993, + "grad_norm": 2.427194884530317, + "learning_rate": 1.8482653971867848e-06, + "loss": 0.0195, + "step": 5942 + }, + { + "epoch": 4.274770724689804, + "grad_norm": 4.608547930797963, + "learning_rate": 1.8479344412114402e-06, + "loss": 0.0513, + "step": 5943 + }, + { + "epoch": 4.275490019780615, + "grad_norm": 1.5944498481186478, + "learning_rate": 1.847603467334257e-06, + "loss": 0.0172, + "step": 5944 + }, + { + "epoch": 4.276209314871426, + "grad_norm": 2.933527337415574, + "learning_rate": 1.8472724755722661e-06, + "loss": 0.0637, + "step": 5945 + }, + { + "epoch": 4.276928609962237, + "grad_norm": 2.1065602370036847, + "learning_rate": 1.8469414659424965e-06, + "loss": 0.0119, + "step": 5946 + }, + { + "epoch": 4.277647905053048, + "grad_norm": 2.5489435125488096, + "learning_rate": 1.8466104384619794e-06, + "loss": 0.032, + "step": 5947 + }, + { + "epoch": 4.278367200143859, + "grad_norm": 4.722277222372356, + "learning_rate": 1.846279393147747e-06, + "loss": 0.0962, + "step": 5948 + }, + { + "epoch": 4.27908649523467, + "grad_norm": 3.562452103772542, + "learning_rate": 1.8459483300168311e-06, + "loss": 0.1, + "step": 5949 + }, + { + "epoch": 4.279805790325481, + "grad_norm": 2.0560653302183054, + "learning_rate": 1.8456172490862673e-06, + "loss": 0.0386, + "step": 5950 + }, + { + "epoch": 4.280525085416292, + "grad_norm": 0.09427731117062917, + "learning_rate": 1.845286150373089e-06, + "loss": 0.0002, + "step": 5951 + }, + { + "epoch": 4.281244380507103, + "grad_norm": 3.697399191165508, + "learning_rate": 1.8449550338943316e-06, + "loss": 0.0795, + "step": 5952 + }, + { + "epoch": 4.281963675597914, + "grad_norm": 1.5651293665812187, + "learning_rate": 1.8446238996670328e-06, + "loss": 0.0134, + "step": 5953 + }, + { + "epoch": 4.282682970688725, + "grad_norm": 1.4265612084098664, + "learning_rate": 1.8442927477082296e-06, + "loss": 0.0117, + "step": 5954 + }, + { + "epoch": 4.283402265779536, + "grad_norm": 3.3888461028952954, + "learning_rate": 1.8439615780349596e-06, + "loss": 0.0655, + "step": 5955 + }, + { + "epoch": 4.284121560870347, + "grad_norm": 4.246477229655259, + "learning_rate": 1.843630390664263e-06, + "loss": 0.1353, + "step": 5956 + }, + { + "epoch": 4.284840855961158, + "grad_norm": 3.7523503196051653, + "learning_rate": 1.843299185613179e-06, + "loss": 0.1037, + "step": 5957 + }, + { + "epoch": 4.285560151051969, + "grad_norm": 3.890918308956497, + "learning_rate": 1.8429679628987496e-06, + "loss": 0.0395, + "step": 5958 + }, + { + "epoch": 4.28627944614278, + "grad_norm": 1.207581898337127, + "learning_rate": 1.8426367225380166e-06, + "loss": 0.017, + "step": 5959 + }, + { + "epoch": 4.2869987412335915, + "grad_norm": 4.550634403125409, + "learning_rate": 1.8423054645480227e-06, + "loss": 0.1245, + "step": 5960 + }, + { + "epoch": 4.287718036324402, + "grad_norm": 4.371258656901886, + "learning_rate": 1.8419741889458117e-06, + "loss": 0.0447, + "step": 5961 + }, + { + "epoch": 4.2884373314152135, + "grad_norm": 2.3896008586709026, + "learning_rate": 1.841642895748428e-06, + "loss": 0.0578, + "step": 5962 + }, + { + "epoch": 4.289156626506024, + "grad_norm": 0.1990231458943505, + "learning_rate": 1.841311584972918e-06, + "loss": 0.0011, + "step": 5963 + }, + { + "epoch": 4.289875921596835, + "grad_norm": 2.204889686563333, + "learning_rate": 1.8409802566363278e-06, + "loss": 0.0211, + "step": 5964 + }, + { + "epoch": 4.290595216687646, + "grad_norm": 4.912818129663357, + "learning_rate": 1.8406489107557043e-06, + "loss": 0.1469, + "step": 5965 + }, + { + "epoch": 4.291314511778457, + "grad_norm": 5.107950483088817, + "learning_rate": 1.840317547348097e-06, + "loss": 0.0543, + "step": 5966 + }, + { + "epoch": 4.292033806869268, + "grad_norm": 2.960310944488645, + "learning_rate": 1.8399861664305542e-06, + "loss": 0.0598, + "step": 5967 + }, + { + "epoch": 4.292753101960079, + "grad_norm": 1.4352693222354085, + "learning_rate": 1.8396547680201267e-06, + "loss": 0.0038, + "step": 5968 + }, + { + "epoch": 4.29347239705089, + "grad_norm": 3.041762068515795, + "learning_rate": 1.839323352133865e-06, + "loss": 0.0722, + "step": 5969 + }, + { + "epoch": 4.294191692141701, + "grad_norm": 4.717683489434875, + "learning_rate": 1.8389919187888204e-06, + "loss": 0.1514, + "step": 5970 + }, + { + "epoch": 4.294910987232512, + "grad_norm": 0.04838206000261159, + "learning_rate": 1.8386604680020473e-06, + "loss": 0.0002, + "step": 5971 + }, + { + "epoch": 4.295630282323323, + "grad_norm": 4.0641081829903545, + "learning_rate": 1.8383289997905984e-06, + "loss": 0.0933, + "step": 5972 + }, + { + "epoch": 4.2963495774141345, + "grad_norm": 3.3636879311710817, + "learning_rate": 1.8379975141715283e-06, + "loss": 0.0066, + "step": 5973 + }, + { + "epoch": 4.297068872504945, + "grad_norm": 5.822034705506147, + "learning_rate": 1.8376660111618932e-06, + "loss": 0.0932, + "step": 5974 + }, + { + "epoch": 4.2977881675957565, + "grad_norm": 2.949823132423696, + "learning_rate": 1.8373344907787489e-06, + "loss": 0.061, + "step": 5975 + }, + { + "epoch": 4.298507462686567, + "grad_norm": 0.4361332283037628, + "learning_rate": 1.8370029530391528e-06, + "loss": 0.0006, + "step": 5976 + }, + { + "epoch": 4.299226757777378, + "grad_norm": 3.305318914942062, + "learning_rate": 1.8366713979601633e-06, + "loss": 0.099, + "step": 5977 + }, + { + "epoch": 4.299946052868189, + "grad_norm": 0.9372212986062975, + "learning_rate": 1.8363398255588391e-06, + "loss": 0.0117, + "step": 5978 + }, + { + "epoch": 4.300665347959, + "grad_norm": 4.8315180221475815, + "learning_rate": 1.8360082358522405e-06, + "loss": 0.0774, + "step": 5979 + }, + { + "epoch": 4.301384643049811, + "grad_norm": 3.047337205004141, + "learning_rate": 1.8356766288574288e-06, + "loss": 0.072, + "step": 5980 + }, + { + "epoch": 4.302103938140622, + "grad_norm": 2.8256533212936614, + "learning_rate": 1.8353450045914645e-06, + "loss": 0.0542, + "step": 5981 + }, + { + "epoch": 4.302823233231433, + "grad_norm": 0.5817324710944679, + "learning_rate": 1.835013363071412e-06, + "loss": 0.0045, + "step": 5982 + }, + { + "epoch": 4.303542528322244, + "grad_norm": 3.106765548293598, + "learning_rate": 1.834681704314333e-06, + "loss": 0.0595, + "step": 5983 + }, + { + "epoch": 4.304261823413055, + "grad_norm": 0.01694649083869301, + "learning_rate": 1.8343500283372933e-06, + "loss": 0.0001, + "step": 5984 + }, + { + "epoch": 4.304981118503866, + "grad_norm": 1.6367311924199759, + "learning_rate": 1.8340183351573576e-06, + "loss": 0.0226, + "step": 5985 + }, + { + "epoch": 4.3057004135946775, + "grad_norm": 2.03276562057214, + "learning_rate": 1.8336866247915922e-06, + "loss": 0.0394, + "step": 5986 + }, + { + "epoch": 4.306419708685488, + "grad_norm": 4.2625127394716875, + "learning_rate": 1.8333548972570642e-06, + "loss": 0.0925, + "step": 5987 + }, + { + "epoch": 4.3071390037762995, + "grad_norm": 0.8118648715772231, + "learning_rate": 1.8330231525708411e-06, + "loss": 0.0055, + "step": 5988 + }, + { + "epoch": 4.30785829886711, + "grad_norm": 4.24255883822549, + "learning_rate": 1.8326913907499926e-06, + "loss": 0.0586, + "step": 5989 + }, + { + "epoch": 4.308577593957922, + "grad_norm": 3.177725088905874, + "learning_rate": 1.8323596118115884e-06, + "loss": 0.0422, + "step": 5990 + }, + { + "epoch": 4.309296889048732, + "grad_norm": 3.5301693010804, + "learning_rate": 1.832027815772698e-06, + "loss": 0.0425, + "step": 5991 + }, + { + "epoch": 4.310016184139544, + "grad_norm": 3.2813534633496264, + "learning_rate": 1.8316960026503941e-06, + "loss": 0.0783, + "step": 5992 + }, + { + "epoch": 4.310735479230354, + "grad_norm": 0.06485571857720598, + "learning_rate": 1.8313641724617488e-06, + "loss": 0.0003, + "step": 5993 + }, + { + "epoch": 4.311454774321165, + "grad_norm": 2.767329281681803, + "learning_rate": 1.8310323252238341e-06, + "loss": 0.0681, + "step": 5994 + }, + { + "epoch": 4.312174069411976, + "grad_norm": 5.305922757234804, + "learning_rate": 1.8307004609537261e-06, + "loss": 0.1364, + "step": 5995 + }, + { + "epoch": 4.312893364502787, + "grad_norm": 3.629846298659511, + "learning_rate": 1.8303685796684987e-06, + "loss": 0.0722, + "step": 5996 + }, + { + "epoch": 4.313612659593598, + "grad_norm": 3.5200835654955607, + "learning_rate": 1.8300366813852279e-06, + "loss": 0.0417, + "step": 5997 + }, + { + "epoch": 4.314331954684409, + "grad_norm": 2.4064988197221435, + "learning_rate": 1.8297047661209906e-06, + "loss": 0.0512, + "step": 5998 + }, + { + "epoch": 4.3150512497752205, + "grad_norm": 4.593143417139112, + "learning_rate": 1.8293728338928638e-06, + "loss": 0.1257, + "step": 5999 + }, + { + "epoch": 4.315770544866031, + "grad_norm": 3.102389738261241, + "learning_rate": 1.829040884717927e-06, + "loss": 0.0442, + "step": 6000 + }, + { + "epoch": 4.3164898399568425, + "grad_norm": 5.585810163326435, + "learning_rate": 1.828708918613259e-06, + "loss": 0.1554, + "step": 6001 + }, + { + "epoch": 4.317209135047653, + "grad_norm": 2.512341703244468, + "learning_rate": 1.8283769355959398e-06, + "loss": 0.0672, + "step": 6002 + }, + { + "epoch": 4.317928430138465, + "grad_norm": 2.86674624122142, + "learning_rate": 1.8280449356830511e-06, + "loss": 0.0723, + "step": 6003 + }, + { + "epoch": 4.318647725229275, + "grad_norm": 3.5898280693199833, + "learning_rate": 1.8277129188916743e-06, + "loss": 0.017, + "step": 6004 + }, + { + "epoch": 4.319367020320087, + "grad_norm": 2.0713591370229065, + "learning_rate": 1.827380885238893e-06, + "loss": 0.041, + "step": 6005 + }, + { + "epoch": 4.320086315410897, + "grad_norm": 2.33466921988587, + "learning_rate": 1.8270488347417903e-06, + "loss": 0.0603, + "step": 6006 + }, + { + "epoch": 4.320805610501708, + "grad_norm": 0.7603137380995526, + "learning_rate": 1.8267167674174506e-06, + "loss": 0.0046, + "step": 6007 + }, + { + "epoch": 4.321524905592519, + "grad_norm": 3.330210622569503, + "learning_rate": 1.82638468328296e-06, + "loss": 0.0814, + "step": 6008 + }, + { + "epoch": 4.32224420068333, + "grad_norm": 3.0451959008739817, + "learning_rate": 1.8260525823554046e-06, + "loss": 0.0752, + "step": 6009 + }, + { + "epoch": 4.322963495774141, + "grad_norm": 0.07827347673911, + "learning_rate": 1.825720464651871e-06, + "loss": 0.0003, + "step": 6010 + }, + { + "epoch": 4.323682790864952, + "grad_norm": 1.5680198821476872, + "learning_rate": 1.825388330189448e-06, + "loss": 0.0078, + "step": 6011 + }, + { + "epoch": 4.3244020859557635, + "grad_norm": 4.581654435423558, + "learning_rate": 1.8250561789852244e-06, + "loss": 0.0791, + "step": 6012 + }, + { + "epoch": 4.325121381046574, + "grad_norm": 2.7577957167359695, + "learning_rate": 1.8247240110562894e-06, + "loss": 0.0419, + "step": 6013 + }, + { + "epoch": 4.3258406761373855, + "grad_norm": 2.099310065484812, + "learning_rate": 1.8243918264197343e-06, + "loss": 0.0421, + "step": 6014 + }, + { + "epoch": 4.326559971228196, + "grad_norm": 1.1280887788070966, + "learning_rate": 1.8240596250926502e-06, + "loss": 0.0109, + "step": 6015 + }, + { + "epoch": 4.327279266319008, + "grad_norm": 2.6504759738519414, + "learning_rate": 1.8237274070921294e-06, + "loss": 0.0367, + "step": 6016 + }, + { + "epoch": 4.327998561409818, + "grad_norm": 3.1539926987731977, + "learning_rate": 1.823395172435265e-06, + "loss": 0.0621, + "step": 6017 + }, + { + "epoch": 4.32871785650063, + "grad_norm": 3.132661276206971, + "learning_rate": 1.8230629211391517e-06, + "loss": 0.0801, + "step": 6018 + }, + { + "epoch": 4.32943715159144, + "grad_norm": 0.1041492481545497, + "learning_rate": 1.8227306532208844e-06, + "loss": 0.0002, + "step": 6019 + }, + { + "epoch": 4.330156446682252, + "grad_norm": 3.0455064456508136, + "learning_rate": 1.8223983686975578e-06, + "loss": 0.0634, + "step": 6020 + }, + { + "epoch": 4.330875741773062, + "grad_norm": 1.106103180682641, + "learning_rate": 1.8220660675862693e-06, + "loss": 0.0149, + "step": 6021 + }, + { + "epoch": 4.331595036863874, + "grad_norm": 3.305099444115452, + "learning_rate": 1.8217337499041164e-06, + "loss": 0.0748, + "step": 6022 + }, + { + "epoch": 4.332314331954684, + "grad_norm": 2.426305704857423, + "learning_rate": 1.8214014156681965e-06, + "loss": 0.0326, + "step": 6023 + }, + { + "epoch": 4.333033627045495, + "grad_norm": 3.6572526189060532, + "learning_rate": 1.8210690648956105e-06, + "loss": 0.0832, + "step": 6024 + }, + { + "epoch": 4.3337529221363065, + "grad_norm": 2.0069434476829273, + "learning_rate": 1.820736697603457e-06, + "loss": 0.0524, + "step": 6025 + }, + { + "epoch": 4.334472217227117, + "grad_norm": 4.669134083117361, + "learning_rate": 1.820404313808838e-06, + "loss": 0.132, + "step": 6026 + }, + { + "epoch": 4.3351915123179285, + "grad_norm": 1.5585604431430462, + "learning_rate": 1.8200719135288544e-06, + "loss": 0.0329, + "step": 6027 + }, + { + "epoch": 4.335910807408739, + "grad_norm": 2.2084355011337427, + "learning_rate": 1.8197394967806082e-06, + "loss": 0.0273, + "step": 6028 + }, + { + "epoch": 4.336630102499551, + "grad_norm": 2.9377709604301248, + "learning_rate": 1.8194070635812044e-06, + "loss": 0.0545, + "step": 6029 + }, + { + "epoch": 4.337349397590361, + "grad_norm": 0.4314663049339562, + "learning_rate": 1.8190746139477462e-06, + "loss": 0.0011, + "step": 6030 + }, + { + "epoch": 4.338068692681173, + "grad_norm": 1.145479982681039, + "learning_rate": 1.8187421478973388e-06, + "loss": 0.0157, + "step": 6031 + }, + { + "epoch": 4.338787987771983, + "grad_norm": 4.5593173854099796, + "learning_rate": 1.8184096654470886e-06, + "loss": 0.1225, + "step": 6032 + }, + { + "epoch": 4.339507282862795, + "grad_norm": 2.722846852571509, + "learning_rate": 1.8180771666141023e-06, + "loss": 0.0511, + "step": 6033 + }, + { + "epoch": 4.340226577953605, + "grad_norm": 6.701232056314367, + "learning_rate": 1.8177446514154873e-06, + "loss": 0.2083, + "step": 6034 + }, + { + "epoch": 4.340945873044417, + "grad_norm": 1.6007150746987873, + "learning_rate": 1.8174121198683523e-06, + "loss": 0.0328, + "step": 6035 + }, + { + "epoch": 4.341665168135227, + "grad_norm": 4.858494825329698, + "learning_rate": 1.817079571989806e-06, + "loss": 0.0838, + "step": 6036 + }, + { + "epoch": 4.342384463226039, + "grad_norm": 0.6621809947334634, + "learning_rate": 1.8167470077969598e-06, + "loss": 0.0075, + "step": 6037 + }, + { + "epoch": 4.3431037583168495, + "grad_norm": 1.0372006108818999, + "learning_rate": 1.8164144273069233e-06, + "loss": 0.0224, + "step": 6038 + }, + { + "epoch": 4.343823053407661, + "grad_norm": 2.6756561819739035, + "learning_rate": 1.8160818305368095e-06, + "loss": 0.0758, + "step": 6039 + }, + { + "epoch": 4.3445423484984715, + "grad_norm": 2.540114420306783, + "learning_rate": 1.8157492175037309e-06, + "loss": 0.0631, + "step": 6040 + }, + { + "epoch": 4.345261643589282, + "grad_norm": 4.081460251858338, + "learning_rate": 1.8154165882248005e-06, + "loss": 0.0174, + "step": 6041 + }, + { + "epoch": 4.345980938680094, + "grad_norm": 0.1623145218256508, + "learning_rate": 1.815083942717133e-06, + "loss": 0.0005, + "step": 6042 + }, + { + "epoch": 4.346700233770904, + "grad_norm": 0.024010338416124766, + "learning_rate": 1.8147512809978434e-06, + "loss": 0.0001, + "step": 6043 + }, + { + "epoch": 4.347419528861716, + "grad_norm": 3.5277651952114604, + "learning_rate": 1.8144186030840479e-06, + "loss": 0.0424, + "step": 6044 + }, + { + "epoch": 4.348138823952526, + "grad_norm": 2.7539793628142384, + "learning_rate": 1.8140859089928637e-06, + "loss": 0.0453, + "step": 6045 + }, + { + "epoch": 4.348858119043338, + "grad_norm": 4.416917647228344, + "learning_rate": 1.8137531987414073e-06, + "loss": 0.0773, + "step": 6046 + }, + { + "epoch": 4.349577414134148, + "grad_norm": 1.9367185249398302, + "learning_rate": 1.8134204723467986e-06, + "loss": 0.0492, + "step": 6047 + }, + { + "epoch": 4.35029670922496, + "grad_norm": 1.2685809495654297, + "learning_rate": 1.8130877298261565e-06, + "loss": 0.0109, + "step": 6048 + }, + { + "epoch": 4.35101600431577, + "grad_norm": 0.6952509715481375, + "learning_rate": 1.8127549711966009e-06, + "loss": 0.0058, + "step": 6049 + }, + { + "epoch": 4.351735299406582, + "grad_norm": 2.305112784199429, + "learning_rate": 1.8124221964752535e-06, + "loss": 0.0255, + "step": 6050 + }, + { + "epoch": 4.3524545944973925, + "grad_norm": 2.398855518782349, + "learning_rate": 1.812089405679235e-06, + "loss": 0.0437, + "step": 6051 + }, + { + "epoch": 4.353173889588204, + "grad_norm": 2.1426717915332567, + "learning_rate": 1.8117565988256685e-06, + "loss": 0.0296, + "step": 6052 + }, + { + "epoch": 4.3538931846790145, + "grad_norm": 3.9929585866166057, + "learning_rate": 1.8114237759316785e-06, + "loss": 0.1187, + "step": 6053 + }, + { + "epoch": 4.354612479769825, + "grad_norm": 0.9174071239434606, + "learning_rate": 1.8110909370143879e-06, + "loss": 0.01, + "step": 6054 + }, + { + "epoch": 4.355331774860637, + "grad_norm": 1.2546723233511283, + "learning_rate": 1.8107580820909233e-06, + "loss": 0.0291, + "step": 6055 + }, + { + "epoch": 4.356051069951447, + "grad_norm": 2.709979716584753, + "learning_rate": 1.8104252111784094e-06, + "loss": 0.0282, + "step": 6056 + }, + { + "epoch": 4.356770365042259, + "grad_norm": 1.0863203048436607, + "learning_rate": 1.810092324293973e-06, + "loss": 0.0251, + "step": 6057 + }, + { + "epoch": 4.357489660133069, + "grad_norm": 2.2768896797083538, + "learning_rate": 1.8097594214547427e-06, + "loss": 0.0511, + "step": 6058 + }, + { + "epoch": 4.358208955223881, + "grad_norm": 1.7874394942505287, + "learning_rate": 1.809426502677846e-06, + "loss": 0.0419, + "step": 6059 + }, + { + "epoch": 4.358928250314691, + "grad_norm": 2.7642826858983867, + "learning_rate": 1.8090935679804127e-06, + "loss": 0.059, + "step": 6060 + }, + { + "epoch": 4.359647545405503, + "grad_norm": 5.52343108867064, + "learning_rate": 1.8087606173795732e-06, + "loss": 0.1276, + "step": 6061 + }, + { + "epoch": 4.360366840496313, + "grad_norm": 1.5647462076908474, + "learning_rate": 1.8084276508924575e-06, + "loss": 0.0245, + "step": 6062 + }, + { + "epoch": 4.361086135587125, + "grad_norm": 3.667425017382983, + "learning_rate": 1.808094668536198e-06, + "loss": 0.0577, + "step": 6063 + }, + { + "epoch": 4.3618054306779355, + "grad_norm": 3.796300151895494, + "learning_rate": 1.8077616703279264e-06, + "loss": 0.0984, + "step": 6064 + }, + { + "epoch": 4.362524725768747, + "grad_norm": 1.9474192449592544, + "learning_rate": 1.807428656284777e-06, + "loss": 0.0486, + "step": 6065 + }, + { + "epoch": 4.3632440208595575, + "grad_norm": 1.7707056877563414, + "learning_rate": 1.8070956264238835e-06, + "loss": 0.0427, + "step": 6066 + }, + { + "epoch": 4.363963315950369, + "grad_norm": 4.8811923008403815, + "learning_rate": 1.8067625807623807e-06, + "loss": 0.1275, + "step": 6067 + }, + { + "epoch": 4.36468261104118, + "grad_norm": 0.3463874120449742, + "learning_rate": 1.806429519317405e-06, + "loss": 0.0008, + "step": 6068 + }, + { + "epoch": 4.365401906131991, + "grad_norm": 5.0356977870711646, + "learning_rate": 1.8060964421060925e-06, + "loss": 0.1404, + "step": 6069 + }, + { + "epoch": 4.366121201222802, + "grad_norm": 2.2382855151929006, + "learning_rate": 1.8057633491455809e-06, + "loss": 0.0639, + "step": 6070 + }, + { + "epoch": 4.366840496313612, + "grad_norm": 0.049567319643360946, + "learning_rate": 1.8054302404530082e-06, + "loss": 0.0002, + "step": 6071 + }, + { + "epoch": 4.367559791404424, + "grad_norm": 0.7730510507138303, + "learning_rate": 1.8050971160455132e-06, + "loss": 0.0053, + "step": 6072 + }, + { + "epoch": 4.368279086495234, + "grad_norm": 2.72533503700837, + "learning_rate": 1.804763975940236e-06, + "loss": 0.066, + "step": 6073 + }, + { + "epoch": 4.368998381586046, + "grad_norm": 2.6044860690722342, + "learning_rate": 1.8044308201543171e-06, + "loss": 0.0628, + "step": 6074 + }, + { + "epoch": 4.369717676676856, + "grad_norm": 2.2966159094388527, + "learning_rate": 1.8040976487048982e-06, + "loss": 0.0237, + "step": 6075 + }, + { + "epoch": 4.370436971767668, + "grad_norm": 3.8370234248598685, + "learning_rate": 1.8037644616091219e-06, + "loss": 0.0624, + "step": 6076 + }, + { + "epoch": 4.3711562668584785, + "grad_norm": 0.5145195219383986, + "learning_rate": 1.8034312588841305e-06, + "loss": 0.0014, + "step": 6077 + }, + { + "epoch": 4.37187556194929, + "grad_norm": 3.8984766957719748, + "learning_rate": 1.8030980405470681e-06, + "loss": 0.1034, + "step": 6078 + }, + { + "epoch": 4.3725948570401005, + "grad_norm": 4.501620064547736, + "learning_rate": 1.8027648066150797e-06, + "loss": 0.0532, + "step": 6079 + }, + { + "epoch": 4.373314152130912, + "grad_norm": 2.8928271982071347, + "learning_rate": 1.80243155710531e-06, + "loss": 0.0566, + "step": 6080 + }, + { + "epoch": 4.374033447221723, + "grad_norm": 1.848440435229474, + "learning_rate": 1.802098292034906e-06, + "loss": 0.022, + "step": 6081 + }, + { + "epoch": 4.374752742312534, + "grad_norm": 4.428182503124213, + "learning_rate": 1.8017650114210146e-06, + "loss": 0.1565, + "step": 6082 + }, + { + "epoch": 4.375472037403345, + "grad_norm": 0.8924430131816987, + "learning_rate": 1.8014317152807838e-06, + "loss": 0.0088, + "step": 6083 + }, + { + "epoch": 4.376191332494155, + "grad_norm": 3.178441498781897, + "learning_rate": 1.8010984036313619e-06, + "loss": 0.0707, + "step": 6084 + }, + { + "epoch": 4.376910627584967, + "grad_norm": 2.5286563243088866, + "learning_rate": 1.800765076489899e-06, + "loss": 0.0199, + "step": 6085 + }, + { + "epoch": 4.377629922675777, + "grad_norm": 2.210918924334111, + "learning_rate": 1.8004317338735446e-06, + "loss": 0.0611, + "step": 6086 + }, + { + "epoch": 4.378349217766589, + "grad_norm": 3.742244842073366, + "learning_rate": 1.8000983757994499e-06, + "loss": 0.1079, + "step": 6087 + }, + { + "epoch": 4.379068512857399, + "grad_norm": 1.7597533548467013, + "learning_rate": 1.7997650022847672e-06, + "loss": 0.0531, + "step": 6088 + }, + { + "epoch": 4.379787807948211, + "grad_norm": 4.122543736935282, + "learning_rate": 1.7994316133466488e-06, + "loss": 0.0873, + "step": 6089 + }, + { + "epoch": 4.3805071030390215, + "grad_norm": 0.8524641953319322, + "learning_rate": 1.7990982090022483e-06, + "loss": 0.0175, + "step": 6090 + }, + { + "epoch": 4.381226398129833, + "grad_norm": 3.57873297594927, + "learning_rate": 1.79876478926872e-06, + "loss": 0.1136, + "step": 6091 + }, + { + "epoch": 4.3819456932206435, + "grad_norm": 1.6886365317279626, + "learning_rate": 1.7984313541632188e-06, + "loss": 0.0299, + "step": 6092 + }, + { + "epoch": 4.382664988311455, + "grad_norm": 3.879001621547681, + "learning_rate": 1.7980979037029002e-06, + "loss": 0.1663, + "step": 6093 + }, + { + "epoch": 4.383384283402266, + "grad_norm": 0.04885495433227142, + "learning_rate": 1.7977644379049217e-06, + "loss": 0.0003, + "step": 6094 + }, + { + "epoch": 4.384103578493077, + "grad_norm": 0.3080941272364588, + "learning_rate": 1.7974309567864398e-06, + "loss": 0.001, + "step": 6095 + }, + { + "epoch": 4.384822873583888, + "grad_norm": 2.2754427419517245, + "learning_rate": 1.797097460364613e-06, + "loss": 0.0642, + "step": 6096 + }, + { + "epoch": 4.385542168674699, + "grad_norm": 2.853673922957916, + "learning_rate": 1.7967639486566009e-06, + "loss": 0.0181, + "step": 6097 + }, + { + "epoch": 4.38626146376551, + "grad_norm": 3.0954230251633335, + "learning_rate": 1.7964304216795622e-06, + "loss": 0.0653, + "step": 6098 + }, + { + "epoch": 4.386980758856321, + "grad_norm": 2.3303528931161046, + "learning_rate": 1.7960968794506582e-06, + "loss": 0.0523, + "step": 6099 + }, + { + "epoch": 4.387700053947132, + "grad_norm": 5.53540657644375, + "learning_rate": 1.7957633219870497e-06, + "loss": 0.0555, + "step": 6100 + }, + { + "epoch": 4.388419349037942, + "grad_norm": 5.230235632218612, + "learning_rate": 1.7954297493058988e-06, + "loss": 0.1631, + "step": 6101 + }, + { + "epoch": 4.389138644128754, + "grad_norm": 3.5510832988923933, + "learning_rate": 1.7950961614243691e-06, + "loss": 0.1085, + "step": 6102 + }, + { + "epoch": 4.3898579392195645, + "grad_norm": 3.634443373654936, + "learning_rate": 1.794762558359624e-06, + "loss": 0.0661, + "step": 6103 + }, + { + "epoch": 4.390577234310376, + "grad_norm": 3.6193553606295557, + "learning_rate": 1.7944289401288277e-06, + "loss": 0.0087, + "step": 6104 + }, + { + "epoch": 4.3912965294011865, + "grad_norm": 3.7108257563320897, + "learning_rate": 1.7940953067491454e-06, + "loss": 0.0453, + "step": 6105 + }, + { + "epoch": 4.392015824491998, + "grad_norm": 2.4816931462708474, + "learning_rate": 1.7937616582377434e-06, + "loss": 0.038, + "step": 6106 + }, + { + "epoch": 4.392735119582809, + "grad_norm": 0.22943908563819748, + "learning_rate": 1.7934279946117888e-06, + "loss": 0.0006, + "step": 6107 + }, + { + "epoch": 4.39345441467362, + "grad_norm": 1.925547466000469, + "learning_rate": 1.7930943158884483e-06, + "loss": 0.044, + "step": 6108 + }, + { + "epoch": 4.394173709764431, + "grad_norm": 1.1861514272963691, + "learning_rate": 1.7927606220848908e-06, + "loss": 0.0205, + "step": 6109 + }, + { + "epoch": 4.394893004855242, + "grad_norm": 3.7964531653587814, + "learning_rate": 1.7924269132182854e-06, + "loss": 0.0594, + "step": 6110 + }, + { + "epoch": 4.395612299946053, + "grad_norm": 2.1824969063445483, + "learning_rate": 1.792093189305802e-06, + "loss": 0.0264, + "step": 6111 + }, + { + "epoch": 4.396331595036864, + "grad_norm": 2.891396124661877, + "learning_rate": 1.791759450364611e-06, + "loss": 0.0383, + "step": 6112 + }, + { + "epoch": 4.397050890127675, + "grad_norm": 3.5199362315689493, + "learning_rate": 1.7914256964118845e-06, + "loss": 0.093, + "step": 6113 + }, + { + "epoch": 4.397770185218486, + "grad_norm": 2.547332243775144, + "learning_rate": 1.7910919274647944e-06, + "loss": 0.0673, + "step": 6114 + }, + { + "epoch": 4.398489480309297, + "grad_norm": 4.419801210172589, + "learning_rate": 1.7907581435405135e-06, + "loss": 0.106, + "step": 6115 + }, + { + "epoch": 4.399208775400108, + "grad_norm": 0.034305537553320754, + "learning_rate": 1.7904243446562155e-06, + "loss": 0.0001, + "step": 6116 + }, + { + "epoch": 4.399928070490919, + "grad_norm": 3.215391658365198, + "learning_rate": 1.7900905308290754e-06, + "loss": 0.0699, + "step": 6117 + }, + { + "epoch": 4.4006473655817295, + "grad_norm": 2.6868505360374044, + "learning_rate": 1.789756702076268e-06, + "loss": 0.0486, + "step": 6118 + }, + { + "epoch": 4.401366660672541, + "grad_norm": 1.0419830032026987, + "learning_rate": 1.78942285841497e-06, + "loss": 0.0076, + "step": 6119 + }, + { + "epoch": 4.402085955763352, + "grad_norm": 2.495737802377356, + "learning_rate": 1.7890889998623578e-06, + "loss": 0.0255, + "step": 6120 + }, + { + "epoch": 4.402805250854163, + "grad_norm": 0.8006658643923864, + "learning_rate": 1.788755126435609e-06, + "loss": 0.003, + "step": 6121 + }, + { + "epoch": 4.403524545944974, + "grad_norm": 1.091323709987208, + "learning_rate": 1.7884212381519017e-06, + "loss": 0.0115, + "step": 6122 + }, + { + "epoch": 4.404243841035785, + "grad_norm": 2.8957687914663985, + "learning_rate": 1.7880873350284154e-06, + "loss": 0.0038, + "step": 6123 + }, + { + "epoch": 4.404963136126596, + "grad_norm": 0.5367827427739006, + "learning_rate": 1.7877534170823305e-06, + "loss": 0.0034, + "step": 6124 + }, + { + "epoch": 4.405682431217407, + "grad_norm": 2.7158457777505975, + "learning_rate": 1.7874194843308267e-06, + "loss": 0.0608, + "step": 6125 + }, + { + "epoch": 4.406401726308218, + "grad_norm": 2.7058016323614935, + "learning_rate": 1.7870855367910863e-06, + "loss": 0.0605, + "step": 6126 + }, + { + "epoch": 4.407121021399029, + "grad_norm": 2.8177960858695292, + "learning_rate": 1.7867515744802908e-06, + "loss": 0.0517, + "step": 6127 + }, + { + "epoch": 4.40784031648984, + "grad_norm": 0.8724775080698774, + "learning_rate": 1.7864175974156236e-06, + "loss": 0.0042, + "step": 6128 + }, + { + "epoch": 4.408559611580651, + "grad_norm": 5.571649608380618, + "learning_rate": 1.7860836056142682e-06, + "loss": 0.1537, + "step": 6129 + }, + { + "epoch": 4.409278906671462, + "grad_norm": 2.15510866445969, + "learning_rate": 1.7857495990934086e-06, + "loss": 0.0512, + "step": 6130 + }, + { + "epoch": 4.4099982017622725, + "grad_norm": 0.5075021552211338, + "learning_rate": 1.785415577870231e-06, + "loss": 0.0011, + "step": 6131 + }, + { + "epoch": 4.410717496853084, + "grad_norm": 5.024786145950193, + "learning_rate": 1.785081541961921e-06, + "loss": 0.0928, + "step": 6132 + }, + { + "epoch": 4.411436791943895, + "grad_norm": 3.4434971673212424, + "learning_rate": 1.7847474913856647e-06, + "loss": 0.0751, + "step": 6133 + }, + { + "epoch": 4.412156087034706, + "grad_norm": 3.3503119135772352, + "learning_rate": 1.7844134261586506e-06, + "loss": 0.0672, + "step": 6134 + }, + { + "epoch": 4.412875382125517, + "grad_norm": 4.6934567552083335, + "learning_rate": 1.7840793462980663e-06, + "loss": 0.1121, + "step": 6135 + }, + { + "epoch": 4.413594677216328, + "grad_norm": 3.0360589493601284, + "learning_rate": 1.7837452518211011e-06, + "loss": 0.0591, + "step": 6136 + }, + { + "epoch": 4.414313972307139, + "grad_norm": 3.776832691909125, + "learning_rate": 1.7834111427449442e-06, + "loss": 0.0913, + "step": 6137 + }, + { + "epoch": 4.41503326739795, + "grad_norm": 2.7265888845238733, + "learning_rate": 1.7830770190867865e-06, + "loss": 0.0898, + "step": 6138 + }, + { + "epoch": 4.415752562488761, + "grad_norm": 2.92763405842064, + "learning_rate": 1.7827428808638197e-06, + "loss": 0.012, + "step": 6139 + }, + { + "epoch": 4.416471857579572, + "grad_norm": 2.390260142651457, + "learning_rate": 1.782408728093235e-06, + "loss": 0.0459, + "step": 6140 + }, + { + "epoch": 4.417191152670383, + "grad_norm": 2.452213909230455, + "learning_rate": 1.7820745607922254e-06, + "loss": 0.0314, + "step": 6141 + }, + { + "epoch": 4.417910447761194, + "grad_norm": 2.742815622846682, + "learning_rate": 1.7817403789779846e-06, + "loss": 0.0644, + "step": 6142 + }, + { + "epoch": 4.418629742852005, + "grad_norm": 7.591778993207834, + "learning_rate": 1.7814061826677067e-06, + "loss": 0.1172, + "step": 6143 + }, + { + "epoch": 4.419349037942816, + "grad_norm": 3.485423330602173, + "learning_rate": 1.7810719718785873e-06, + "loss": 0.1295, + "step": 6144 + }, + { + "epoch": 4.420068333033627, + "grad_norm": 1.604394604620596, + "learning_rate": 1.780737746627821e-06, + "loss": 0.0044, + "step": 6145 + }, + { + "epoch": 4.4207876281244385, + "grad_norm": 2.452989373663375, + "learning_rate": 1.7804035069326044e-06, + "loss": 0.0426, + "step": 6146 + }, + { + "epoch": 4.421506923215249, + "grad_norm": 2.461810532081175, + "learning_rate": 1.7800692528101357e-06, + "loss": 0.0366, + "step": 6147 + }, + { + "epoch": 4.42222621830606, + "grad_norm": 2.394372377380664, + "learning_rate": 1.7797349842776124e-06, + "loss": 0.0685, + "step": 6148 + }, + { + "epoch": 4.422945513396871, + "grad_norm": 3.659609516771946, + "learning_rate": 1.7794007013522325e-06, + "loss": 0.0913, + "step": 6149 + }, + { + "epoch": 4.423664808487682, + "grad_norm": 0.599705933416922, + "learning_rate": 1.779066404051197e-06, + "loss": 0.0024, + "step": 6150 + }, + { + "epoch": 4.424384103578493, + "grad_norm": 3.112812576135934, + "learning_rate": 1.778732092391704e-06, + "loss": 0.0913, + "step": 6151 + }, + { + "epoch": 4.425103398669304, + "grad_norm": 0.9118382420069862, + "learning_rate": 1.7783977663909562e-06, + "loss": 0.0026, + "step": 6152 + }, + { + "epoch": 4.425822693760115, + "grad_norm": 4.501649580773762, + "learning_rate": 1.7780634260661546e-06, + "loss": 0.0331, + "step": 6153 + }, + { + "epoch": 4.426541988850926, + "grad_norm": 3.559104403514001, + "learning_rate": 1.7777290714345015e-06, + "loss": 0.0982, + "step": 6154 + }, + { + "epoch": 4.427261283941737, + "grad_norm": 4.678837949644355, + "learning_rate": 1.7773947025132003e-06, + "loss": 0.1833, + "step": 6155 + }, + { + "epoch": 4.427980579032548, + "grad_norm": 4.236169002178598, + "learning_rate": 1.7770603193194545e-06, + "loss": 0.0465, + "step": 6156 + }, + { + "epoch": 4.428699874123359, + "grad_norm": 2.890506217724969, + "learning_rate": 1.776725921870469e-06, + "loss": 0.0858, + "step": 6157 + }, + { + "epoch": 4.42941916921417, + "grad_norm": 1.8319593263432643, + "learning_rate": 1.776391510183449e-06, + "loss": 0.0412, + "step": 6158 + }, + { + "epoch": 4.4301384643049815, + "grad_norm": 3.521782003524683, + "learning_rate": 1.7760570842756004e-06, + "loss": 0.0842, + "step": 6159 + }, + { + "epoch": 4.430857759395792, + "grad_norm": 3.848876831578218, + "learning_rate": 1.7757226441641304e-06, + "loss": 0.1027, + "step": 6160 + }, + { + "epoch": 4.431577054486603, + "grad_norm": 1.6779098326130952, + "learning_rate": 1.7753881898662464e-06, + "loss": 0.0267, + "step": 6161 + }, + { + "epoch": 4.432296349577414, + "grad_norm": 0.26367630749206666, + "learning_rate": 1.7750537213991561e-06, + "loss": 0.0003, + "step": 6162 + }, + { + "epoch": 4.433015644668225, + "grad_norm": 1.193406021147191, + "learning_rate": 1.7747192387800697e-06, + "loss": 0.0236, + "step": 6163 + }, + { + "epoch": 4.433734939759036, + "grad_norm": 4.723526608417339, + "learning_rate": 1.774384742026196e-06, + "loss": 0.1249, + "step": 6164 + }, + { + "epoch": 4.434454234849847, + "grad_norm": 0.11266843034537984, + "learning_rate": 1.7740502311547458e-06, + "loss": 0.0003, + "step": 6165 + }, + { + "epoch": 4.435173529940658, + "grad_norm": 0.4209404854139488, + "learning_rate": 1.7737157061829302e-06, + "loss": 0.0014, + "step": 6166 + }, + { + "epoch": 4.435892825031469, + "grad_norm": 4.020354636210274, + "learning_rate": 1.7733811671279604e-06, + "loss": 0.0502, + "step": 6167 + }, + { + "epoch": 4.43661212012228, + "grad_norm": 1.9570077346470218, + "learning_rate": 1.7730466140070502e-06, + "loss": 0.0408, + "step": 6168 + }, + { + "epoch": 4.437331415213091, + "grad_norm": 1.0026881282793554, + "learning_rate": 1.7727120468374122e-06, + "loss": 0.0025, + "step": 6169 + }, + { + "epoch": 4.438050710303902, + "grad_norm": 4.046591123921115, + "learning_rate": 1.7723774656362603e-06, + "loss": 0.1943, + "step": 6170 + }, + { + "epoch": 4.438770005394713, + "grad_norm": 2.554449833240542, + "learning_rate": 1.77204287042081e-06, + "loss": 0.0229, + "step": 6171 + }, + { + "epoch": 4.4394893004855245, + "grad_norm": 0.08936641913790921, + "learning_rate": 1.7717082612082768e-06, + "loss": 0.0004, + "step": 6172 + }, + { + "epoch": 4.440208595576335, + "grad_norm": 3.192545804838201, + "learning_rate": 1.7713736380158761e-06, + "loss": 0.0494, + "step": 6173 + }, + { + "epoch": 4.4409278906671465, + "grad_norm": 6.344781450833596, + "learning_rate": 1.7710390008608257e-06, + "loss": 0.1403, + "step": 6174 + }, + { + "epoch": 4.441647185757957, + "grad_norm": 2.938959881494209, + "learning_rate": 1.7707043497603426e-06, + "loss": 0.0447, + "step": 6175 + }, + { + "epoch": 4.442366480848769, + "grad_norm": 6.752451999144458, + "learning_rate": 1.7703696847316456e-06, + "loss": 0.0943, + "step": 6176 + }, + { + "epoch": 4.443085775939579, + "grad_norm": 0.13337032336901095, + "learning_rate": 1.7700350057919534e-06, + "loss": 0.0004, + "step": 6177 + }, + { + "epoch": 4.44380507103039, + "grad_norm": 4.345937788328146, + "learning_rate": 1.7697003129584869e-06, + "loss": 0.1339, + "step": 6178 + }, + { + "epoch": 4.444524366121201, + "grad_norm": 5.212907465245972, + "learning_rate": 1.7693656062484656e-06, + "loss": 0.1629, + "step": 6179 + }, + { + "epoch": 4.445243661212012, + "grad_norm": 1.9742559946817129, + "learning_rate": 1.7690308856791106e-06, + "loss": 0.0377, + "step": 6180 + }, + { + "epoch": 4.445962956302823, + "grad_norm": 2.3866329834994993, + "learning_rate": 1.7686961512676442e-06, + "loss": 0.0556, + "step": 6181 + }, + { + "epoch": 4.446682251393634, + "grad_norm": 2.409744291061262, + "learning_rate": 1.7683614030312898e-06, + "loss": 0.0665, + "step": 6182 + }, + { + "epoch": 4.447401546484445, + "grad_norm": 3.257340889654492, + "learning_rate": 1.7680266409872693e-06, + "loss": 0.064, + "step": 6183 + }, + { + "epoch": 4.448120841575256, + "grad_norm": 0.03909276854895944, + "learning_rate": 1.767691865152808e-06, + "loss": 0.0001, + "step": 6184 + }, + { + "epoch": 4.4488401366660675, + "grad_norm": 2.0696779747637124, + "learning_rate": 1.7673570755451303e-06, + "loss": 0.045, + "step": 6185 + }, + { + "epoch": 4.449559431756878, + "grad_norm": 3.631375685245719, + "learning_rate": 1.767022272181462e-06, + "loss": 0.0772, + "step": 6186 + }, + { + "epoch": 4.4502787268476895, + "grad_norm": 4.306395135236247, + "learning_rate": 1.7666874550790289e-06, + "loss": 0.0539, + "step": 6187 + }, + { + "epoch": 4.4509980219385, + "grad_norm": 2.147545738063925, + "learning_rate": 1.766352624255058e-06, + "loss": 0.0086, + "step": 6188 + }, + { + "epoch": 4.451717317029312, + "grad_norm": 4.712429967130418, + "learning_rate": 1.7660177797267768e-06, + "loss": 0.1063, + "step": 6189 + }, + { + "epoch": 4.452436612120122, + "grad_norm": 2.732169568981602, + "learning_rate": 1.765682921511414e-06, + "loss": 0.0329, + "step": 6190 + }, + { + "epoch": 4.453155907210934, + "grad_norm": 0.4123680806666529, + "learning_rate": 1.7653480496261987e-06, + "loss": 0.0011, + "step": 6191 + }, + { + "epoch": 4.453875202301744, + "grad_norm": 3.564698975039429, + "learning_rate": 1.7650131640883605e-06, + "loss": 0.1043, + "step": 6192 + }, + { + "epoch": 4.454594497392556, + "grad_norm": 4.971551055593641, + "learning_rate": 1.7646782649151297e-06, + "loss": 0.1578, + "step": 6193 + }, + { + "epoch": 4.455313792483366, + "grad_norm": 3.7799866254248364, + "learning_rate": 1.7643433521237374e-06, + "loss": 0.0308, + "step": 6194 + }, + { + "epoch": 4.456033087574177, + "grad_norm": 1.2096603202167822, + "learning_rate": 1.7640084257314158e-06, + "loss": 0.0119, + "step": 6195 + }, + { + "epoch": 4.456752382664988, + "grad_norm": 3.1725028435679157, + "learning_rate": 1.7636734857553967e-06, + "loss": 0.0519, + "step": 6196 + }, + { + "epoch": 4.457471677755799, + "grad_norm": 3.511813830587288, + "learning_rate": 1.7633385322129144e-06, + "loss": 0.0968, + "step": 6197 + }, + { + "epoch": 4.4581909728466105, + "grad_norm": 3.273910599156768, + "learning_rate": 1.763003565121202e-06, + "loss": 0.0828, + "step": 6198 + }, + { + "epoch": 4.458910267937421, + "grad_norm": 0.44886004958524206, + "learning_rate": 1.7626685844974951e-06, + "loss": 0.0031, + "step": 6199 + }, + { + "epoch": 4.4596295630282325, + "grad_norm": 0.03618440135567915, + "learning_rate": 1.7623335903590282e-06, + "loss": 0.0001, + "step": 6200 + }, + { + "epoch": 4.460348858119043, + "grad_norm": 2.673310139014084, + "learning_rate": 1.7619985827230375e-06, + "loss": 0.0659, + "step": 6201 + }, + { + "epoch": 4.461068153209855, + "grad_norm": 3.196494895381535, + "learning_rate": 1.7616635616067594e-06, + "loss": 0.0391, + "step": 6202 + }, + { + "epoch": 4.461787448300665, + "grad_norm": 3.369690923917561, + "learning_rate": 1.7613285270274322e-06, + "loss": 0.0811, + "step": 6203 + }, + { + "epoch": 4.462506743391477, + "grad_norm": 3.282607205493699, + "learning_rate": 1.760993479002293e-06, + "loss": 0.064, + "step": 6204 + }, + { + "epoch": 4.463226038482287, + "grad_norm": 2.888893141100408, + "learning_rate": 1.760658417548582e-06, + "loss": 0.0349, + "step": 6205 + }, + { + "epoch": 4.463945333573099, + "grad_norm": 1.1154187039685335, + "learning_rate": 1.7603233426835368e-06, + "loss": 0.0066, + "step": 6206 + }, + { + "epoch": 4.464664628663909, + "grad_norm": 0.5801202732930397, + "learning_rate": 1.7599882544243993e-06, + "loss": 0.0015, + "step": 6207 + }, + { + "epoch": 4.46538392375472, + "grad_norm": 5.106461781417456, + "learning_rate": 1.7596531527884098e-06, + "loss": 0.0855, + "step": 6208 + }, + { + "epoch": 4.466103218845531, + "grad_norm": 3.735804854269762, + "learning_rate": 1.7593180377928094e-06, + "loss": 0.0955, + "step": 6209 + }, + { + "epoch": 4.466822513936342, + "grad_norm": 2.4028830959143637, + "learning_rate": 1.758982909454841e-06, + "loss": 0.0683, + "step": 6210 + }, + { + "epoch": 4.4675418090271535, + "grad_norm": 7.558723036873182, + "learning_rate": 1.758647767791747e-06, + "loss": 0.0115, + "step": 6211 + }, + { + "epoch": 4.468261104117964, + "grad_norm": 1.3063276048774657, + "learning_rate": 1.7583126128207717e-06, + "loss": 0.0084, + "step": 6212 + }, + { + "epoch": 4.4689803992087755, + "grad_norm": 2.4893784949909263, + "learning_rate": 1.757977444559159e-06, + "loss": 0.0474, + "step": 6213 + }, + { + "epoch": 4.469699694299586, + "grad_norm": 1.7676481411507112, + "learning_rate": 1.7576422630241534e-06, + "loss": 0.0214, + "step": 6214 + }, + { + "epoch": 4.470418989390398, + "grad_norm": 2.389656482588371, + "learning_rate": 1.757307068233002e-06, + "loss": 0.0602, + "step": 6215 + }, + { + "epoch": 4.471138284481208, + "grad_norm": 0.4687441698758167, + "learning_rate": 1.7569718602029496e-06, + "loss": 0.001, + "step": 6216 + }, + { + "epoch": 4.47185757957202, + "grad_norm": 1.856185255132, + "learning_rate": 1.7566366389512438e-06, + "loss": 0.0423, + "step": 6217 + }, + { + "epoch": 4.47257687466283, + "grad_norm": 0.08927003278858267, + "learning_rate": 1.7563014044951327e-06, + "loss": 0.0005, + "step": 6218 + }, + { + "epoch": 4.473296169753642, + "grad_norm": 0.30074305831455234, + "learning_rate": 1.7559661568518638e-06, + "loss": 0.0004, + "step": 6219 + }, + { + "epoch": 4.474015464844452, + "grad_norm": 0.2828249655495781, + "learning_rate": 1.7556308960386874e-06, + "loss": 0.0007, + "step": 6220 + }, + { + "epoch": 4.474734759935264, + "grad_norm": 2.193181655706693, + "learning_rate": 1.7552956220728528e-06, + "loss": 0.0206, + "step": 6221 + }, + { + "epoch": 4.475454055026074, + "grad_norm": 4.947844336288775, + "learning_rate": 1.7549603349716101e-06, + "loss": 0.096, + "step": 6222 + }, + { + "epoch": 4.476173350116886, + "grad_norm": 4.771519048161475, + "learning_rate": 1.7546250347522106e-06, + "loss": 0.0701, + "step": 6223 + }, + { + "epoch": 4.4768926452076965, + "grad_norm": 1.7354680980068953, + "learning_rate": 1.754289721431906e-06, + "loss": 0.0346, + "step": 6224 + }, + { + "epoch": 4.477611940298507, + "grad_norm": 4.38794363200039, + "learning_rate": 1.7539543950279486e-06, + "loss": 0.0629, + "step": 6225 + }, + { + "epoch": 4.4783312353893185, + "grad_norm": 1.495398121440337, + "learning_rate": 1.7536190555575923e-06, + "loss": 0.0029, + "step": 6226 + }, + { + "epoch": 4.479050530480129, + "grad_norm": 6.592618061064994, + "learning_rate": 1.7532837030380899e-06, + "loss": 0.1663, + "step": 6227 + }, + { + "epoch": 4.479769825570941, + "grad_norm": 2.6775091937411553, + "learning_rate": 1.7529483374866965e-06, + "loss": 0.0443, + "step": 6228 + }, + { + "epoch": 4.480489120661751, + "grad_norm": 3.2918075880001765, + "learning_rate": 1.7526129589206672e-06, + "loss": 0.0443, + "step": 6229 + }, + { + "epoch": 4.481208415752563, + "grad_norm": 0.04575260062742102, + "learning_rate": 1.752277567357258e-06, + "loss": 0.0001, + "step": 6230 + }, + { + "epoch": 4.481927710843373, + "grad_norm": 3.641729124938048, + "learning_rate": 1.7519421628137246e-06, + "loss": 0.073, + "step": 6231 + }, + { + "epoch": 4.482647005934185, + "grad_norm": 4.61798865832625, + "learning_rate": 1.7516067453073246e-06, + "loss": 0.039, + "step": 6232 + }, + { + "epoch": 4.483366301024995, + "grad_norm": 1.3066812486927704, + "learning_rate": 1.751271314855316e-06, + "loss": 0.0144, + "step": 6233 + }, + { + "epoch": 4.484085596115807, + "grad_norm": 3.5020499393975673, + "learning_rate": 1.7509358714749575e-06, + "loss": 0.0609, + "step": 6234 + }, + { + "epoch": 4.484804891206617, + "grad_norm": 5.976843173390546, + "learning_rate": 1.7506004151835073e-06, + "loss": 0.0333, + "step": 6235 + }, + { + "epoch": 4.485524186297429, + "grad_norm": 3.173765477311921, + "learning_rate": 1.7502649459982264e-06, + "loss": 0.0948, + "step": 6236 + }, + { + "epoch": 4.4862434813882395, + "grad_norm": 1.4877889565317082, + "learning_rate": 1.7499294639363746e-06, + "loss": 0.0249, + "step": 6237 + }, + { + "epoch": 4.48696277647905, + "grad_norm": 5.247910399634501, + "learning_rate": 1.749593969015213e-06, + "loss": 0.1574, + "step": 6238 + }, + { + "epoch": 4.4876820715698615, + "grad_norm": 2.680575772813916, + "learning_rate": 1.7492584612520036e-06, + "loss": 0.058, + "step": 6239 + }, + { + "epoch": 4.488401366660672, + "grad_norm": 2.291676808288746, + "learning_rate": 1.7489229406640087e-06, + "loss": 0.0331, + "step": 6240 + }, + { + "epoch": 4.489120661751484, + "grad_norm": 7.634014355271276, + "learning_rate": 1.7485874072684917e-06, + "loss": 0.1448, + "step": 6241 + }, + { + "epoch": 4.489839956842294, + "grad_norm": 2.6071409849572422, + "learning_rate": 1.7482518610827162e-06, + "loss": 0.0418, + "step": 6242 + }, + { + "epoch": 4.490559251933106, + "grad_norm": 1.173723825807178, + "learning_rate": 1.7479163021239462e-06, + "loss": 0.0027, + "step": 6243 + }, + { + "epoch": 4.491278547023916, + "grad_norm": 3.558249898980863, + "learning_rate": 1.7475807304094479e-06, + "loss": 0.085, + "step": 6244 + }, + { + "epoch": 4.491997842114728, + "grad_norm": 6.0047447408269425, + "learning_rate": 1.7472451459564867e-06, + "loss": 0.1857, + "step": 6245 + }, + { + "epoch": 4.492717137205538, + "grad_norm": 3.977262858104638, + "learning_rate": 1.7469095487823277e-06, + "loss": 0.1193, + "step": 6246 + }, + { + "epoch": 4.49343643229635, + "grad_norm": 4.3075207075631, + "learning_rate": 1.7465739389042395e-06, + "loss": 0.1908, + "step": 6247 + }, + { + "epoch": 4.49415572738716, + "grad_norm": 1.3701326433414747, + "learning_rate": 1.7462383163394889e-06, + "loss": 0.0203, + "step": 6248 + }, + { + "epoch": 4.494875022477972, + "grad_norm": 3.3283957160556192, + "learning_rate": 1.7459026811053452e-06, + "loss": 0.065, + "step": 6249 + }, + { + "epoch": 4.4955943175687825, + "grad_norm": 0.8913660219018149, + "learning_rate": 1.7455670332190768e-06, + "loss": 0.0023, + "step": 6250 + }, + { + "epoch": 4.496313612659594, + "grad_norm": 1.6677606631128188, + "learning_rate": 1.7452313726979536e-06, + "loss": 0.0239, + "step": 6251 + }, + { + "epoch": 4.4970329077504045, + "grad_norm": 2.779504953886069, + "learning_rate": 1.7448956995592456e-06, + "loss": 0.0662, + "step": 6252 + }, + { + "epoch": 4.497752202841216, + "grad_norm": 7.201444189975691, + "learning_rate": 1.7445600138202234e-06, + "loss": 0.1361, + "step": 6253 + }, + { + "epoch": 4.498471497932027, + "grad_norm": 2.2225731734420267, + "learning_rate": 1.7442243154981597e-06, + "loss": 0.0516, + "step": 6254 + }, + { + "epoch": 4.499190793022837, + "grad_norm": 0.4920602564603982, + "learning_rate": 1.7438886046103265e-06, + "loss": 0.0031, + "step": 6255 + }, + { + "epoch": 4.499910088113649, + "grad_norm": 2.6233741318024952, + "learning_rate": 1.743552881173996e-06, + "loss": 0.0547, + "step": 6256 + }, + { + "epoch": 4.500629383204459, + "grad_norm": 3.342272510217458, + "learning_rate": 1.7432171452064424e-06, + "loss": 0.1216, + "step": 6257 + }, + { + "epoch": 4.501348678295271, + "grad_norm": 1.8515665773925818, + "learning_rate": 1.7428813967249399e-06, + "loss": 0.006, + "step": 6258 + }, + { + "epoch": 4.502067973386081, + "grad_norm": 1.9994449499240616, + "learning_rate": 1.7425456357467628e-06, + "loss": 0.0391, + "step": 6259 + }, + { + "epoch": 4.502787268476893, + "grad_norm": 0.020298254618424376, + "learning_rate": 1.7422098622891872e-06, + "loss": 0.0001, + "step": 6260 + }, + { + "epoch": 4.503506563567703, + "grad_norm": 2.6801692999876967, + "learning_rate": 1.7418740763694889e-06, + "loss": 0.0724, + "step": 6261 + }, + { + "epoch": 4.504225858658515, + "grad_norm": 0.09144893190564646, + "learning_rate": 1.7415382780049449e-06, + "loss": 0.0003, + "step": 6262 + }, + { + "epoch": 4.5049451537493255, + "grad_norm": 4.489958799865371, + "learning_rate": 1.7412024672128326e-06, + "loss": 0.1244, + "step": 6263 + }, + { + "epoch": 4.505664448840137, + "grad_norm": 1.7606511669504146, + "learning_rate": 1.7408666440104294e-06, + "loss": 0.0183, + "step": 6264 + }, + { + "epoch": 4.5063837439309475, + "grad_norm": 3.2023304289355523, + "learning_rate": 1.740530808415015e-06, + "loss": 0.0694, + "step": 6265 + }, + { + "epoch": 4.507103039021759, + "grad_norm": 3.996239514048809, + "learning_rate": 1.7401949604438681e-06, + "loss": 0.0792, + "step": 6266 + }, + { + "epoch": 4.50782233411257, + "grad_norm": 3.481264298520459, + "learning_rate": 1.7398591001142691e-06, + "loss": 0.0056, + "step": 6267 + }, + { + "epoch": 4.50854162920338, + "grad_norm": 1.5559703729720973, + "learning_rate": 1.7395232274434983e-06, + "loss": 0.0202, + "step": 6268 + }, + { + "epoch": 4.509260924294192, + "grad_norm": 1.3295607553664106, + "learning_rate": 1.7391873424488365e-06, + "loss": 0.0018, + "step": 6269 + }, + { + "epoch": 4.509980219385003, + "grad_norm": 2.798293634095384, + "learning_rate": 1.7388514451475666e-06, + "loss": 0.0579, + "step": 6270 + }, + { + "epoch": 4.510699514475814, + "grad_norm": 2.354096587016648, + "learning_rate": 1.7385155355569702e-06, + "loss": 0.0201, + "step": 6271 + }, + { + "epoch": 4.511418809566624, + "grad_norm": 2.043149089924992, + "learning_rate": 1.738179613694331e-06, + "loss": 0.0034, + "step": 6272 + }, + { + "epoch": 4.512138104657436, + "grad_norm": 1.1103677291734486, + "learning_rate": 1.7378436795769326e-06, + "loss": 0.0103, + "step": 6273 + }, + { + "epoch": 4.512857399748246, + "grad_norm": 2.454534480785061, + "learning_rate": 1.7375077332220593e-06, + "loss": 0.058, + "step": 6274 + }, + { + "epoch": 4.513576694839058, + "grad_norm": 0.2829005851630867, + "learning_rate": 1.7371717746469961e-06, + "loss": 0.0004, + "step": 6275 + }, + { + "epoch": 4.5142959899298685, + "grad_norm": 6.322759404126412, + "learning_rate": 1.736835803869029e-06, + "loss": 0.1479, + "step": 6276 + }, + { + "epoch": 4.51501528502068, + "grad_norm": 3.590388139525511, + "learning_rate": 1.7364998209054435e-06, + "loss": 0.0668, + "step": 6277 + }, + { + "epoch": 4.5157345801114905, + "grad_norm": 2.4543613800836903, + "learning_rate": 1.7361638257735274e-06, + "loss": 0.0477, + "step": 6278 + }, + { + "epoch": 4.516453875202302, + "grad_norm": 4.425665333557274, + "learning_rate": 1.735827818490568e-06, + "loss": 0.0742, + "step": 6279 + }, + { + "epoch": 4.517173170293113, + "grad_norm": 0.8640939869510368, + "learning_rate": 1.7354917990738532e-06, + "loss": 0.0021, + "step": 6280 + }, + { + "epoch": 4.517892465383924, + "grad_norm": 3.1348749669814864, + "learning_rate": 1.7351557675406722e-06, + "loss": 0.0647, + "step": 6281 + }, + { + "epoch": 4.518611760474735, + "grad_norm": 0.13553215264099983, + "learning_rate": 1.7348197239083138e-06, + "loss": 0.0005, + "step": 6282 + }, + { + "epoch": 4.519331055565546, + "grad_norm": 1.7705025103514191, + "learning_rate": 1.7344836681940685e-06, + "loss": 0.0254, + "step": 6283 + }, + { + "epoch": 4.520050350656357, + "grad_norm": 0.8592278828209304, + "learning_rate": 1.7341476004152271e-06, + "loss": 0.002, + "step": 6284 + }, + { + "epoch": 4.520769645747167, + "grad_norm": 4.669750240749038, + "learning_rate": 1.7338115205890804e-06, + "loss": 0.135, + "step": 6285 + }, + { + "epoch": 4.521488940837979, + "grad_norm": 1.897896690643785, + "learning_rate": 1.7334754287329203e-06, + "loss": 0.0053, + "step": 6286 + }, + { + "epoch": 4.522208235928789, + "grad_norm": 2.003314923076362, + "learning_rate": 1.7331393248640398e-06, + "loss": 0.0091, + "step": 6287 + }, + { + "epoch": 4.522927531019601, + "grad_norm": 3.1396319213441326, + "learning_rate": 1.732803208999732e-06, + "loss": 0.0705, + "step": 6288 + }, + { + "epoch": 4.5236468261104115, + "grad_norm": 1.9551884941035242, + "learning_rate": 1.7324670811572903e-06, + "loss": 0.0476, + "step": 6289 + }, + { + "epoch": 4.524366121201223, + "grad_norm": 2.6990759929565535, + "learning_rate": 1.7321309413540088e-06, + "loss": 0.054, + "step": 6290 + }, + { + "epoch": 4.5250854162920335, + "grad_norm": 0.46372399851378665, + "learning_rate": 1.7317947896071832e-06, + "loss": 0.0018, + "step": 6291 + }, + { + "epoch": 4.525804711382845, + "grad_norm": 2.352917751377274, + "learning_rate": 1.7314586259341088e-06, + "loss": 0.0592, + "step": 6292 + }, + { + "epoch": 4.526524006473656, + "grad_norm": 5.597519111830127, + "learning_rate": 1.7311224503520813e-06, + "loss": 0.1602, + "step": 6293 + }, + { + "epoch": 4.527243301564467, + "grad_norm": 1.7530556107263253, + "learning_rate": 1.7307862628783987e-06, + "loss": 0.0247, + "step": 6294 + }, + { + "epoch": 4.527962596655278, + "grad_norm": 0.46623070233967, + "learning_rate": 1.7304500635303572e-06, + "loss": 0.0009, + "step": 6295 + }, + { + "epoch": 4.528681891746089, + "grad_norm": 0.7477840268091221, + "learning_rate": 1.7301138523252558e-06, + "loss": 0.0023, + "step": 6296 + }, + { + "epoch": 4.5294011868369, + "grad_norm": 1.2426079840873712, + "learning_rate": 1.7297776292803927e-06, + "loss": 0.0237, + "step": 6297 + }, + { + "epoch": 4.530120481927711, + "grad_norm": 2.992311808156613, + "learning_rate": 1.729441394413067e-06, + "loss": 0.0285, + "step": 6298 + }, + { + "epoch": 4.530839777018522, + "grad_norm": 2.3258048164216487, + "learning_rate": 1.729105147740579e-06, + "loss": 0.042, + "step": 6299 + }, + { + "epoch": 4.531559072109333, + "grad_norm": 3.769940362262411, + "learning_rate": 1.7287688892802288e-06, + "loss": 0.0915, + "step": 6300 + }, + { + "epoch": 4.532278367200144, + "grad_norm": 2.361305184873462, + "learning_rate": 1.7284326190493175e-06, + "loss": 0.0122, + "step": 6301 + }, + { + "epoch": 4.5329976622909545, + "grad_norm": 4.983858187455307, + "learning_rate": 1.7280963370651472e-06, + "loss": 0.0833, + "step": 6302 + }, + { + "epoch": 4.533716957381766, + "grad_norm": 1.8362457081079446, + "learning_rate": 1.7277600433450198e-06, + "loss": 0.0335, + "step": 6303 + }, + { + "epoch": 4.5344362524725765, + "grad_norm": 0.22129454408246874, + "learning_rate": 1.7274237379062386e-06, + "loss": 0.0009, + "step": 6304 + }, + { + "epoch": 4.535155547563388, + "grad_norm": 3.5562679391776144, + "learning_rate": 1.7270874207661068e-06, + "loss": 0.0446, + "step": 6305 + }, + { + "epoch": 4.535874842654199, + "grad_norm": 0.7399113699900612, + "learning_rate": 1.7267510919419282e-06, + "loss": 0.012, + "step": 6306 + }, + { + "epoch": 4.53659413774501, + "grad_norm": 3.767546073635469, + "learning_rate": 1.7264147514510081e-06, + "loss": 0.1188, + "step": 6307 + }, + { + "epoch": 4.537313432835821, + "grad_norm": 0.08478537934269573, + "learning_rate": 1.7260783993106516e-06, + "loss": 0.0003, + "step": 6308 + }, + { + "epoch": 4.538032727926632, + "grad_norm": 1.6790208374702382, + "learning_rate": 1.7257420355381649e-06, + "loss": 0.0149, + "step": 6309 + }, + { + "epoch": 4.538752023017443, + "grad_norm": 0.008830063980295549, + "learning_rate": 1.7254056601508538e-06, + "loss": 0.0, + "step": 6310 + }, + { + "epoch": 4.539471318108254, + "grad_norm": 0.06791300496938531, + "learning_rate": 1.7250692731660261e-06, + "loss": 0.0002, + "step": 6311 + }, + { + "epoch": 4.540190613199065, + "grad_norm": 3.766978876639846, + "learning_rate": 1.7247328746009887e-06, + "loss": 0.0452, + "step": 6312 + }, + { + "epoch": 4.540909908289876, + "grad_norm": 3.1830551826910085, + "learning_rate": 1.7243964644730508e-06, + "loss": 0.0576, + "step": 6313 + }, + { + "epoch": 4.541629203380687, + "grad_norm": 3.0709005902708784, + "learning_rate": 1.7240600427995206e-06, + "loss": 0.0827, + "step": 6314 + }, + { + "epoch": 4.5423484984714975, + "grad_norm": 1.6672194844908301, + "learning_rate": 1.7237236095977083e-06, + "loss": 0.0398, + "step": 6315 + }, + { + "epoch": 4.543067793562309, + "grad_norm": 1.7174963262669307, + "learning_rate": 1.723387164884923e-06, + "loss": 0.0223, + "step": 6316 + }, + { + "epoch": 4.54378708865312, + "grad_norm": 0.053276710463450914, + "learning_rate": 1.7230507086784763e-06, + "loss": 0.0001, + "step": 6317 + }, + { + "epoch": 4.544506383743931, + "grad_norm": 5.172631194970555, + "learning_rate": 1.7227142409956792e-06, + "loss": 0.0982, + "step": 6318 + }, + { + "epoch": 4.545225678834742, + "grad_norm": 2.214229383697837, + "learning_rate": 1.722377761853843e-06, + "loss": 0.0054, + "step": 6319 + }, + { + "epoch": 4.545944973925553, + "grad_norm": 8.272181521695684, + "learning_rate": 1.7220412712702809e-06, + "loss": 0.1933, + "step": 6320 + }, + { + "epoch": 4.546664269016364, + "grad_norm": 0.33712557526001086, + "learning_rate": 1.7217047692623058e-06, + "loss": 0.0014, + "step": 6321 + }, + { + "epoch": 4.547383564107175, + "grad_norm": 4.151031726093295, + "learning_rate": 1.7213682558472306e-06, + "loss": 0.2269, + "step": 6322 + }, + { + "epoch": 4.548102859197986, + "grad_norm": 4.393701437946786, + "learning_rate": 1.7210317310423704e-06, + "loss": 0.1134, + "step": 6323 + }, + { + "epoch": 4.548822154288797, + "grad_norm": 2.3474640206820068, + "learning_rate": 1.7206951948650394e-06, + "loss": 0.0462, + "step": 6324 + }, + { + "epoch": 4.549541449379608, + "grad_norm": 0.0387312331949139, + "learning_rate": 1.7203586473325535e-06, + "loss": 0.0002, + "step": 6325 + }, + { + "epoch": 4.550260744470419, + "grad_norm": 2.861412390336307, + "learning_rate": 1.7200220884622283e-06, + "loss": 0.0328, + "step": 6326 + }, + { + "epoch": 4.55098003956123, + "grad_norm": 4.8970162216487845, + "learning_rate": 1.7196855182713803e-06, + "loss": 0.1061, + "step": 6327 + }, + { + "epoch": 4.551699334652041, + "grad_norm": 1.9642939039711684, + "learning_rate": 1.7193489367773266e-06, + "loss": 0.03, + "step": 6328 + }, + { + "epoch": 4.552418629742852, + "grad_norm": 0.2964300819935056, + "learning_rate": 1.7190123439973849e-06, + "loss": 0.0005, + "step": 6329 + }, + { + "epoch": 4.553137924833663, + "grad_norm": 2.4633388237818177, + "learning_rate": 1.7186757399488742e-06, + "loss": 0.0656, + "step": 6330 + }, + { + "epoch": 4.553857219924474, + "grad_norm": 2.3834397767857998, + "learning_rate": 1.7183391246491124e-06, + "loss": 0.0105, + "step": 6331 + }, + { + "epoch": 4.554576515015285, + "grad_norm": 1.8674646683277059, + "learning_rate": 1.71800249811542e-06, + "loss": 0.0439, + "step": 6332 + }, + { + "epoch": 4.555295810106096, + "grad_norm": 2.4390928001852332, + "learning_rate": 1.7176658603651156e-06, + "loss": 0.017, + "step": 6333 + }, + { + "epoch": 4.556015105196907, + "grad_norm": 0.6064509053845807, + "learning_rate": 1.7173292114155212e-06, + "loss": 0.0014, + "step": 6334 + }, + { + "epoch": 4.556734400287718, + "grad_norm": 3.1415079408243756, + "learning_rate": 1.7169925512839568e-06, + "loss": 0.0418, + "step": 6335 + }, + { + "epoch": 4.557453695378529, + "grad_norm": 1.9189781918809548, + "learning_rate": 1.716655879987745e-06, + "loss": 0.0038, + "step": 6336 + }, + { + "epoch": 4.55817299046934, + "grad_norm": 2.9866557972994605, + "learning_rate": 1.7163191975442075e-06, + "loss": 0.0091, + "step": 6337 + }, + { + "epoch": 4.558892285560151, + "grad_norm": 0.022386245268241556, + "learning_rate": 1.715982503970668e-06, + "loss": 0.0001, + "step": 6338 + }, + { + "epoch": 4.559611580650962, + "grad_norm": 1.376249801563846, + "learning_rate": 1.7156457992844493e-06, + "loss": 0.0032, + "step": 6339 + }, + { + "epoch": 4.560330875741773, + "grad_norm": 5.784986880650363, + "learning_rate": 1.7153090835028758e-06, + "loss": 0.105, + "step": 6340 + }, + { + "epoch": 4.561050170832584, + "grad_norm": 1.8122427038930298, + "learning_rate": 1.714972356643272e-06, + "loss": 0.0388, + "step": 6341 + }, + { + "epoch": 4.561769465923395, + "grad_norm": 2.8581417205379576, + "learning_rate": 1.7146356187229628e-06, + "loss": 0.0336, + "step": 6342 + }, + { + "epoch": 4.562488761014206, + "grad_norm": 1.7262910159942757, + "learning_rate": 1.7142988697592745e-06, + "loss": 0.0224, + "step": 6343 + }, + { + "epoch": 4.563208056105017, + "grad_norm": 5.998667644936431, + "learning_rate": 1.7139621097695334e-06, + "loss": 0.2079, + "step": 6344 + }, + { + "epoch": 4.563927351195828, + "grad_norm": 3.9369683409346785, + "learning_rate": 1.7136253387710655e-06, + "loss": 0.1094, + "step": 6345 + }, + { + "epoch": 4.564646646286639, + "grad_norm": 3.0533624665728647, + "learning_rate": 1.7132885567811996e-06, + "loss": 0.066, + "step": 6346 + }, + { + "epoch": 4.5653659413774506, + "grad_norm": 1.3732213120533545, + "learning_rate": 1.712951763817263e-06, + "loss": 0.0209, + "step": 6347 + }, + { + "epoch": 4.566085236468261, + "grad_norm": 5.599851480597732, + "learning_rate": 1.7126149598965834e-06, + "loss": 0.2591, + "step": 6348 + }, + { + "epoch": 4.566804531559072, + "grad_norm": 3.0860219889584513, + "learning_rate": 1.7122781450364918e-06, + "loss": 0.023, + "step": 6349 + }, + { + "epoch": 4.567523826649883, + "grad_norm": 1.4562176547768193, + "learning_rate": 1.7119413192543165e-06, + "loss": 0.0024, + "step": 6350 + }, + { + "epoch": 4.568243121740694, + "grad_norm": 2.462369726708238, + "learning_rate": 1.7116044825673886e-06, + "loss": 0.0393, + "step": 6351 + }, + { + "epoch": 4.568962416831505, + "grad_norm": 0.06742006554016401, + "learning_rate": 1.7112676349930387e-06, + "loss": 0.0003, + "step": 6352 + }, + { + "epoch": 4.569681711922316, + "grad_norm": 1.8614849894392347, + "learning_rate": 1.7109307765485977e-06, + "loss": 0.0346, + "step": 6353 + }, + { + "epoch": 4.570401007013127, + "grad_norm": 1.0384566923366498, + "learning_rate": 1.7105939072513985e-06, + "loss": 0.0024, + "step": 6354 + }, + { + "epoch": 4.571120302103938, + "grad_norm": 2.243819309112382, + "learning_rate": 1.7102570271187727e-06, + "loss": 0.0111, + "step": 6355 + }, + { + "epoch": 4.571839597194749, + "grad_norm": 4.2600105353399425, + "learning_rate": 1.7099201361680536e-06, + "loss": 0.0444, + "step": 6356 + }, + { + "epoch": 4.57255889228556, + "grad_norm": 0.4656614056035723, + "learning_rate": 1.7095832344165753e-06, + "loss": 0.0013, + "step": 6357 + }, + { + "epoch": 4.5732781873763715, + "grad_norm": 2.1488087192452716, + "learning_rate": 1.7092463218816714e-06, + "loss": 0.0102, + "step": 6358 + }, + { + "epoch": 4.573997482467182, + "grad_norm": 2.9257876146427884, + "learning_rate": 1.7089093985806771e-06, + "loss": 0.0358, + "step": 6359 + }, + { + "epoch": 4.5747167775579936, + "grad_norm": 4.041435327159601, + "learning_rate": 1.7085724645309276e-06, + "loss": 0.0755, + "step": 6360 + }, + { + "epoch": 4.575436072648804, + "grad_norm": 4.1431577379361775, + "learning_rate": 1.7082355197497588e-06, + "loss": 0.091, + "step": 6361 + }, + { + "epoch": 4.576155367739615, + "grad_norm": 2.463844120697995, + "learning_rate": 1.7078985642545066e-06, + "loss": 0.072, + "step": 6362 + }, + { + "epoch": 4.576874662830426, + "grad_norm": 2.146272023670015, + "learning_rate": 1.7075615980625085e-06, + "loss": 0.0545, + "step": 6363 + }, + { + "epoch": 4.577593957921237, + "grad_norm": 5.07723610723675, + "learning_rate": 1.7072246211911018e-06, + "loss": 0.1121, + "step": 6364 + }, + { + "epoch": 4.578313253012048, + "grad_norm": 2.365544655568069, + "learning_rate": 1.7068876336576244e-06, + "loss": 0.0412, + "step": 6365 + }, + { + "epoch": 4.579032548102859, + "grad_norm": 2.9663042232499772, + "learning_rate": 1.7065506354794154e-06, + "loss": 0.0435, + "step": 6366 + }, + { + "epoch": 4.57975184319367, + "grad_norm": 4.935301238756384, + "learning_rate": 1.7062136266738134e-06, + "loss": 0.1569, + "step": 6367 + }, + { + "epoch": 4.580471138284481, + "grad_norm": 5.103291951279044, + "learning_rate": 1.7058766072581585e-06, + "loss": 0.1108, + "step": 6368 + }, + { + "epoch": 4.581190433375292, + "grad_norm": 0.19662149583934405, + "learning_rate": 1.7055395772497906e-06, + "loss": 0.0003, + "step": 6369 + }, + { + "epoch": 4.581909728466103, + "grad_norm": 1.5288590452129167, + "learning_rate": 1.7052025366660507e-06, + "loss": 0.02, + "step": 6370 + }, + { + "epoch": 4.5826290235569145, + "grad_norm": 2.426125993558333, + "learning_rate": 1.7048654855242798e-06, + "loss": 0.0352, + "step": 6371 + }, + { + "epoch": 4.583348318647725, + "grad_norm": 4.484020108547524, + "learning_rate": 1.7045284238418203e-06, + "loss": 0.1316, + "step": 6372 + }, + { + "epoch": 4.5840676137385366, + "grad_norm": 4.24065320300311, + "learning_rate": 1.7041913516360142e-06, + "loss": 0.0979, + "step": 6373 + }, + { + "epoch": 4.584786908829347, + "grad_norm": 2.212847675868232, + "learning_rate": 1.7038542689242047e-06, + "loss": 0.032, + "step": 6374 + }, + { + "epoch": 4.585506203920159, + "grad_norm": 2.098119058118541, + "learning_rate": 1.7035171757237353e-06, + "loss": 0.0304, + "step": 6375 + }, + { + "epoch": 4.586225499010969, + "grad_norm": 3.768596442791888, + "learning_rate": 1.7031800720519502e-06, + "loss": 0.1104, + "step": 6376 + }, + { + "epoch": 4.586944794101781, + "grad_norm": 1.2685023350793818, + "learning_rate": 1.7028429579261934e-06, + "loss": 0.0207, + "step": 6377 + }, + { + "epoch": 4.587664089192591, + "grad_norm": 3.9240481194950716, + "learning_rate": 1.70250583336381e-06, + "loss": 0.0568, + "step": 6378 + }, + { + "epoch": 4.588383384283402, + "grad_norm": 2.2059420320946312, + "learning_rate": 1.702168698382146e-06, + "loss": 0.0535, + "step": 6379 + }, + { + "epoch": 4.589102679374213, + "grad_norm": 4.809980764927814, + "learning_rate": 1.7018315529985481e-06, + "loss": 0.1125, + "step": 6380 + }, + { + "epoch": 4.589821974465024, + "grad_norm": 4.9475180296983305, + "learning_rate": 1.7014943972303623e-06, + "loss": 0.1178, + "step": 6381 + }, + { + "epoch": 4.590541269555835, + "grad_norm": 1.1630241117162583, + "learning_rate": 1.7011572310949364e-06, + "loss": 0.0167, + "step": 6382 + }, + { + "epoch": 4.591260564646646, + "grad_norm": 3.721232716208791, + "learning_rate": 1.7008200546096174e-06, + "loss": 0.0865, + "step": 6383 + }, + { + "epoch": 4.5919798597374575, + "grad_norm": 0.12387535787338994, + "learning_rate": 1.700482867791754e-06, + "loss": 0.0003, + "step": 6384 + }, + { + "epoch": 4.592699154828268, + "grad_norm": 2.245359754391687, + "learning_rate": 1.7001456706586953e-06, + "loss": 0.0544, + "step": 6385 + }, + { + "epoch": 4.5934184499190795, + "grad_norm": 1.6304856707540605, + "learning_rate": 1.6998084632277904e-06, + "loss": 0.0542, + "step": 6386 + }, + { + "epoch": 4.59413774500989, + "grad_norm": 2.9737710290565555, + "learning_rate": 1.6994712455163892e-06, + "loss": 0.1096, + "step": 6387 + }, + { + "epoch": 4.594857040100702, + "grad_norm": 8.003670517969848, + "learning_rate": 1.6991340175418425e-06, + "loss": 0.2861, + "step": 6388 + }, + { + "epoch": 4.595576335191512, + "grad_norm": 1.7359559423111652, + "learning_rate": 1.6987967793215009e-06, + "loss": 0.0308, + "step": 6389 + }, + { + "epoch": 4.596295630282324, + "grad_norm": 0.08070810472962923, + "learning_rate": 1.6984595308727161e-06, + "loss": 0.0002, + "step": 6390 + }, + { + "epoch": 4.597014925373134, + "grad_norm": 2.61916831758499, + "learning_rate": 1.6981222722128403e-06, + "loss": 0.0647, + "step": 6391 + }, + { + "epoch": 4.597734220463945, + "grad_norm": 0.5110886318644844, + "learning_rate": 1.697785003359225e-06, + "loss": 0.0046, + "step": 6392 + }, + { + "epoch": 4.598453515554756, + "grad_norm": 1.0469197103578936, + "learning_rate": 1.6974477243292247e-06, + "loss": 0.003, + "step": 6393 + }, + { + "epoch": 4.599172810645568, + "grad_norm": 2.282684557681592, + "learning_rate": 1.6971104351401925e-06, + "loss": 0.0613, + "step": 6394 + }, + { + "epoch": 4.599892105736378, + "grad_norm": 1.8087255715410981, + "learning_rate": 1.696773135809482e-06, + "loss": 0.0213, + "step": 6395 + }, + { + "epoch": 4.600611400827189, + "grad_norm": 1.4752868480831236, + "learning_rate": 1.6964358263544485e-06, + "loss": 0.0144, + "step": 6396 + }, + { + "epoch": 4.6013306959180005, + "grad_norm": 2.4224250168983943, + "learning_rate": 1.6960985067924468e-06, + "loss": 0.0492, + "step": 6397 + }, + { + "epoch": 4.602049991008811, + "grad_norm": 3.6046069352609607, + "learning_rate": 1.6957611771408326e-06, + "loss": 0.0444, + "step": 6398 + }, + { + "epoch": 4.6027692860996225, + "grad_norm": 3.950578262960119, + "learning_rate": 1.6954238374169624e-06, + "loss": 0.1697, + "step": 6399 + }, + { + "epoch": 4.603488581190433, + "grad_norm": 0.5331100319792457, + "learning_rate": 1.6950864876381924e-06, + "loss": 0.0023, + "step": 6400 + }, + { + "epoch": 4.604207876281245, + "grad_norm": 3.6777520689301273, + "learning_rate": 1.6947491278218806e-06, + "loss": 0.1252, + "step": 6401 + }, + { + "epoch": 4.604927171372055, + "grad_norm": 5.968393024760236, + "learning_rate": 1.694411757985384e-06, + "loss": 0.0982, + "step": 6402 + }, + { + "epoch": 4.605646466462867, + "grad_norm": 3.845041795984026, + "learning_rate": 1.6940743781460611e-06, + "loss": 0.0785, + "step": 6403 + }, + { + "epoch": 4.606365761553677, + "grad_norm": 3.9016796747578573, + "learning_rate": 1.6937369883212708e-06, + "loss": 0.0826, + "step": 6404 + }, + { + "epoch": 4.607085056644489, + "grad_norm": 8.886737238464391, + "learning_rate": 1.6933995885283725e-06, + "loss": 0.201, + "step": 6405 + }, + { + "epoch": 4.607804351735299, + "grad_norm": 0.5825870432164348, + "learning_rate": 1.6930621787847258e-06, + "loss": 0.0012, + "step": 6406 + }, + { + "epoch": 4.608523646826111, + "grad_norm": 6.585919430155223, + "learning_rate": 1.6927247591076912e-06, + "loss": 0.2221, + "step": 6407 + }, + { + "epoch": 4.609242941916921, + "grad_norm": 17.96644791973082, + "learning_rate": 1.6923873295146292e-06, + "loss": 0.3756, + "step": 6408 + }, + { + "epoch": 4.609962237007732, + "grad_norm": 0.009522622247460524, + "learning_rate": 1.6920498900229014e-06, + "loss": 0.0, + "step": 6409 + }, + { + "epoch": 4.6106815320985435, + "grad_norm": 1.722079530305576, + "learning_rate": 1.6917124406498698e-06, + "loss": 0.0425, + "step": 6410 + }, + { + "epoch": 4.611400827189354, + "grad_norm": 2.4941829306120713, + "learning_rate": 1.691374981412897e-06, + "loss": 0.0845, + "step": 6411 + }, + { + "epoch": 4.6121201222801655, + "grad_norm": 2.9042204463326033, + "learning_rate": 1.6910375123293447e-06, + "loss": 0.0418, + "step": 6412 + }, + { + "epoch": 4.612839417370976, + "grad_norm": 2.9564907903639517, + "learning_rate": 1.6907000334165771e-06, + "loss": 0.0922, + "step": 6413 + }, + { + "epoch": 4.613558712461788, + "grad_norm": 3.5708927006993774, + "learning_rate": 1.6903625446919582e-06, + "loss": 0.1703, + "step": 6414 + }, + { + "epoch": 4.614278007552598, + "grad_norm": 1.4323914213368578, + "learning_rate": 1.6900250461728525e-06, + "loss": 0.0187, + "step": 6415 + }, + { + "epoch": 4.61499730264341, + "grad_norm": 1.8150288710823208, + "learning_rate": 1.689687537876624e-06, + "loss": 0.038, + "step": 6416 + }, + { + "epoch": 4.61571659773422, + "grad_norm": 3.0967128928391614, + "learning_rate": 1.689350019820639e-06, + "loss": 0.0839, + "step": 6417 + }, + { + "epoch": 4.616435892825032, + "grad_norm": 2.523301771092086, + "learning_rate": 1.6890124920222633e-06, + "loss": 0.0862, + "step": 6418 + }, + { + "epoch": 4.617155187915842, + "grad_norm": 0.6899907454967285, + "learning_rate": 1.6886749544988628e-06, + "loss": 0.0127, + "step": 6419 + }, + { + "epoch": 4.617874483006654, + "grad_norm": 4.568476689690237, + "learning_rate": 1.688337407267805e-06, + "loss": 0.0713, + "step": 6420 + }, + { + "epoch": 4.618593778097464, + "grad_norm": 2.0841371591235087, + "learning_rate": 1.6879998503464564e-06, + "loss": 0.0384, + "step": 6421 + }, + { + "epoch": 4.619313073188275, + "grad_norm": 1.956080555368812, + "learning_rate": 1.6876622837521858e-06, + "loss": 0.0366, + "step": 6422 + }, + { + "epoch": 4.6200323682790865, + "grad_norm": 5.88090715093223, + "learning_rate": 1.6873247075023616e-06, + "loss": 0.1499, + "step": 6423 + }, + { + "epoch": 4.620751663369898, + "grad_norm": 22.43145861221353, + "learning_rate": 1.6869871216143516e-06, + "loss": 0.0547, + "step": 6424 + }, + { + "epoch": 4.6214709584607085, + "grad_norm": 2.7972962177955134, + "learning_rate": 1.6866495261055265e-06, + "loss": 0.0756, + "step": 6425 + }, + { + "epoch": 4.622190253551519, + "grad_norm": 4.019345237603542, + "learning_rate": 1.6863119209932557e-06, + "loss": 0.1203, + "step": 6426 + }, + { + "epoch": 4.622909548642331, + "grad_norm": 0.028391151817713137, + "learning_rate": 1.6859743062949093e-06, + "loss": 0.0002, + "step": 6427 + }, + { + "epoch": 4.623628843733141, + "grad_norm": 2.314254711288784, + "learning_rate": 1.6856366820278586e-06, + "loss": 0.0701, + "step": 6428 + }, + { + "epoch": 4.624348138823953, + "grad_norm": 0.27805588618505833, + "learning_rate": 1.6852990482094741e-06, + "loss": 0.0005, + "step": 6429 + }, + { + "epoch": 4.625067433914763, + "grad_norm": 2.886781115257202, + "learning_rate": 1.684961404857129e-06, + "loss": 0.0319, + "step": 6430 + }, + { + "epoch": 4.625786729005575, + "grad_norm": 2.3753506609565704, + "learning_rate": 1.6846237519881946e-06, + "loss": 0.0266, + "step": 6431 + }, + { + "epoch": 4.626506024096385, + "grad_norm": 0.0497670858888453, + "learning_rate": 1.6842860896200437e-06, + "loss": 0.0003, + "step": 6432 + }, + { + "epoch": 4.627225319187197, + "grad_norm": 0.09754745153720625, + "learning_rate": 1.6839484177700506e-06, + "loss": 0.0004, + "step": 6433 + }, + { + "epoch": 4.627944614278007, + "grad_norm": 0.6861238855778203, + "learning_rate": 1.6836107364555884e-06, + "loss": 0.0058, + "step": 6434 + }, + { + "epoch": 4.628663909368819, + "grad_norm": 2.3827781568444837, + "learning_rate": 1.6832730456940317e-06, + "loss": 0.0308, + "step": 6435 + }, + { + "epoch": 4.6293832044596295, + "grad_norm": 1.6155758095085693, + "learning_rate": 1.682935345502755e-06, + "loss": 0.0135, + "step": 6436 + }, + { + "epoch": 4.630102499550441, + "grad_norm": 1.7503332796108342, + "learning_rate": 1.6825976358991329e-06, + "loss": 0.0435, + "step": 6437 + }, + { + "epoch": 4.6308217946412515, + "grad_norm": 3.2527724425857625, + "learning_rate": 1.6822599169005424e-06, + "loss": 0.0412, + "step": 6438 + }, + { + "epoch": 4.631541089732062, + "grad_norm": 3.9919858673472977, + "learning_rate": 1.6819221885243589e-06, + "loss": 0.0899, + "step": 6439 + }, + { + "epoch": 4.632260384822874, + "grad_norm": 3.312716381057069, + "learning_rate": 1.68158445078796e-06, + "loss": 0.0731, + "step": 6440 + }, + { + "epoch": 4.632979679913684, + "grad_norm": 3.705600037136754, + "learning_rate": 1.6812467037087224e-06, + "loss": 0.1195, + "step": 6441 + }, + { + "epoch": 4.633698975004496, + "grad_norm": 1.800907115735062, + "learning_rate": 1.680908947304023e-06, + "loss": 0.0507, + "step": 6442 + }, + { + "epoch": 4.634418270095306, + "grad_norm": 2.891179896416794, + "learning_rate": 1.6805711815912415e-06, + "loss": 0.0799, + "step": 6443 + }, + { + "epoch": 4.635137565186118, + "grad_norm": 3.6030594460549397, + "learning_rate": 1.6802334065877552e-06, + "loss": 0.0839, + "step": 6444 + }, + { + "epoch": 4.635856860276928, + "grad_norm": 1.2405927257335982, + "learning_rate": 1.6798956223109439e-06, + "loss": 0.0184, + "step": 6445 + }, + { + "epoch": 4.63657615536774, + "grad_norm": 0.17917492380205313, + "learning_rate": 1.679557828778187e-06, + "loss": 0.0009, + "step": 6446 + }, + { + "epoch": 4.63729545045855, + "grad_norm": 1.2980576579826169, + "learning_rate": 1.6792200260068647e-06, + "loss": 0.0016, + "step": 6447 + }, + { + "epoch": 4.638014745549362, + "grad_norm": 0.07824030662322237, + "learning_rate": 1.6788822140143574e-06, + "loss": 0.0001, + "step": 6448 + }, + { + "epoch": 4.6387340406401725, + "grad_norm": 7.050029608024736, + "learning_rate": 1.6785443928180462e-06, + "loss": 0.2155, + "step": 6449 + }, + { + "epoch": 4.639453335730984, + "grad_norm": 0.6252965650211079, + "learning_rate": 1.6782065624353126e-06, + "loss": 0.0103, + "step": 6450 + }, + { + "epoch": 4.6401726308217945, + "grad_norm": 3.2545456866227633, + "learning_rate": 1.6778687228835383e-06, + "loss": 0.0417, + "step": 6451 + }, + { + "epoch": 4.640891925912606, + "grad_norm": 0.03911162633330471, + "learning_rate": 1.6775308741801063e-06, + "loss": 0.0002, + "step": 6452 + }, + { + "epoch": 4.641611221003417, + "grad_norm": 2.702309196644279, + "learning_rate": 1.6771930163423985e-06, + "loss": 0.0387, + "step": 6453 + }, + { + "epoch": 4.642330516094228, + "grad_norm": 2.2383014453706584, + "learning_rate": 1.6768551493877997e-06, + "loss": 0.0468, + "step": 6454 + }, + { + "epoch": 4.643049811185039, + "grad_norm": 2.286917024894287, + "learning_rate": 1.676517273333693e-06, + "loss": 0.0362, + "step": 6455 + }, + { + "epoch": 4.643769106275849, + "grad_norm": 2.9243900927549498, + "learning_rate": 1.6761793881974626e-06, + "loss": 0.0627, + "step": 6456 + }, + { + "epoch": 4.644488401366661, + "grad_norm": 1.9230894328439392, + "learning_rate": 1.6758414939964931e-06, + "loss": 0.0322, + "step": 6457 + }, + { + "epoch": 4.645207696457471, + "grad_norm": 3.2981680183706, + "learning_rate": 1.6755035907481702e-06, + "loss": 0.0671, + "step": 6458 + }, + { + "epoch": 4.645926991548283, + "grad_norm": 0.0662817001218068, + "learning_rate": 1.6751656784698795e-06, + "loss": 0.0001, + "step": 6459 + }, + { + "epoch": 4.646646286639093, + "grad_norm": 1.8865292398822195, + "learning_rate": 1.6748277571790067e-06, + "loss": 0.0287, + "step": 6460 + }, + { + "epoch": 4.647365581729905, + "grad_norm": 2.250393981036212, + "learning_rate": 1.6744898268929398e-06, + "loss": 0.0128, + "step": 6461 + }, + { + "epoch": 4.6480848768207155, + "grad_norm": 1.6599113455662595, + "learning_rate": 1.6741518876290643e-06, + "loss": 0.0091, + "step": 6462 + }, + { + "epoch": 4.648804171911527, + "grad_norm": 2.1654205263575355, + "learning_rate": 1.673813939404769e-06, + "loss": 0.0354, + "step": 6463 + }, + { + "epoch": 4.6495234670023375, + "grad_norm": 4.127316288415784, + "learning_rate": 1.6734759822374412e-06, + "loss": 0.0149, + "step": 6464 + }, + { + "epoch": 4.650242762093149, + "grad_norm": 5.077644107662487, + "learning_rate": 1.6731380161444695e-06, + "loss": 0.1195, + "step": 6465 + }, + { + "epoch": 4.65096205718396, + "grad_norm": 4.509468941259139, + "learning_rate": 1.6728000411432427e-06, + "loss": 0.1199, + "step": 6466 + }, + { + "epoch": 4.651681352274771, + "grad_norm": 4.746631117859775, + "learning_rate": 1.672462057251151e-06, + "loss": 0.1781, + "step": 6467 + }, + { + "epoch": 4.652400647365582, + "grad_norm": 3.9290437333955364, + "learning_rate": 1.672124064485583e-06, + "loss": 0.0843, + "step": 6468 + }, + { + "epoch": 4.653119942456392, + "grad_norm": 2.305976696167017, + "learning_rate": 1.6717860628639299e-06, + "loss": 0.0371, + "step": 6469 + }, + { + "epoch": 4.653839237547204, + "grad_norm": 6.757965650852733, + "learning_rate": 1.671448052403583e-06, + "loss": 0.1101, + "step": 6470 + }, + { + "epoch": 4.654558532638015, + "grad_norm": 3.454404935754181, + "learning_rate": 1.6711100331219321e-06, + "loss": 0.0746, + "step": 6471 + }, + { + "epoch": 4.655277827728826, + "grad_norm": 3.3215185777885083, + "learning_rate": 1.67077200503637e-06, + "loss": 0.1126, + "step": 6472 + }, + { + "epoch": 4.655997122819636, + "grad_norm": 2.5547002626729176, + "learning_rate": 1.6704339681642883e-06, + "loss": 0.0371, + "step": 6473 + }, + { + "epoch": 4.656716417910448, + "grad_norm": 2.817536958545492, + "learning_rate": 1.6700959225230796e-06, + "loss": 0.0662, + "step": 6474 + }, + { + "epoch": 4.6574357130012585, + "grad_norm": 0.47017069874844103, + "learning_rate": 1.6697578681301374e-06, + "loss": 0.002, + "step": 6475 + }, + { + "epoch": 4.65815500809207, + "grad_norm": 2.1716361480339526, + "learning_rate": 1.6694198050028543e-06, + "loss": 0.0438, + "step": 6476 + }, + { + "epoch": 4.6588743031828805, + "grad_norm": 2.6131740373834584, + "learning_rate": 1.669081733158626e-06, + "loss": 0.0431, + "step": 6477 + }, + { + "epoch": 4.659593598273692, + "grad_norm": 0.7055456596764852, + "learning_rate": 1.6687436526148449e-06, + "loss": 0.0037, + "step": 6478 + }, + { + "epoch": 4.660312893364503, + "grad_norm": 2.0081073293487517, + "learning_rate": 1.6684055633889066e-06, + "loss": 0.0455, + "step": 6479 + }, + { + "epoch": 4.661032188455314, + "grad_norm": 3.2705195171895824, + "learning_rate": 1.6680674654982066e-06, + "loss": 0.057, + "step": 6480 + }, + { + "epoch": 4.661751483546125, + "grad_norm": 1.2434661312460495, + "learning_rate": 1.6677293589601401e-06, + "loss": 0.04, + "step": 6481 + }, + { + "epoch": 4.662470778636936, + "grad_norm": 2.737435544940386, + "learning_rate": 1.6673912437921041e-06, + "loss": 0.028, + "step": 6482 + }, + { + "epoch": 4.663190073727747, + "grad_norm": 2.301653241346803, + "learning_rate": 1.6670531200114947e-06, + "loss": 0.05, + "step": 6483 + }, + { + "epoch": 4.663909368818558, + "grad_norm": 1.8949843586470898, + "learning_rate": 1.6667149876357091e-06, + "loss": 0.0409, + "step": 6484 + }, + { + "epoch": 4.664628663909369, + "grad_norm": 0.8129505296085334, + "learning_rate": 1.6663768466821448e-06, + "loss": 0.0072, + "step": 6485 + }, + { + "epoch": 4.665347959000179, + "grad_norm": 1.9352495398721783, + "learning_rate": 1.6660386971681997e-06, + "loss": 0.0542, + "step": 6486 + }, + { + "epoch": 4.666067254090991, + "grad_norm": 4.969561645298901, + "learning_rate": 1.6657005391112717e-06, + "loss": 0.0539, + "step": 6487 + }, + { + "epoch": 4.6667865491818015, + "grad_norm": 2.2425355333669494, + "learning_rate": 1.6653623725287604e-06, + "loss": 0.0349, + "step": 6488 + }, + { + "epoch": 4.667505844272613, + "grad_norm": 1.5689687793180613, + "learning_rate": 1.6650241974380644e-06, + "loss": 0.0309, + "step": 6489 + }, + { + "epoch": 4.6682251393634235, + "grad_norm": 3.5180048019088774, + "learning_rate": 1.6646860138565842e-06, + "loss": 0.0443, + "step": 6490 + }, + { + "epoch": 4.668944434454235, + "grad_norm": 3.779573379609225, + "learning_rate": 1.6643478218017193e-06, + "loss": 0.0145, + "step": 6491 + }, + { + "epoch": 4.669663729545046, + "grad_norm": 0.5328943600730488, + "learning_rate": 1.6640096212908706e-06, + "loss": 0.0022, + "step": 6492 + }, + { + "epoch": 4.670383024635857, + "grad_norm": 1.5349199844301735, + "learning_rate": 1.6636714123414389e-06, + "loss": 0.0035, + "step": 6493 + }, + { + "epoch": 4.671102319726668, + "grad_norm": 1.3560439919225464, + "learning_rate": 1.663333194970826e-06, + "loss": 0.0133, + "step": 6494 + }, + { + "epoch": 4.671821614817479, + "grad_norm": 2.557951105273391, + "learning_rate": 1.6629949691964327e-06, + "loss": 0.0624, + "step": 6495 + }, + { + "epoch": 4.67254090990829, + "grad_norm": 2.0388377566642055, + "learning_rate": 1.6626567350356627e-06, + "loss": 0.0477, + "step": 6496 + }, + { + "epoch": 4.673260204999101, + "grad_norm": 4.68744958029039, + "learning_rate": 1.6623184925059176e-06, + "loss": 0.2017, + "step": 6497 + }, + { + "epoch": 4.673979500089912, + "grad_norm": 3.7839192141679914, + "learning_rate": 1.6619802416246018e-06, + "loss": 0.0069, + "step": 6498 + }, + { + "epoch": 4.674698795180722, + "grad_norm": 4.900317358229887, + "learning_rate": 1.6616419824091181e-06, + "loss": 0.1108, + "step": 6499 + }, + { + "epoch": 4.675418090271534, + "grad_norm": 0.3917938797390461, + "learning_rate": 1.6613037148768703e-06, + "loss": 0.001, + "step": 6500 + }, + { + "epoch": 4.676137385362345, + "grad_norm": 4.080609772824777, + "learning_rate": 1.6609654390452635e-06, + "loss": 0.0726, + "step": 6501 + }, + { + "epoch": 4.676856680453156, + "grad_norm": 2.4713889198163117, + "learning_rate": 1.6606271549317022e-06, + "loss": 0.06, + "step": 6502 + }, + { + "epoch": 4.6775759755439665, + "grad_norm": 3.376885830901478, + "learning_rate": 1.660288862553592e-06, + "loss": 0.0781, + "step": 6503 + }, + { + "epoch": 4.678295270634778, + "grad_norm": 2.0433016228276624, + "learning_rate": 1.6599505619283383e-06, + "loss": 0.0506, + "step": 6504 + }, + { + "epoch": 4.679014565725589, + "grad_norm": 1.0117691944742784, + "learning_rate": 1.6596122530733472e-06, + "loss": 0.0106, + "step": 6505 + }, + { + "epoch": 4.6797338608164, + "grad_norm": 1.6355895766047686, + "learning_rate": 1.6592739360060256e-06, + "loss": 0.0168, + "step": 6506 + }, + { + "epoch": 4.680453155907211, + "grad_norm": 2.2407489112095242, + "learning_rate": 1.6589356107437806e-06, + "loss": 0.0596, + "step": 6507 + }, + { + "epoch": 4.681172450998022, + "grad_norm": 2.4493652803613783, + "learning_rate": 1.6585972773040192e-06, + "loss": 0.0483, + "step": 6508 + }, + { + "epoch": 4.681891746088833, + "grad_norm": 5.0042279190238625, + "learning_rate": 1.6582589357041496e-06, + "loss": 0.0526, + "step": 6509 + }, + { + "epoch": 4.682611041179644, + "grad_norm": 0.3236709597739673, + "learning_rate": 1.6579205859615796e-06, + "loss": 0.0006, + "step": 6510 + }, + { + "epoch": 4.683330336270455, + "grad_norm": 5.094253162878108, + "learning_rate": 1.6575822280937188e-06, + "loss": 0.1749, + "step": 6511 + }, + { + "epoch": 4.684049631361266, + "grad_norm": 0.4114322956620578, + "learning_rate": 1.6572438621179755e-06, + "loss": 0.0007, + "step": 6512 + }, + { + "epoch": 4.684768926452077, + "grad_norm": 1.4623650056214867, + "learning_rate": 1.6569054880517594e-06, + "loss": 0.0145, + "step": 6513 + }, + { + "epoch": 4.685488221542888, + "grad_norm": 4.1619088885070505, + "learning_rate": 1.6565671059124806e-06, + "loss": 0.0687, + "step": 6514 + }, + { + "epoch": 4.686207516633699, + "grad_norm": 4.4345606118117935, + "learning_rate": 1.6562287157175493e-06, + "loss": 0.0418, + "step": 6515 + }, + { + "epoch": 4.6869268117245095, + "grad_norm": 3.020195460869355, + "learning_rate": 1.6558903174843762e-06, + "loss": 0.0811, + "step": 6516 + }, + { + "epoch": 4.687646106815321, + "grad_norm": 5.101888375478918, + "learning_rate": 1.6555519112303729e-06, + "loss": 0.0451, + "step": 6517 + }, + { + "epoch": 4.6883654019061325, + "grad_norm": 2.5684520292936486, + "learning_rate": 1.6552134969729502e-06, + "loss": 0.0468, + "step": 6518 + }, + { + "epoch": 4.689084696996943, + "grad_norm": 1.0078344726564128, + "learning_rate": 1.6548750747295211e-06, + "loss": 0.0148, + "step": 6519 + }, + { + "epoch": 4.689803992087754, + "grad_norm": 2.1480243306677154, + "learning_rate": 1.6545366445174975e-06, + "loss": 0.0255, + "step": 6520 + }, + { + "epoch": 4.690523287178565, + "grad_norm": 1.5647473952593058, + "learning_rate": 1.654198206354292e-06, + "loss": 0.0032, + "step": 6521 + }, + { + "epoch": 4.691242582269376, + "grad_norm": 4.399672241350049, + "learning_rate": 1.6538597602573185e-06, + "loss": 0.0854, + "step": 6522 + }, + { + "epoch": 4.691961877360187, + "grad_norm": 2.209220960246672, + "learning_rate": 1.6535213062439897e-06, + "loss": 0.0348, + "step": 6523 + }, + { + "epoch": 4.692681172450998, + "grad_norm": 7.140097594830647, + "learning_rate": 1.6531828443317205e-06, + "loss": 0.2117, + "step": 6524 + }, + { + "epoch": 4.693400467541809, + "grad_norm": 3.606408054918426, + "learning_rate": 1.6528443745379251e-06, + "loss": 0.0644, + "step": 6525 + }, + { + "epoch": 4.69411976263262, + "grad_norm": 3.1735751340369873, + "learning_rate": 1.652505896880018e-06, + "loss": 0.0436, + "step": 6526 + }, + { + "epoch": 4.694839057723431, + "grad_norm": 0.198592663023344, + "learning_rate": 1.652167411375415e-06, + "loss": 0.0003, + "step": 6527 + }, + { + "epoch": 4.695558352814242, + "grad_norm": 3.971547734780975, + "learning_rate": 1.6518289180415316e-06, + "loss": 0.064, + "step": 6528 + }, + { + "epoch": 4.696277647905053, + "grad_norm": 3.708014511899447, + "learning_rate": 1.651490416895784e-06, + "loss": 0.0718, + "step": 6529 + }, + { + "epoch": 4.696996942995864, + "grad_norm": 0.020267620705528608, + "learning_rate": 1.6511519079555886e-06, + "loss": 0.0001, + "step": 6530 + }, + { + "epoch": 4.6977162380866755, + "grad_norm": 3.0013982511980175, + "learning_rate": 1.6508133912383615e-06, + "loss": 0.0082, + "step": 6531 + }, + { + "epoch": 4.698435533177486, + "grad_norm": 3.028172516227196, + "learning_rate": 1.6504748667615214e-06, + "loss": 0.0453, + "step": 6532 + }, + { + "epoch": 4.699154828268297, + "grad_norm": 1.5249826238036401, + "learning_rate": 1.650136334542485e-06, + "loss": 0.0239, + "step": 6533 + }, + { + "epoch": 4.699874123359108, + "grad_norm": 0.28462234150811333, + "learning_rate": 1.6497977945986703e-06, + "loss": 0.0005, + "step": 6534 + }, + { + "epoch": 4.700593418449919, + "grad_norm": 1.1257685169977625, + "learning_rate": 1.6494592469474966e-06, + "loss": 0.0123, + "step": 6535 + }, + { + "epoch": 4.70131271354073, + "grad_norm": 0.006158691904360141, + "learning_rate": 1.6491206916063822e-06, + "loss": 0.0, + "step": 6536 + }, + { + "epoch": 4.702032008631541, + "grad_norm": 2.8865414487225065, + "learning_rate": 1.6487821285927462e-06, + "loss": 0.0695, + "step": 6537 + }, + { + "epoch": 4.702751303722352, + "grad_norm": 0.011013578127726378, + "learning_rate": 1.6484435579240086e-06, + "loss": 0.0001, + "step": 6538 + }, + { + "epoch": 4.703470598813163, + "grad_norm": 3.039821422397097, + "learning_rate": 1.6481049796175894e-06, + "loss": 0.0165, + "step": 6539 + }, + { + "epoch": 4.704189893903974, + "grad_norm": 0.4117452852756537, + "learning_rate": 1.6477663936909092e-06, + "loss": 0.0006, + "step": 6540 + }, + { + "epoch": 4.704909188994785, + "grad_norm": 0.8756478784006266, + "learning_rate": 1.6474278001613883e-06, + "loss": 0.0029, + "step": 6541 + }, + { + "epoch": 4.705628484085596, + "grad_norm": 2.696448236974491, + "learning_rate": 1.6470891990464487e-06, + "loss": 0.0328, + "step": 6542 + }, + { + "epoch": 4.706347779176407, + "grad_norm": 0.24159987023481005, + "learning_rate": 1.6467505903635114e-06, + "loss": 0.0007, + "step": 6543 + }, + { + "epoch": 4.7070670742672185, + "grad_norm": 2.808714426147137, + "learning_rate": 1.646411974129998e-06, + "loss": 0.0679, + "step": 6544 + }, + { + "epoch": 4.707786369358029, + "grad_norm": 2.924684879359135, + "learning_rate": 1.6460733503633326e-06, + "loss": 0.0721, + "step": 6545 + }, + { + "epoch": 4.70850566444884, + "grad_norm": 2.4795727895121527, + "learning_rate": 1.6457347190809361e-06, + "loss": 0.009, + "step": 6546 + }, + { + "epoch": 4.709224959539651, + "grad_norm": 4.3319672386380335, + "learning_rate": 1.6453960803002327e-06, + "loss": 0.1279, + "step": 6547 + }, + { + "epoch": 4.709944254630463, + "grad_norm": 2.3639184805477864, + "learning_rate": 1.6450574340386458e-06, + "loss": 0.0071, + "step": 6548 + }, + { + "epoch": 4.710663549721273, + "grad_norm": 3.860762390887824, + "learning_rate": 1.6447187803135994e-06, + "loss": 0.0891, + "step": 6549 + }, + { + "epoch": 4.711382844812084, + "grad_norm": 5.619334620226762, + "learning_rate": 1.6443801191425176e-06, + "loss": 0.0895, + "step": 6550 + }, + { + "epoch": 4.712102139902895, + "grad_norm": 3.5082983332688196, + "learning_rate": 1.6440414505428254e-06, + "loss": 0.0056, + "step": 6551 + }, + { + "epoch": 4.712821434993706, + "grad_norm": 0.015211413751784073, + "learning_rate": 1.643702774531947e-06, + "loss": 0.0, + "step": 6552 + }, + { + "epoch": 4.713540730084517, + "grad_norm": 0.11714295231073918, + "learning_rate": 1.6433640911273091e-06, + "loss": 0.0003, + "step": 6553 + }, + { + "epoch": 4.714260025175328, + "grad_norm": 3.2747817038404374, + "learning_rate": 1.6430254003463372e-06, + "loss": 0.0275, + "step": 6554 + }, + { + "epoch": 4.714979320266139, + "grad_norm": 4.250666728443854, + "learning_rate": 1.642686702206457e-06, + "loss": 0.0846, + "step": 6555 + }, + { + "epoch": 4.71569861535695, + "grad_norm": 6.045174307541767, + "learning_rate": 1.6423479967250957e-06, + "loss": 0.1839, + "step": 6556 + }, + { + "epoch": 4.7164179104477615, + "grad_norm": 7.572205434697697, + "learning_rate": 1.64200928391968e-06, + "loss": 0.1795, + "step": 6557 + }, + { + "epoch": 4.717137205538572, + "grad_norm": 3.119438471682456, + "learning_rate": 1.6416705638076378e-06, + "loss": 0.0619, + "step": 6558 + }, + { + "epoch": 4.7178565006293836, + "grad_norm": 4.553132435300819, + "learning_rate": 1.641331836406396e-06, + "loss": 0.1249, + "step": 6559 + }, + { + "epoch": 4.718575795720194, + "grad_norm": 1.0764592810621612, + "learning_rate": 1.6409931017333828e-06, + "loss": 0.013, + "step": 6560 + }, + { + "epoch": 4.719295090811006, + "grad_norm": 3.4793045344452342, + "learning_rate": 1.6406543598060278e-06, + "loss": 0.1069, + "step": 6561 + }, + { + "epoch": 4.720014385901816, + "grad_norm": 6.916298392486317, + "learning_rate": 1.6403156106417584e-06, + "loss": 0.228, + "step": 6562 + }, + { + "epoch": 4.720733680992627, + "grad_norm": 2.1302706667150813, + "learning_rate": 1.6399768542580046e-06, + "loss": 0.0092, + "step": 6563 + }, + { + "epoch": 4.721452976083438, + "grad_norm": 0.6513184695423738, + "learning_rate": 1.639638090672196e-06, + "loss": 0.0063, + "step": 6564 + }, + { + "epoch": 4.722172271174249, + "grad_norm": 1.8818930955650184, + "learning_rate": 1.639299319901763e-06, + "loss": 0.0451, + "step": 6565 + }, + { + "epoch": 4.72289156626506, + "grad_norm": 0.8815150738577637, + "learning_rate": 1.638960541964135e-06, + "loss": 0.0012, + "step": 6566 + }, + { + "epoch": 4.723610861355871, + "grad_norm": 7.751911394340983, + "learning_rate": 1.6386217568767433e-06, + "loss": 0.1096, + "step": 6567 + }, + { + "epoch": 4.724330156446682, + "grad_norm": 1.2781880345518477, + "learning_rate": 1.6382829646570186e-06, + "loss": 0.0046, + "step": 6568 + }, + { + "epoch": 4.725049451537493, + "grad_norm": 2.9657444856219897, + "learning_rate": 1.6379441653223928e-06, + "loss": 0.0687, + "step": 6569 + }, + { + "epoch": 4.7257687466283045, + "grad_norm": 2.6072359994105576, + "learning_rate": 1.6376053588902977e-06, + "loss": 0.0374, + "step": 6570 + }, + { + "epoch": 4.726488041719115, + "grad_norm": 4.443316569690892, + "learning_rate": 1.6372665453781647e-06, + "loss": 0.0857, + "step": 6571 + }, + { + "epoch": 4.7272073368099266, + "grad_norm": 3.864337894981979, + "learning_rate": 1.636927724803428e-06, + "loss": 0.0795, + "step": 6572 + }, + { + "epoch": 4.727926631900737, + "grad_norm": 0.8778242631280578, + "learning_rate": 1.6365888971835188e-06, + "loss": 0.0084, + "step": 6573 + }, + { + "epoch": 4.728645926991549, + "grad_norm": 2.3967458878654213, + "learning_rate": 1.6362500625358712e-06, + "loss": 0.0453, + "step": 6574 + }, + { + "epoch": 4.729365222082359, + "grad_norm": 0.06003333759975409, + "learning_rate": 1.6359112208779189e-06, + "loss": 0.0002, + "step": 6575 + }, + { + "epoch": 4.73008451717317, + "grad_norm": 4.056048693757284, + "learning_rate": 1.6355723722270953e-06, + "loss": 0.0542, + "step": 6576 + }, + { + "epoch": 4.730803812263981, + "grad_norm": 2.4577947166269323, + "learning_rate": 1.6352335166008353e-06, + "loss": 0.0541, + "step": 6577 + }, + { + "epoch": 4.731523107354793, + "grad_norm": 4.416574326409258, + "learning_rate": 1.6348946540165736e-06, + "loss": 0.0751, + "step": 6578 + }, + { + "epoch": 4.732242402445603, + "grad_norm": 2.7248817382047985, + "learning_rate": 1.634555784491745e-06, + "loss": 0.0318, + "step": 6579 + }, + { + "epoch": 4.732961697536414, + "grad_norm": 4.205021373120423, + "learning_rate": 1.6342169080437856e-06, + "loss": 0.1051, + "step": 6580 + }, + { + "epoch": 4.733680992627225, + "grad_norm": 1.3773545305028412, + "learning_rate": 1.6338780246901299e-06, + "loss": 0.0306, + "step": 6581 + }, + { + "epoch": 4.734400287718036, + "grad_norm": 0.21989141972407308, + "learning_rate": 1.6335391344482153e-06, + "loss": 0.0003, + "step": 6582 + }, + { + "epoch": 4.7351195828088475, + "grad_norm": 2.4734122965524623, + "learning_rate": 1.6332002373354776e-06, + "loss": 0.0489, + "step": 6583 + }, + { + "epoch": 4.735838877899658, + "grad_norm": 2.2841165637039, + "learning_rate": 1.6328613333693537e-06, + "loss": 0.0453, + "step": 6584 + }, + { + "epoch": 4.7365581729904696, + "grad_norm": 2.7695349714766966, + "learning_rate": 1.6325224225672813e-06, + "loss": 0.0145, + "step": 6585 + }, + { + "epoch": 4.73727746808128, + "grad_norm": 2.147380135046358, + "learning_rate": 1.6321835049466975e-06, + "loss": 0.004, + "step": 6586 + }, + { + "epoch": 4.737996763172092, + "grad_norm": 4.088802444960838, + "learning_rate": 1.6318445805250407e-06, + "loss": 0.1374, + "step": 6587 + }, + { + "epoch": 4.738716058262902, + "grad_norm": 2.5213739018562746, + "learning_rate": 1.6315056493197486e-06, + "loss": 0.0331, + "step": 6588 + }, + { + "epoch": 4.739435353353714, + "grad_norm": 5.716880193238295, + "learning_rate": 1.6311667113482595e-06, + "loss": 0.2216, + "step": 6589 + }, + { + "epoch": 4.740154648444524, + "grad_norm": 1.568316471621854, + "learning_rate": 1.6308277666280131e-06, + "loss": 0.0165, + "step": 6590 + }, + { + "epoch": 4.740873943535336, + "grad_norm": 1.4315450144532678, + "learning_rate": 1.6304888151764487e-06, + "loss": 0.025, + "step": 6591 + }, + { + "epoch": 4.741593238626146, + "grad_norm": 4.251320553948076, + "learning_rate": 1.6301498570110055e-06, + "loss": 0.0771, + "step": 6592 + }, + { + "epoch": 4.742312533716957, + "grad_norm": 3.2824248969746344, + "learning_rate": 1.6298108921491237e-06, + "loss": 0.029, + "step": 6593 + }, + { + "epoch": 4.743031828807768, + "grad_norm": 2.9486056485803016, + "learning_rate": 1.6294719206082436e-06, + "loss": 0.0813, + "step": 6594 + }, + { + "epoch": 4.74375112389858, + "grad_norm": 2.9897686005391058, + "learning_rate": 1.629132942405806e-06, + "loss": 0.0551, + "step": 6595 + }, + { + "epoch": 4.7444704189893905, + "grad_norm": 3.2771272382716594, + "learning_rate": 1.628793957559252e-06, + "loss": 0.1067, + "step": 6596 + }, + { + "epoch": 4.745189714080201, + "grad_norm": 3.754369800403645, + "learning_rate": 1.6284549660860222e-06, + "loss": 0.0221, + "step": 6597 + }, + { + "epoch": 4.7459090091710125, + "grad_norm": 3.46843292847435, + "learning_rate": 1.6281159680035593e-06, + "loss": 0.0447, + "step": 6598 + }, + { + "epoch": 4.746628304261823, + "grad_norm": 2.753179116635653, + "learning_rate": 1.6277769633293047e-06, + "loss": 0.0382, + "step": 6599 + }, + { + "epoch": 4.747347599352635, + "grad_norm": 4.6835557863794195, + "learning_rate": 1.6274379520807014e-06, + "loss": 0.172, + "step": 6600 + }, + { + "epoch": 4.748066894443445, + "grad_norm": 2.899085891995275, + "learning_rate": 1.627098934275192e-06, + "loss": 0.0565, + "step": 6601 + }, + { + "epoch": 4.748786189534257, + "grad_norm": 2.2751243017449228, + "learning_rate": 1.6267599099302185e-06, + "loss": 0.0672, + "step": 6602 + }, + { + "epoch": 4.749505484625067, + "grad_norm": 3.458029907078376, + "learning_rate": 1.6264208790632258e-06, + "loss": 0.0718, + "step": 6603 + }, + { + "epoch": 4.750224779715879, + "grad_norm": 2.355306839577528, + "learning_rate": 1.6260818416916565e-06, + "loss": 0.0097, + "step": 6604 + }, + { + "epoch": 4.750944074806689, + "grad_norm": 0.8458728399000682, + "learning_rate": 1.6257427978329551e-06, + "loss": 0.0079, + "step": 6605 + }, + { + "epoch": 4.751663369897501, + "grad_norm": 3.1108443080007406, + "learning_rate": 1.6254037475045662e-06, + "loss": 0.0912, + "step": 6606 + }, + { + "epoch": 4.752382664988311, + "grad_norm": 4.624818512357233, + "learning_rate": 1.6250646907239342e-06, + "loss": 0.098, + "step": 6607 + }, + { + "epoch": 4.753101960079123, + "grad_norm": 1.0871994527167856, + "learning_rate": 1.624725627508505e-06, + "loss": 0.0126, + "step": 6608 + }, + { + "epoch": 4.7538212551699335, + "grad_norm": 0.1200205607032814, + "learning_rate": 1.6243865578757227e-06, + "loss": 0.0007, + "step": 6609 + }, + { + "epoch": 4.754540550260744, + "grad_norm": 1.4175858823441438, + "learning_rate": 1.6240474818430337e-06, + "loss": 0.0297, + "step": 6610 + }, + { + "epoch": 4.7552598453515555, + "grad_norm": 2.0861543308192227, + "learning_rate": 1.6237083994278839e-06, + "loss": 0.0409, + "step": 6611 + }, + { + "epoch": 4.755979140442366, + "grad_norm": 5.556103694696101, + "learning_rate": 1.6233693106477196e-06, + "loss": 0.0443, + "step": 6612 + }, + { + "epoch": 4.756698435533178, + "grad_norm": 2.4314551125713577, + "learning_rate": 1.6230302155199882e-06, + "loss": 0.068, + "step": 6613 + }, + { + "epoch": 4.757417730623988, + "grad_norm": 4.67423897403326, + "learning_rate": 1.6226911140621363e-06, + "loss": 0.145, + "step": 6614 + }, + { + "epoch": 4.7581370257148, + "grad_norm": 4.206076874298304, + "learning_rate": 1.622352006291611e-06, + "loss": 0.114, + "step": 6615 + }, + { + "epoch": 4.75885632080561, + "grad_norm": 0.9649288585055931, + "learning_rate": 1.6220128922258601e-06, + "loss": 0.0076, + "step": 6616 + }, + { + "epoch": 4.759575615896422, + "grad_norm": 2.061743780723609, + "learning_rate": 1.6216737718823319e-06, + "loss": 0.0174, + "step": 6617 + }, + { + "epoch": 4.760294910987232, + "grad_norm": 1.7425140258316574, + "learning_rate": 1.621334645278474e-06, + "loss": 0.0373, + "step": 6618 + }, + { + "epoch": 4.761014206078044, + "grad_norm": 1.3780723540255881, + "learning_rate": 1.6209955124317363e-06, + "loss": 0.0208, + "step": 6619 + }, + { + "epoch": 4.761733501168854, + "grad_norm": 2.474572012351872, + "learning_rate": 1.6206563733595666e-06, + "loss": 0.0449, + "step": 6620 + }, + { + "epoch": 4.762452796259666, + "grad_norm": 3.464544214917929, + "learning_rate": 1.6203172280794148e-06, + "loss": 0.0948, + "step": 6621 + }, + { + "epoch": 4.7631720913504765, + "grad_norm": 6.119721926203527, + "learning_rate": 1.6199780766087309e-06, + "loss": 0.1919, + "step": 6622 + }, + { + "epoch": 4.763891386441287, + "grad_norm": 2.213435499373046, + "learning_rate": 1.6196389189649637e-06, + "loss": 0.0426, + "step": 6623 + }, + { + "epoch": 4.7646106815320985, + "grad_norm": 1.341804031199311, + "learning_rate": 1.6192997551655644e-06, + "loss": 0.0263, + "step": 6624 + }, + { + "epoch": 4.76532997662291, + "grad_norm": 1.256856760838146, + "learning_rate": 1.6189605852279833e-06, + "loss": 0.0185, + "step": 6625 + }, + { + "epoch": 4.766049271713721, + "grad_norm": 2.3814426699080173, + "learning_rate": 1.618621409169671e-06, + "loss": 0.0379, + "step": 6626 + }, + { + "epoch": 4.766768566804531, + "grad_norm": 1.5527251994107472, + "learning_rate": 1.6182822270080793e-06, + "loss": 0.0304, + "step": 6627 + }, + { + "epoch": 4.767487861895343, + "grad_norm": 0.8966741504973903, + "learning_rate": 1.6179430387606587e-06, + "loss": 0.0065, + "step": 6628 + }, + { + "epoch": 4.768207156986153, + "grad_norm": 0.007416735009964656, + "learning_rate": 1.6176038444448625e-06, + "loss": 0.0, + "step": 6629 + }, + { + "epoch": 4.768926452076965, + "grad_norm": 0.00927455194000191, + "learning_rate": 1.617264644078142e-06, + "loss": 0.0001, + "step": 6630 + }, + { + "epoch": 4.769645747167775, + "grad_norm": 1.3318793604752466, + "learning_rate": 1.6169254376779493e-06, + "loss": 0.0227, + "step": 6631 + }, + { + "epoch": 4.770365042258587, + "grad_norm": 4.662475413843208, + "learning_rate": 1.6165862252617379e-06, + "loss": 0.0668, + "step": 6632 + }, + { + "epoch": 4.771084337349397, + "grad_norm": 3.3695521901643244, + "learning_rate": 1.6162470068469602e-06, + "loss": 0.007, + "step": 6633 + }, + { + "epoch": 4.771803632440209, + "grad_norm": 0.03204678716148783, + "learning_rate": 1.6159077824510702e-06, + "loss": 0.0001, + "step": 6634 + }, + { + "epoch": 4.7725229275310195, + "grad_norm": 2.7948644910970883, + "learning_rate": 1.6155685520915214e-06, + "loss": 0.0566, + "step": 6635 + }, + { + "epoch": 4.773242222621831, + "grad_norm": 2.1981543310442557, + "learning_rate": 1.6152293157857672e-06, + "loss": 0.0286, + "step": 6636 + }, + { + "epoch": 4.7739615177126415, + "grad_norm": 3.408815245376647, + "learning_rate": 1.6148900735512633e-06, + "loss": 0.0741, + "step": 6637 + }, + { + "epoch": 4.774680812803453, + "grad_norm": 5.215023502092897, + "learning_rate": 1.614550825405463e-06, + "loss": 0.2012, + "step": 6638 + }, + { + "epoch": 4.775400107894264, + "grad_norm": 1.5627966811306928, + "learning_rate": 1.6142115713658213e-06, + "loss": 0.0234, + "step": 6639 + }, + { + "epoch": 4.776119402985074, + "grad_norm": 4.105601484263425, + "learning_rate": 1.6138723114497941e-06, + "loss": 0.1129, + "step": 6640 + }, + { + "epoch": 4.776838698075886, + "grad_norm": 2.400756237021966, + "learning_rate": 1.6135330456748364e-06, + "loss": 0.0426, + "step": 6641 + }, + { + "epoch": 4.777557993166696, + "grad_norm": 3.0173375364636774, + "learning_rate": 1.6131937740584043e-06, + "loss": 0.0587, + "step": 6642 + }, + { + "epoch": 4.778277288257508, + "grad_norm": 4.015171378068447, + "learning_rate": 1.6128544966179539e-06, + "loss": 0.1025, + "step": 6643 + }, + { + "epoch": 4.778996583348318, + "grad_norm": 2.2693778133251272, + "learning_rate": 1.6125152133709416e-06, + "loss": 0.0288, + "step": 6644 + }, + { + "epoch": 4.77971587843913, + "grad_norm": 1.2078672462036353, + "learning_rate": 1.6121759243348242e-06, + "loss": 0.0167, + "step": 6645 + }, + { + "epoch": 4.78043517352994, + "grad_norm": 0.12200278628204757, + "learning_rate": 1.6118366295270582e-06, + "loss": 0.0004, + "step": 6646 + }, + { + "epoch": 4.781154468620752, + "grad_norm": 2.576685560461319, + "learning_rate": 1.611497328965101e-06, + "loss": 0.0696, + "step": 6647 + }, + { + "epoch": 4.7818737637115625, + "grad_norm": 3.015459251016891, + "learning_rate": 1.6111580226664113e-06, + "loss": 0.0654, + "step": 6648 + }, + { + "epoch": 4.782593058802374, + "grad_norm": 0.6534022054944869, + "learning_rate": 1.6108187106484454e-06, + "loss": 0.002, + "step": 6649 + }, + { + "epoch": 4.7833123538931845, + "grad_norm": 2.5447090037473443, + "learning_rate": 1.6104793929286629e-06, + "loss": 0.0505, + "step": 6650 + }, + { + "epoch": 4.784031648983996, + "grad_norm": 1.5056322790472445, + "learning_rate": 1.6101400695245216e-06, + "loss": 0.0166, + "step": 6651 + }, + { + "epoch": 4.784750944074807, + "grad_norm": 2.4814757863648986, + "learning_rate": 1.6098007404534804e-06, + "loss": 0.0276, + "step": 6652 + }, + { + "epoch": 4.785470239165617, + "grad_norm": 3.597848195857391, + "learning_rate": 1.6094614057329985e-06, + "loss": 0.057, + "step": 6653 + }, + { + "epoch": 4.786189534256429, + "grad_norm": 1.7976814023218555, + "learning_rate": 1.6091220653805345e-06, + "loss": 0.0369, + "step": 6654 + }, + { + "epoch": 4.78690882934724, + "grad_norm": 4.27885369435141, + "learning_rate": 1.6087827194135489e-06, + "loss": 0.1011, + "step": 6655 + }, + { + "epoch": 4.787628124438051, + "grad_norm": 1.1844980229198272, + "learning_rate": 1.6084433678495016e-06, + "loss": 0.0175, + "step": 6656 + }, + { + "epoch": 4.788347419528861, + "grad_norm": 3.8417402565314296, + "learning_rate": 1.6081040107058525e-06, + "loss": 0.1252, + "step": 6657 + }, + { + "epoch": 4.789066714619673, + "grad_norm": 0.4136756546685581, + "learning_rate": 1.6077646480000624e-06, + "loss": 0.003, + "step": 6658 + }, + { + "epoch": 4.789786009710483, + "grad_norm": 7.067699443507985, + "learning_rate": 1.6074252797495921e-06, + "loss": 0.1817, + "step": 6659 + }, + { + "epoch": 4.790505304801295, + "grad_norm": 3.673213006226901, + "learning_rate": 1.6070859059719026e-06, + "loss": 0.0317, + "step": 6660 + }, + { + "epoch": 4.7912245998921055, + "grad_norm": 3.301898867547547, + "learning_rate": 1.6067465266844551e-06, + "loss": 0.082, + "step": 6661 + }, + { + "epoch": 4.791943894982917, + "grad_norm": 3.7912665979150155, + "learning_rate": 1.6064071419047112e-06, + "loss": 0.1165, + "step": 6662 + }, + { + "epoch": 4.7926631900737275, + "grad_norm": 1.7071195760473994, + "learning_rate": 1.6060677516501335e-06, + "loss": 0.0307, + "step": 6663 + }, + { + "epoch": 4.793382485164539, + "grad_norm": 4.463282618717407, + "learning_rate": 1.6057283559381836e-06, + "loss": 0.1879, + "step": 6664 + }, + { + "epoch": 4.79410178025535, + "grad_norm": 6.722305708790678, + "learning_rate": 1.605388954786324e-06, + "loss": 0.2093, + "step": 6665 + }, + { + "epoch": 4.794821075346161, + "grad_norm": 6.130977914513703, + "learning_rate": 1.6050495482120177e-06, + "loss": 0.0528, + "step": 6666 + }, + { + "epoch": 4.795540370436972, + "grad_norm": 2.221727334818281, + "learning_rate": 1.6047101362327284e-06, + "loss": 0.0485, + "step": 6667 + }, + { + "epoch": 4.796259665527783, + "grad_norm": 1.7877962116165145, + "learning_rate": 1.604370718865918e-06, + "loss": 0.0258, + "step": 6668 + }, + { + "epoch": 4.796978960618594, + "grad_norm": 0.14443511670909398, + "learning_rate": 1.6040312961290512e-06, + "loss": 0.0004, + "step": 6669 + }, + { + "epoch": 4.797698255709404, + "grad_norm": 2.315660000417737, + "learning_rate": 1.6036918680395913e-06, + "loss": 0.0346, + "step": 6670 + }, + { + "epoch": 4.798417550800216, + "grad_norm": 1.7676768374267036, + "learning_rate": 1.6033524346150029e-06, + "loss": 0.0392, + "step": 6671 + }, + { + "epoch": 4.799136845891027, + "grad_norm": 3.048823354087359, + "learning_rate": 1.6030129958727505e-06, + "loss": 0.0339, + "step": 6672 + }, + { + "epoch": 4.799856140981838, + "grad_norm": 3.6571598073350917, + "learning_rate": 1.6026735518302988e-06, + "loss": 0.0717, + "step": 6673 + }, + { + "epoch": 4.8005754360726485, + "grad_norm": 1.9238298256789736, + "learning_rate": 1.6023341025051124e-06, + "loss": 0.0283, + "step": 6674 + }, + { + "epoch": 4.80129473116346, + "grad_norm": 4.065709083245872, + "learning_rate": 1.6019946479146565e-06, + "loss": 0.1061, + "step": 6675 + }, + { + "epoch": 4.8020140262542705, + "grad_norm": 0.44516085083774765, + "learning_rate": 1.6016551880763974e-06, + "loss": 0.0031, + "step": 6676 + }, + { + "epoch": 4.802733321345082, + "grad_norm": 1.8349103885294602, + "learning_rate": 1.6013157230078e-06, + "loss": 0.0459, + "step": 6677 + }, + { + "epoch": 4.803452616435893, + "grad_norm": 2.9382208570386674, + "learning_rate": 1.6009762527263308e-06, + "loss": 0.1104, + "step": 6678 + }, + { + "epoch": 4.804171911526704, + "grad_norm": 2.516803366270564, + "learning_rate": 1.6006367772494564e-06, + "loss": 0.0352, + "step": 6679 + }, + { + "epoch": 4.804891206617515, + "grad_norm": 2.835494286673339, + "learning_rate": 1.600297296594643e-06, + "loss": 0.0542, + "step": 6680 + }, + { + "epoch": 4.805610501708326, + "grad_norm": 2.0419836980110535, + "learning_rate": 1.599957810779358e-06, + "loss": 0.0559, + "step": 6681 + }, + { + "epoch": 4.806329796799137, + "grad_norm": 0.5506690152676075, + "learning_rate": 1.599618319821068e-06, + "loss": 0.0049, + "step": 6682 + }, + { + "epoch": 4.807049091889948, + "grad_norm": 1.6544184634123194, + "learning_rate": 1.5992788237372401e-06, + "loss": 0.0185, + "step": 6683 + }, + { + "epoch": 4.807768386980759, + "grad_norm": 3.789498259079538, + "learning_rate": 1.5989393225453427e-06, + "loss": 0.0969, + "step": 6684 + }, + { + "epoch": 4.80848768207157, + "grad_norm": 3.036617442515559, + "learning_rate": 1.5985998162628436e-06, + "loss": 0.0334, + "step": 6685 + }, + { + "epoch": 4.809206977162381, + "grad_norm": 4.9237217123372305, + "learning_rate": 1.5982603049072108e-06, + "loss": 0.086, + "step": 6686 + }, + { + "epoch": 4.8099262722531915, + "grad_norm": 0.3278359206722524, + "learning_rate": 1.5979207884959126e-06, + "loss": 0.0007, + "step": 6687 + }, + { + "epoch": 4.810645567344003, + "grad_norm": 3.5271264577970065, + "learning_rate": 1.5975812670464183e-06, + "loss": 0.0991, + "step": 6688 + }, + { + "epoch": 4.8113648624348135, + "grad_norm": 4.434018934871093, + "learning_rate": 1.5972417405761964e-06, + "loss": 0.054, + "step": 6689 + }, + { + "epoch": 4.812084157525625, + "grad_norm": 2.714429501774179, + "learning_rate": 1.5969022091027165e-06, + "loss": 0.0569, + "step": 6690 + }, + { + "epoch": 4.812803452616436, + "grad_norm": 6.313264875806313, + "learning_rate": 1.5965626726434473e-06, + "loss": 0.2117, + "step": 6691 + }, + { + "epoch": 4.813522747707247, + "grad_norm": 0.7509971022915536, + "learning_rate": 1.5962231312158596e-06, + "loss": 0.0078, + "step": 6692 + }, + { + "epoch": 4.814242042798058, + "grad_norm": 2.2870025004196752, + "learning_rate": 1.5958835848374225e-06, + "loss": 0.028, + "step": 6693 + }, + { + "epoch": 4.814961337888869, + "grad_norm": 2.934685449229407, + "learning_rate": 1.5955440335256066e-06, + "loss": 0.0498, + "step": 6694 + }, + { + "epoch": 4.81568063297968, + "grad_norm": 2.775169653099012, + "learning_rate": 1.5952044772978827e-06, + "loss": 0.0659, + "step": 6695 + }, + { + "epoch": 4.816399928070491, + "grad_norm": 0.1606972841576902, + "learning_rate": 1.5948649161717213e-06, + "loss": 0.0009, + "step": 6696 + }, + { + "epoch": 4.817119223161302, + "grad_norm": 4.178416607132933, + "learning_rate": 1.5945253501645935e-06, + "loss": 0.1009, + "step": 6697 + }, + { + "epoch": 4.817838518252113, + "grad_norm": 1.6209432725269848, + "learning_rate": 1.5941857792939703e-06, + "loss": 0.0317, + "step": 6698 + }, + { + "epoch": 4.818557813342924, + "grad_norm": 0.07655337408442657, + "learning_rate": 1.5938462035773231e-06, + "loss": 0.0001, + "step": 6699 + }, + { + "epoch": 4.8192771084337345, + "grad_norm": 2.215698900457913, + "learning_rate": 1.5935066230321244e-06, + "loss": 0.0396, + "step": 6700 + }, + { + "epoch": 4.819996403524546, + "grad_norm": 4.059331039769088, + "learning_rate": 1.593167037675846e-06, + "loss": 0.1531, + "step": 6701 + }, + { + "epoch": 4.820715698615357, + "grad_norm": 5.184959794900499, + "learning_rate": 1.59282744752596e-06, + "loss": 0.0801, + "step": 6702 + }, + { + "epoch": 4.821434993706168, + "grad_norm": 2.40373487367325, + "learning_rate": 1.5924878525999386e-06, + "loss": 0.0219, + "step": 6703 + }, + { + "epoch": 4.822154288796979, + "grad_norm": 1.7549032107798144, + "learning_rate": 1.5921482529152548e-06, + "loss": 0.0418, + "step": 6704 + }, + { + "epoch": 4.82287358388779, + "grad_norm": 0.7313247213243055, + "learning_rate": 1.5918086484893815e-06, + "loss": 0.0084, + "step": 6705 + }, + { + "epoch": 4.823592878978601, + "grad_norm": 2.6734887196778656, + "learning_rate": 1.5914690393397923e-06, + "loss": 0.038, + "step": 6706 + }, + { + "epoch": 4.824312174069412, + "grad_norm": 1.2298033749821438, + "learning_rate": 1.5911294254839601e-06, + "loss": 0.0045, + "step": 6707 + }, + { + "epoch": 4.825031469160223, + "grad_norm": 0.015289410634952233, + "learning_rate": 1.5907898069393595e-06, + "loss": 0.0001, + "step": 6708 + }, + { + "epoch": 4.825750764251034, + "grad_norm": 2.4835781099639704, + "learning_rate": 1.5904501837234634e-06, + "loss": 0.0491, + "step": 6709 + }, + { + "epoch": 4.826470059341845, + "grad_norm": 4.1198955430691395, + "learning_rate": 1.5901105558537473e-06, + "loss": 0.0832, + "step": 6710 + }, + { + "epoch": 4.827189354432656, + "grad_norm": 3.0671893815539257, + "learning_rate": 1.5897709233476847e-06, + "loss": 0.0627, + "step": 6711 + }, + { + "epoch": 4.827908649523467, + "grad_norm": 0.1282277107326838, + "learning_rate": 1.58943128622275e-06, + "loss": 0.0003, + "step": 6712 + }, + { + "epoch": 4.828627944614278, + "grad_norm": 4.507925985578353, + "learning_rate": 1.5890916444964188e-06, + "loss": 0.1634, + "step": 6713 + }, + { + "epoch": 4.829347239705089, + "grad_norm": 1.8474056479367535, + "learning_rate": 1.5887519981861665e-06, + "loss": 0.0363, + "step": 6714 + }, + { + "epoch": 4.8300665347959, + "grad_norm": 3.8054742298307254, + "learning_rate": 1.5884123473094674e-06, + "loss": 0.0947, + "step": 6715 + }, + { + "epoch": 4.830785829886711, + "grad_norm": 2.303788837422596, + "learning_rate": 1.5880726918837984e-06, + "loss": 0.0626, + "step": 6716 + }, + { + "epoch": 4.831505124977522, + "grad_norm": 1.9612485039303238, + "learning_rate": 1.5877330319266347e-06, + "loss": 0.0207, + "step": 6717 + }, + { + "epoch": 4.832224420068333, + "grad_norm": 4.37496206345724, + "learning_rate": 1.5873933674554526e-06, + "loss": 0.1327, + "step": 6718 + }, + { + "epoch": 4.832943715159144, + "grad_norm": 0.03428609791445765, + "learning_rate": 1.5870536984877281e-06, + "loss": 0.0002, + "step": 6719 + }, + { + "epoch": 4.833663010249955, + "grad_norm": 0.8805898729077196, + "learning_rate": 1.586714025040938e-06, + "loss": 0.0091, + "step": 6720 + }, + { + "epoch": 4.834382305340766, + "grad_norm": 2.643412072545456, + "learning_rate": 1.5863743471325592e-06, + "loss": 0.0443, + "step": 6721 + }, + { + "epoch": 4.835101600431577, + "grad_norm": 2.2575744354464593, + "learning_rate": 1.5860346647800687e-06, + "loss": 0.0394, + "step": 6722 + }, + { + "epoch": 4.835820895522388, + "grad_norm": 6.2174515266789045, + "learning_rate": 1.5856949780009433e-06, + "loss": 0.1205, + "step": 6723 + }, + { + "epoch": 4.836540190613199, + "grad_norm": 1.7891573071431193, + "learning_rate": 1.5853552868126613e-06, + "loss": 0.0292, + "step": 6724 + }, + { + "epoch": 4.83725948570401, + "grad_norm": 3.2932179622499027, + "learning_rate": 1.5850155912326998e-06, + "loss": 0.0846, + "step": 6725 + }, + { + "epoch": 4.837978780794821, + "grad_norm": 1.876191813987112, + "learning_rate": 1.5846758912785368e-06, + "loss": 0.0423, + "step": 6726 + }, + { + "epoch": 4.838698075885632, + "grad_norm": 2.7518088113484658, + "learning_rate": 1.5843361869676509e-06, + "loss": 0.0402, + "step": 6727 + }, + { + "epoch": 4.839417370976443, + "grad_norm": 2.727303040615984, + "learning_rate": 1.5839964783175195e-06, + "loss": 0.0382, + "step": 6728 + }, + { + "epoch": 4.840136666067254, + "grad_norm": 2.7589392985457035, + "learning_rate": 1.5836567653456222e-06, + "loss": 0.073, + "step": 6729 + }, + { + "epoch": 4.840855961158065, + "grad_norm": 0.0031332090144005546, + "learning_rate": 1.5833170480694374e-06, + "loss": 0.0, + "step": 6730 + }, + { + "epoch": 4.841575256248876, + "grad_norm": 2.281074313500855, + "learning_rate": 1.5829773265064442e-06, + "loss": 0.0714, + "step": 6731 + }, + { + "epoch": 4.842294551339688, + "grad_norm": 1.3228380082450644, + "learning_rate": 1.5826376006741223e-06, + "loss": 0.0018, + "step": 6732 + }, + { + "epoch": 4.843013846430498, + "grad_norm": 0.24867106229151, + "learning_rate": 1.5822978705899504e-06, + "loss": 0.0006, + "step": 6733 + }, + { + "epoch": 4.843733141521309, + "grad_norm": 3.6081619845568977, + "learning_rate": 1.5819581362714087e-06, + "loss": 0.0342, + "step": 6734 + }, + { + "epoch": 4.84445243661212, + "grad_norm": 1.1816887693230518, + "learning_rate": 1.5816183977359767e-06, + "loss": 0.0234, + "step": 6735 + }, + { + "epoch": 4.845171731702931, + "grad_norm": 2.860108427909, + "learning_rate": 1.581278655001135e-06, + "loss": 0.0423, + "step": 6736 + }, + { + "epoch": 4.845891026793742, + "grad_norm": 3.8270835480208665, + "learning_rate": 1.5809389080843637e-06, + "loss": 0.1006, + "step": 6737 + }, + { + "epoch": 4.846610321884553, + "grad_norm": 0.5171648169457038, + "learning_rate": 1.5805991570031442e-06, + "loss": 0.0038, + "step": 6738 + }, + { + "epoch": 4.847329616975364, + "grad_norm": 4.508824547734458, + "learning_rate": 1.580259401774956e-06, + "loss": 0.1066, + "step": 6739 + }, + { + "epoch": 4.848048912066175, + "grad_norm": 0.13867238440594634, + "learning_rate": 1.5799196424172808e-06, + "loss": 0.0005, + "step": 6740 + }, + { + "epoch": 4.848768207156986, + "grad_norm": 0.009981787244968697, + "learning_rate": 1.5795798789475994e-06, + "loss": 0.0, + "step": 6741 + }, + { + "epoch": 4.849487502247797, + "grad_norm": 2.0473196702520755, + "learning_rate": 1.579240111383394e-06, + "loss": 0.0289, + "step": 6742 + }, + { + "epoch": 4.8502067973386085, + "grad_norm": 4.7860440940780675, + "learning_rate": 1.5789003397421458e-06, + "loss": 0.1162, + "step": 6743 + }, + { + "epoch": 4.850926092429419, + "grad_norm": 3.8453114105691903, + "learning_rate": 1.5785605640413362e-06, + "loss": 0.0693, + "step": 6744 + }, + { + "epoch": 4.851645387520231, + "grad_norm": 2.295803215923815, + "learning_rate": 1.5782207842984478e-06, + "loss": 0.0136, + "step": 6745 + }, + { + "epoch": 4.852364682611041, + "grad_norm": 2.385872563058216, + "learning_rate": 1.5778810005309627e-06, + "loss": 0.0483, + "step": 6746 + }, + { + "epoch": 4.853083977701852, + "grad_norm": 1.0016805517332124, + "learning_rate": 1.5775412127563637e-06, + "loss": 0.016, + "step": 6747 + }, + { + "epoch": 4.853803272792663, + "grad_norm": 1.9086144059640766, + "learning_rate": 1.577201420992133e-06, + "loss": 0.0524, + "step": 6748 + }, + { + "epoch": 4.854522567883475, + "grad_norm": 0.05985454497483043, + "learning_rate": 1.5768616252557537e-06, + "loss": 0.0002, + "step": 6749 + }, + { + "epoch": 4.855241862974285, + "grad_norm": 2.241063415003864, + "learning_rate": 1.5765218255647088e-06, + "loss": 0.0471, + "step": 6750 + }, + { + "epoch": 4.855961158065096, + "grad_norm": 0.03914530016355235, + "learning_rate": 1.5761820219364815e-06, + "loss": 0.0002, + "step": 6751 + }, + { + "epoch": 4.856680453155907, + "grad_norm": 1.9603991067594069, + "learning_rate": 1.5758422143885554e-06, + "loss": 0.0467, + "step": 6752 + }, + { + "epoch": 4.857399748246718, + "grad_norm": 2.4142428949079995, + "learning_rate": 1.5755024029384147e-06, + "loss": 0.0528, + "step": 6753 + }, + { + "epoch": 4.858119043337529, + "grad_norm": 0.15980872162603618, + "learning_rate": 1.5751625876035426e-06, + "loss": 0.0004, + "step": 6754 + }, + { + "epoch": 4.85883833842834, + "grad_norm": 1.5424121743583805, + "learning_rate": 1.5748227684014233e-06, + "loss": 0.0046, + "step": 6755 + }, + { + "epoch": 4.8595576335191515, + "grad_norm": 3.421520457249513, + "learning_rate": 1.5744829453495417e-06, + "loss": 0.0923, + "step": 6756 + }, + { + "epoch": 4.860276928609962, + "grad_norm": 1.7584653453639438, + "learning_rate": 1.574143118465381e-06, + "loss": 0.0047, + "step": 6757 + }, + { + "epoch": 4.8609962237007736, + "grad_norm": 0.04040458372983333, + "learning_rate": 1.573803287766427e-06, + "loss": 0.0001, + "step": 6758 + }, + { + "epoch": 4.861715518791584, + "grad_norm": 1.2396920650883414, + "learning_rate": 1.5734634532701643e-06, + "loss": 0.0055, + "step": 6759 + }, + { + "epoch": 4.862434813882396, + "grad_norm": 3.668628688669138, + "learning_rate": 1.5731236149940782e-06, + "loss": 0.0739, + "step": 6760 + }, + { + "epoch": 4.863154108973206, + "grad_norm": 1.947808345233405, + "learning_rate": 1.5727837729556541e-06, + "loss": 0.02, + "step": 6761 + }, + { + "epoch": 4.863873404064018, + "grad_norm": 0.3131161894987748, + "learning_rate": 1.5724439271723763e-06, + "loss": 0.0009, + "step": 6762 + }, + { + "epoch": 4.864592699154828, + "grad_norm": 0.14837396181846638, + "learning_rate": 1.5721040776617316e-06, + "loss": 0.0005, + "step": 6763 + }, + { + "epoch": 4.865311994245639, + "grad_norm": 5.05633388827718, + "learning_rate": 1.5717642244412054e-06, + "loss": 0.1103, + "step": 6764 + }, + { + "epoch": 4.86603128933645, + "grad_norm": 3.296029893557865, + "learning_rate": 1.571424367528284e-06, + "loss": 0.0502, + "step": 6765 + }, + { + "epoch": 4.866750584427261, + "grad_norm": 1.5034182249763814, + "learning_rate": 1.5710845069404535e-06, + "loss": 0.0318, + "step": 6766 + }, + { + "epoch": 4.867469879518072, + "grad_norm": 3.4931636147623215, + "learning_rate": 1.5707446426952002e-06, + "loss": 0.0701, + "step": 6767 + }, + { + "epoch": 4.868189174608883, + "grad_norm": 4.130023136522517, + "learning_rate": 1.5704047748100112e-06, + "loss": 0.1197, + "step": 6768 + }, + { + "epoch": 4.8689084696996945, + "grad_norm": 1.406744536501054, + "learning_rate": 1.5700649033023733e-06, + "loss": 0.0061, + "step": 6769 + }, + { + "epoch": 4.869627764790505, + "grad_norm": 0.6813038004434989, + "learning_rate": 1.569725028189772e-06, + "loss": 0.0122, + "step": 6770 + }, + { + "epoch": 4.8703470598813166, + "grad_norm": 2.364111117349552, + "learning_rate": 1.5693851494896962e-06, + "loss": 0.0351, + "step": 6771 + }, + { + "epoch": 4.871066354972127, + "grad_norm": 1.0885206565159502, + "learning_rate": 1.5690452672196324e-06, + "loss": 0.0015, + "step": 6772 + }, + { + "epoch": 4.871785650062939, + "grad_norm": 2.306566482825947, + "learning_rate": 1.5687053813970688e-06, + "loss": 0.0524, + "step": 6773 + }, + { + "epoch": 4.872504945153749, + "grad_norm": 2.624708962894925, + "learning_rate": 1.5683654920394928e-06, + "loss": 0.0455, + "step": 6774 + }, + { + "epoch": 4.873224240244561, + "grad_norm": 2.763830102260979, + "learning_rate": 1.5680255991643918e-06, + "loss": 0.0369, + "step": 6775 + }, + { + "epoch": 4.873943535335371, + "grad_norm": 2.4111615891626594, + "learning_rate": 1.5676857027892546e-06, + "loss": 0.033, + "step": 6776 + }, + { + "epoch": 4.874662830426182, + "grad_norm": 2.844824887893612, + "learning_rate": 1.567345802931569e-06, + "loss": 0.0047, + "step": 6777 + }, + { + "epoch": 4.875382125516993, + "grad_norm": 1.6690620826364997, + "learning_rate": 1.5670058996088235e-06, + "loss": 0.0072, + "step": 6778 + }, + { + "epoch": 4.876101420607805, + "grad_norm": 2.1304999699797835, + "learning_rate": 1.5666659928385072e-06, + "loss": 0.0254, + "step": 6779 + }, + { + "epoch": 4.876820715698615, + "grad_norm": 5.605024202328447, + "learning_rate": 1.5663260826381081e-06, + "loss": 0.0835, + "step": 6780 + }, + { + "epoch": 4.877540010789426, + "grad_norm": 5.516250013263064, + "learning_rate": 1.565986169025116e-06, + "loss": 0.1662, + "step": 6781 + }, + { + "epoch": 4.8782593058802375, + "grad_norm": 2.306806329493952, + "learning_rate": 1.5656462520170196e-06, + "loss": 0.0084, + "step": 6782 + }, + { + "epoch": 4.878978600971048, + "grad_norm": 5.009063165486984, + "learning_rate": 1.5653063316313081e-06, + "loss": 0.1474, + "step": 6783 + }, + { + "epoch": 4.8796978960618596, + "grad_norm": 1.4391763759732037, + "learning_rate": 1.5649664078854716e-06, + "loss": 0.0284, + "step": 6784 + }, + { + "epoch": 4.88041719115267, + "grad_norm": 1.6230774475575134, + "learning_rate": 1.5646264807969989e-06, + "loss": 0.0309, + "step": 6785 + }, + { + "epoch": 4.881136486243482, + "grad_norm": 3.3265371848397125, + "learning_rate": 1.5642865503833806e-06, + "loss": 0.0118, + "step": 6786 + }, + { + "epoch": 4.881855781334292, + "grad_norm": 2.4153887620420083, + "learning_rate": 1.563946616662106e-06, + "loss": 0.0453, + "step": 6787 + }, + { + "epoch": 4.882575076425104, + "grad_norm": 3.1033658921779574, + "learning_rate": 1.5636066796506659e-06, + "loss": 0.0589, + "step": 6788 + }, + { + "epoch": 4.883294371515914, + "grad_norm": 2.749860961892385, + "learning_rate": 1.563266739366551e-06, + "loss": 0.075, + "step": 6789 + }, + { + "epoch": 4.884013666606726, + "grad_norm": 1.7531060659849267, + "learning_rate": 1.562926795827251e-06, + "loss": 0.0069, + "step": 6790 + }, + { + "epoch": 4.884732961697536, + "grad_norm": 3.8853106353944478, + "learning_rate": 1.5625868490502567e-06, + "loss": 0.0207, + "step": 6791 + }, + { + "epoch": 4.885452256788348, + "grad_norm": 2.608701012833949, + "learning_rate": 1.5622468990530592e-06, + "loss": 0.0422, + "step": 6792 + }, + { + "epoch": 4.886171551879158, + "grad_norm": 2.9395674952631303, + "learning_rate": 1.5619069458531495e-06, + "loss": 0.0149, + "step": 6793 + }, + { + "epoch": 4.886890846969969, + "grad_norm": 4.63299871978225, + "learning_rate": 1.561566989468019e-06, + "loss": 0.1554, + "step": 6794 + }, + { + "epoch": 4.8876101420607805, + "grad_norm": 1.1577120788477975, + "learning_rate": 1.5612270299151584e-06, + "loss": 0.0156, + "step": 6795 + }, + { + "epoch": 4.888329437151591, + "grad_norm": 2.0286762206620272, + "learning_rate": 1.5608870672120597e-06, + "loss": 0.0082, + "step": 6796 + }, + { + "epoch": 4.8890487322424026, + "grad_norm": 2.2177455426753556, + "learning_rate": 1.560547101376215e-06, + "loss": 0.0451, + "step": 6797 + }, + { + "epoch": 4.889768027333213, + "grad_norm": 0.10479035632150054, + "learning_rate": 1.5602071324251156e-06, + "loss": 0.0003, + "step": 6798 + }, + { + "epoch": 4.890487322424025, + "grad_norm": 0.4041101659945557, + "learning_rate": 1.5598671603762529e-06, + "loss": 0.0007, + "step": 6799 + }, + { + "epoch": 4.891206617514835, + "grad_norm": 0.09520209931675638, + "learning_rate": 1.5595271852471203e-06, + "loss": 0.0002, + "step": 6800 + }, + { + "epoch": 4.891925912605647, + "grad_norm": 1.2303424840449964, + "learning_rate": 1.559187207055209e-06, + "loss": 0.0204, + "step": 6801 + }, + { + "epoch": 4.892645207696457, + "grad_norm": 1.4036969305778464, + "learning_rate": 1.558847225818012e-06, + "loss": 0.0025, + "step": 6802 + }, + { + "epoch": 4.893364502787269, + "grad_norm": 0.5370353080442806, + "learning_rate": 1.5585072415530224e-06, + "loss": 0.0012, + "step": 6803 + }, + { + "epoch": 4.894083797878079, + "grad_norm": 2.9161157941880984, + "learning_rate": 1.5581672542777325e-06, + "loss": 0.0648, + "step": 6804 + }, + { + "epoch": 4.894803092968891, + "grad_norm": 0.021172740058380844, + "learning_rate": 1.5578272640096354e-06, + "loss": 0.0001, + "step": 6805 + }, + { + "epoch": 4.895522388059701, + "grad_norm": 3.0422042904667204, + "learning_rate": 1.5574872707662234e-06, + "loss": 0.0398, + "step": 6806 + }, + { + "epoch": 4.896241683150512, + "grad_norm": 0.006305823412490692, + "learning_rate": 1.5571472745649907e-06, + "loss": 0.0, + "step": 6807 + }, + { + "epoch": 4.8969609782413235, + "grad_norm": 1.2751153545328349, + "learning_rate": 1.5568072754234304e-06, + "loss": 0.0054, + "step": 6808 + }, + { + "epoch": 4.897680273332135, + "grad_norm": 2.471470351777139, + "learning_rate": 1.556467273359036e-06, + "loss": 0.0412, + "step": 6809 + }, + { + "epoch": 4.8983995684229455, + "grad_norm": 4.361595918649916, + "learning_rate": 1.5561272683893015e-06, + "loss": 0.1269, + "step": 6810 + }, + { + "epoch": 4.899118863513756, + "grad_norm": 1.7718177288138364, + "learning_rate": 1.5557872605317202e-06, + "loss": 0.0436, + "step": 6811 + }, + { + "epoch": 4.899838158604568, + "grad_norm": 0.606423212954094, + "learning_rate": 1.5554472498037868e-06, + "loss": 0.002, + "step": 6812 + }, + { + "epoch": 4.900557453695378, + "grad_norm": 2.0450345018638, + "learning_rate": 1.5551072362229947e-06, + "loss": 0.0384, + "step": 6813 + }, + { + "epoch": 4.90127674878619, + "grad_norm": 3.68452994553118, + "learning_rate": 1.5547672198068384e-06, + "loss": 0.0979, + "step": 6814 + }, + { + "epoch": 4.901996043877, + "grad_norm": 3.9341350904400167, + "learning_rate": 1.5544272005728131e-06, + "loss": 0.1466, + "step": 6815 + }, + { + "epoch": 4.902715338967812, + "grad_norm": 4.54455080837031, + "learning_rate": 1.5540871785384126e-06, + "loss": 0.1149, + "step": 6816 + }, + { + "epoch": 4.903434634058622, + "grad_norm": 2.995312840791445, + "learning_rate": 1.5537471537211316e-06, + "loss": 0.0687, + "step": 6817 + }, + { + "epoch": 4.904153929149434, + "grad_norm": 1.4438035512298597, + "learning_rate": 1.5534071261384652e-06, + "loss": 0.0241, + "step": 6818 + }, + { + "epoch": 4.904873224240244, + "grad_norm": 4.845698356701442, + "learning_rate": 1.5530670958079085e-06, + "loss": 0.091, + "step": 6819 + }, + { + "epoch": 4.905592519331056, + "grad_norm": 4.376790139665013, + "learning_rate": 1.5527270627469567e-06, + "loss": 0.0906, + "step": 6820 + }, + { + "epoch": 4.9063118144218665, + "grad_norm": 2.5921786328343255, + "learning_rate": 1.5523870269731048e-06, + "loss": 0.0476, + "step": 6821 + }, + { + "epoch": 4.907031109512678, + "grad_norm": 1.405431743259159, + "learning_rate": 1.5520469885038481e-06, + "loss": 0.0299, + "step": 6822 + }, + { + "epoch": 4.9077504046034885, + "grad_norm": 3.4606823228672217, + "learning_rate": 1.551706947356683e-06, + "loss": 0.0705, + "step": 6823 + }, + { + "epoch": 4.908469699694299, + "grad_norm": 0.9885039438600399, + "learning_rate": 1.5513669035491046e-06, + "loss": 0.0033, + "step": 6824 + }, + { + "epoch": 4.909188994785111, + "grad_norm": 0.8145226904131523, + "learning_rate": 1.5510268570986086e-06, + "loss": 0.0106, + "step": 6825 + }, + { + "epoch": 4.909908289875922, + "grad_norm": 3.191926080144873, + "learning_rate": 1.5506868080226912e-06, + "loss": 0.0607, + "step": 6826 + }, + { + "epoch": 4.910627584966733, + "grad_norm": 0.35714563508454894, + "learning_rate": 1.5503467563388492e-06, + "loss": 0.0019, + "step": 6827 + }, + { + "epoch": 4.911346880057543, + "grad_norm": 0.019216018125860414, + "learning_rate": 1.5500067020645775e-06, + "loss": 0.0001, + "step": 6828 + }, + { + "epoch": 4.912066175148355, + "grad_norm": 0.8906067954096623, + "learning_rate": 1.5496666452173736e-06, + "loss": 0.0097, + "step": 6829 + }, + { + "epoch": 4.912785470239165, + "grad_norm": 1.128598884431842, + "learning_rate": 1.5493265858147334e-06, + "loss": 0.024, + "step": 6830 + }, + { + "epoch": 4.913504765329977, + "grad_norm": 0.6936147175360826, + "learning_rate": 1.5489865238741537e-06, + "loss": 0.0008, + "step": 6831 + }, + { + "epoch": 4.914224060420787, + "grad_norm": 1.4392103251204083, + "learning_rate": 1.5486464594131314e-06, + "loss": 0.016, + "step": 6832 + }, + { + "epoch": 4.914943355511599, + "grad_norm": 1.1767084469197764, + "learning_rate": 1.5483063924491637e-06, + "loss": 0.0018, + "step": 6833 + }, + { + "epoch": 4.9156626506024095, + "grad_norm": 4.470761469104431, + "learning_rate": 1.547966322999747e-06, + "loss": 0.1454, + "step": 6834 + }, + { + "epoch": 4.916381945693221, + "grad_norm": 2.087639907476344, + "learning_rate": 1.5476262510823787e-06, + "loss": 0.0068, + "step": 6835 + }, + { + "epoch": 4.9171012407840315, + "grad_norm": 4.17006729547468, + "learning_rate": 1.5472861767145562e-06, + "loss": 0.0335, + "step": 6836 + }, + { + "epoch": 4.917820535874843, + "grad_norm": 4.745605228794661, + "learning_rate": 1.5469460999137772e-06, + "loss": 0.0687, + "step": 6837 + }, + { + "epoch": 4.918539830965654, + "grad_norm": 4.022028929577227, + "learning_rate": 1.5466060206975384e-06, + "loss": 0.1186, + "step": 6838 + }, + { + "epoch": 4.919259126056465, + "grad_norm": 0.21628886223224286, + "learning_rate": 1.5462659390833384e-06, + "loss": 0.0006, + "step": 6839 + }, + { + "epoch": 4.919978421147276, + "grad_norm": 1.4417220798797727, + "learning_rate": 1.5459258550886746e-06, + "loss": 0.033, + "step": 6840 + }, + { + "epoch": 4.920697716238086, + "grad_norm": 0.014093693898596977, + "learning_rate": 1.545585768731045e-06, + "loss": 0.0, + "step": 6841 + }, + { + "epoch": 4.921417011328898, + "grad_norm": 1.2158304217358737, + "learning_rate": 1.5452456800279476e-06, + "loss": 0.012, + "step": 6842 + }, + { + "epoch": 4.922136306419708, + "grad_norm": 0.11831387711133581, + "learning_rate": 1.54490558899688e-06, + "loss": 0.0002, + "step": 6843 + }, + { + "epoch": 4.92285560151052, + "grad_norm": 0.44473159100189796, + "learning_rate": 1.5445654956553416e-06, + "loss": 0.0009, + "step": 6844 + }, + { + "epoch": 4.92357489660133, + "grad_norm": 1.6040807511943787, + "learning_rate": 1.5442254000208302e-06, + "loss": 0.0167, + "step": 6845 + }, + { + "epoch": 4.924294191692142, + "grad_norm": 1.3872383785428901, + "learning_rate": 1.5438853021108438e-06, + "loss": 0.0082, + "step": 6846 + }, + { + "epoch": 4.9250134867829525, + "grad_norm": 1.5471976467366901, + "learning_rate": 1.5435452019428822e-06, + "loss": 0.0195, + "step": 6847 + }, + { + "epoch": 4.925732781873764, + "grad_norm": 0.09351558119133471, + "learning_rate": 1.5432050995344433e-06, + "loss": 0.0003, + "step": 6848 + }, + { + "epoch": 4.9264520769645745, + "grad_norm": 3.19573090777938, + "learning_rate": 1.5428649949030262e-06, + "loss": 0.0745, + "step": 6849 + }, + { + "epoch": 4.927171372055386, + "grad_norm": 3.6540825450450765, + "learning_rate": 1.5425248880661301e-06, + "loss": 0.109, + "step": 6850 + }, + { + "epoch": 4.927890667146197, + "grad_norm": 3.4470564608307317, + "learning_rate": 1.5421847790412532e-06, + "loss": 0.1044, + "step": 6851 + }, + { + "epoch": 4.928609962237008, + "grad_norm": 0.6407811520777118, + "learning_rate": 1.5418446678458957e-06, + "loss": 0.0055, + "step": 6852 + }, + { + "epoch": 4.929329257327819, + "grad_norm": 2.6038079008288917, + "learning_rate": 1.5415045544975569e-06, + "loss": 0.0525, + "step": 6853 + }, + { + "epoch": 4.930048552418629, + "grad_norm": 4.26794499287925, + "learning_rate": 1.541164439013735e-06, + "loss": 0.1229, + "step": 6854 + }, + { + "epoch": 4.930767847509441, + "grad_norm": 5.291478806056253, + "learning_rate": 1.5408243214119316e-06, + "loss": 0.121, + "step": 6855 + }, + { + "epoch": 4.931487142600252, + "grad_norm": 0.021231511876692224, + "learning_rate": 1.5404842017096447e-06, + "loss": 0.0001, + "step": 6856 + }, + { + "epoch": 4.932206437691063, + "grad_norm": 3.1083118483693415, + "learning_rate": 1.5401440799243745e-06, + "loss": 0.0562, + "step": 6857 + }, + { + "epoch": 4.932925732781873, + "grad_norm": 4.865171674515226, + "learning_rate": 1.539803956073621e-06, + "loss": 0.113, + "step": 6858 + }, + { + "epoch": 4.933645027872685, + "grad_norm": 3.5721351893045514, + "learning_rate": 1.539463830174884e-06, + "loss": 0.0669, + "step": 6859 + }, + { + "epoch": 4.9343643229634955, + "grad_norm": 3.7706037643711223, + "learning_rate": 1.5391237022456635e-06, + "loss": 0.0872, + "step": 6860 + }, + { + "epoch": 4.935083618054307, + "grad_norm": 2.1211413371014536, + "learning_rate": 1.53878357230346e-06, + "loss": 0.0401, + "step": 6861 + }, + { + "epoch": 4.9358029131451175, + "grad_norm": 0.05623793921511756, + "learning_rate": 1.5384434403657736e-06, + "loss": 0.0001, + "step": 6862 + }, + { + "epoch": 4.936522208235929, + "grad_norm": 1.8009525382859437, + "learning_rate": 1.5381033064501051e-06, + "loss": 0.0379, + "step": 6863 + }, + { + "epoch": 4.93724150332674, + "grad_norm": 1.6706229106815698, + "learning_rate": 1.5377631705739543e-06, + "loss": 0.03, + "step": 6864 + }, + { + "epoch": 4.937960798417551, + "grad_norm": 4.445991236280691, + "learning_rate": 1.537423032754822e-06, + "loss": 0.062, + "step": 6865 + }, + { + "epoch": 4.938680093508362, + "grad_norm": 0.6465238146317415, + "learning_rate": 1.5370828930102092e-06, + "loss": 0.0059, + "step": 6866 + }, + { + "epoch": 4.939399388599173, + "grad_norm": 0.4301890186135692, + "learning_rate": 1.536742751357616e-06, + "loss": 0.0016, + "step": 6867 + }, + { + "epoch": 4.940118683689984, + "grad_norm": 0.7296715944327, + "learning_rate": 1.536402607814544e-06, + "loss": 0.0079, + "step": 6868 + }, + { + "epoch": 4.940837978780795, + "grad_norm": 2.816003822748717, + "learning_rate": 1.5360624623984943e-06, + "loss": 0.0497, + "step": 6869 + }, + { + "epoch": 4.941557273871606, + "grad_norm": 3.5308067526895672, + "learning_rate": 1.5357223151269674e-06, + "loss": 0.0883, + "step": 6870 + }, + { + "epoch": 4.942276568962416, + "grad_norm": 3.6582913383736972, + "learning_rate": 1.5353821660174644e-06, + "loss": 0.0914, + "step": 6871 + }, + { + "epoch": 4.942995864053228, + "grad_norm": 2.103508500520179, + "learning_rate": 1.5350420150874866e-06, + "loss": 0.0451, + "step": 6872 + }, + { + "epoch": 4.9437151591440385, + "grad_norm": 3.1450086553375654, + "learning_rate": 1.534701862354536e-06, + "loss": 0.0482, + "step": 6873 + }, + { + "epoch": 4.94443445423485, + "grad_norm": 2.375756196359212, + "learning_rate": 1.5343617078361138e-06, + "loss": 0.0667, + "step": 6874 + }, + { + "epoch": 4.9451537493256605, + "grad_norm": 0.10482086444082286, + "learning_rate": 1.5340215515497212e-06, + "loss": 0.0004, + "step": 6875 + }, + { + "epoch": 4.945873044416472, + "grad_norm": 1.8792049349245135, + "learning_rate": 1.5336813935128602e-06, + "loss": 0.039, + "step": 6876 + }, + { + "epoch": 4.946592339507283, + "grad_norm": 3.7837883758675015, + "learning_rate": 1.5333412337430323e-06, + "loss": 0.0948, + "step": 6877 + }, + { + "epoch": 4.947311634598094, + "grad_norm": 1.8013284951434603, + "learning_rate": 1.5330010722577397e-06, + "loss": 0.0294, + "step": 6878 + }, + { + "epoch": 4.948030929688905, + "grad_norm": 0.013977343709042965, + "learning_rate": 1.5326609090744836e-06, + "loss": 0.0001, + "step": 6879 + }, + { + "epoch": 4.948750224779716, + "grad_norm": 0.015068628500523186, + "learning_rate": 1.5323207442107664e-06, + "loss": 0.0001, + "step": 6880 + }, + { + "epoch": 4.949469519870527, + "grad_norm": 0.82902885626921, + "learning_rate": 1.5319805776840903e-06, + "loss": 0.0015, + "step": 6881 + }, + { + "epoch": 4.950188814961338, + "grad_norm": 3.3959412050776914, + "learning_rate": 1.531640409511957e-06, + "loss": 0.0414, + "step": 6882 + }, + { + "epoch": 4.950908110052149, + "grad_norm": 6.543047135449496, + "learning_rate": 1.5313002397118693e-06, + "loss": 0.138, + "step": 6883 + }, + { + "epoch": 4.951627405142959, + "grad_norm": 3.6583792879896464, + "learning_rate": 1.5309600683013297e-06, + "loss": 0.0301, + "step": 6884 + }, + { + "epoch": 4.952346700233771, + "grad_norm": 5.159219433986411, + "learning_rate": 1.53061989529784e-06, + "loss": 0.0865, + "step": 6885 + }, + { + "epoch": 4.953065995324582, + "grad_norm": 4.737919140455368, + "learning_rate": 1.530279720718903e-06, + "loss": 0.1739, + "step": 6886 + }, + { + "epoch": 4.953785290415393, + "grad_norm": 0.22235651324933206, + "learning_rate": 1.529939544582021e-06, + "loss": 0.0006, + "step": 6887 + }, + { + "epoch": 4.9545045855062035, + "grad_norm": 2.5363079023554826, + "learning_rate": 1.529599366904697e-06, + "loss": 0.0399, + "step": 6888 + }, + { + "epoch": 4.955223880597015, + "grad_norm": 1.9424052859020458, + "learning_rate": 1.5292591877044335e-06, + "loss": 0.0371, + "step": 6889 + }, + { + "epoch": 4.955943175687826, + "grad_norm": 1.9305322056325551, + "learning_rate": 1.5289190069987332e-06, + "loss": 0.0275, + "step": 6890 + }, + { + "epoch": 4.956662470778637, + "grad_norm": 3.4836925917356765, + "learning_rate": 1.5285788248050996e-06, + "loss": 0.0434, + "step": 6891 + }, + { + "epoch": 4.957381765869448, + "grad_norm": 1.59818340794383, + "learning_rate": 1.5282386411410356e-06, + "loss": 0.0285, + "step": 6892 + }, + { + "epoch": 4.958101060960259, + "grad_norm": 4.8503395191243035, + "learning_rate": 1.5278984560240432e-06, + "loss": 0.112, + "step": 6893 + }, + { + "epoch": 4.95882035605107, + "grad_norm": 3.8866216651595136, + "learning_rate": 1.5275582694716268e-06, + "loss": 0.0987, + "step": 6894 + }, + { + "epoch": 4.959539651141881, + "grad_norm": 1.1130804230740916, + "learning_rate": 1.527218081501289e-06, + "loss": 0.0169, + "step": 6895 + }, + { + "epoch": 4.960258946232692, + "grad_norm": 2.102665920345144, + "learning_rate": 1.526877892130533e-06, + "loss": 0.0648, + "step": 6896 + }, + { + "epoch": 4.960978241323503, + "grad_norm": 4.249523907985297, + "learning_rate": 1.5265377013768623e-06, + "loss": 0.1292, + "step": 6897 + }, + { + "epoch": 4.961697536414314, + "grad_norm": 1.7565385154891466, + "learning_rate": 1.52619750925778e-06, + "loss": 0.0168, + "step": 6898 + }, + { + "epoch": 4.962416831505125, + "grad_norm": 1.6476834101054918, + "learning_rate": 1.5258573157907903e-06, + "loss": 0.0423, + "step": 6899 + }, + { + "epoch": 4.963136126595936, + "grad_norm": 2.5323289564870914, + "learning_rate": 1.5255171209933963e-06, + "loss": 0.0403, + "step": 6900 + }, + { + "epoch": 4.9638554216867465, + "grad_norm": 2.7741636905932676, + "learning_rate": 1.5251769248831013e-06, + "loss": 0.0495, + "step": 6901 + }, + { + "epoch": 4.964574716777558, + "grad_norm": 0.7939099805686383, + "learning_rate": 1.5248367274774097e-06, + "loss": 0.0036, + "step": 6902 + }, + { + "epoch": 4.9652940118683695, + "grad_norm": 7.669067167460385, + "learning_rate": 1.5244965287938241e-06, + "loss": 0.0332, + "step": 6903 + }, + { + "epoch": 4.96601330695918, + "grad_norm": 3.265073390465103, + "learning_rate": 1.5241563288498499e-06, + "loss": 0.0605, + "step": 6904 + }, + { + "epoch": 4.966732602049991, + "grad_norm": 0.12841773392909034, + "learning_rate": 1.5238161276629901e-06, + "loss": 0.0008, + "step": 6905 + }, + { + "epoch": 4.967451897140802, + "grad_norm": 1.560587669387678, + "learning_rate": 1.5234759252507486e-06, + "loss": 0.0395, + "step": 6906 + }, + { + "epoch": 4.968171192231613, + "grad_norm": 6.699382166070695, + "learning_rate": 1.5231357216306295e-06, + "loss": 0.1192, + "step": 6907 + }, + { + "epoch": 4.968890487322424, + "grad_norm": 3.1934131761666187, + "learning_rate": 1.5227955168201368e-06, + "loss": 0.0822, + "step": 6908 + }, + { + "epoch": 4.969609782413235, + "grad_norm": 0.354577060076399, + "learning_rate": 1.5224553108367749e-06, + "loss": 0.0009, + "step": 6909 + }, + { + "epoch": 4.970329077504046, + "grad_norm": 3.7019041394084446, + "learning_rate": 1.522115103698048e-06, + "loss": 0.0359, + "step": 6910 + }, + { + "epoch": 4.971048372594857, + "grad_norm": 2.428133843634875, + "learning_rate": 1.5217748954214593e-06, + "loss": 0.0211, + "step": 6911 + }, + { + "epoch": 4.971767667685668, + "grad_norm": 5.347104022252348, + "learning_rate": 1.521434686024515e-06, + "loss": 0.168, + "step": 6912 + }, + { + "epoch": 4.972486962776479, + "grad_norm": 3.578993508549843, + "learning_rate": 1.5210944755247182e-06, + "loss": 0.0901, + "step": 6913 + }, + { + "epoch": 4.97320625786729, + "grad_norm": 4.614501184593706, + "learning_rate": 1.5207542639395735e-06, + "loss": 0.1347, + "step": 6914 + }, + { + "epoch": 4.973925552958101, + "grad_norm": 2.08844913730273, + "learning_rate": 1.5204140512865857e-06, + "loss": 0.0149, + "step": 6915 + }, + { + "epoch": 4.9746448480489125, + "grad_norm": 2.92751586746207, + "learning_rate": 1.5200738375832592e-06, + "loss": 0.078, + "step": 6916 + }, + { + "epoch": 4.975364143139723, + "grad_norm": 0.5419971008328754, + "learning_rate": 1.5197336228470978e-06, + "loss": 0.0042, + "step": 6917 + }, + { + "epoch": 4.976083438230534, + "grad_norm": 2.9538319327882396, + "learning_rate": 1.5193934070956075e-06, + "loss": 0.0356, + "step": 6918 + }, + { + "epoch": 4.976802733321345, + "grad_norm": 6.578253276598189, + "learning_rate": 1.519053190346292e-06, + "loss": 0.1919, + "step": 6919 + }, + { + "epoch": 4.977522028412156, + "grad_norm": 0.49130172552389945, + "learning_rate": 1.5187129726166565e-06, + "loss": 0.0031, + "step": 6920 + }, + { + "epoch": 4.978241323502967, + "grad_norm": 5.9613324899580045, + "learning_rate": 1.5183727539242062e-06, + "loss": 0.0998, + "step": 6921 + }, + { + "epoch": 4.978960618593778, + "grad_norm": 3.4076711137137927, + "learning_rate": 1.518032534286445e-06, + "loss": 0.0604, + "step": 6922 + }, + { + "epoch": 4.979679913684589, + "grad_norm": 5.2752093847476775, + "learning_rate": 1.5176923137208787e-06, + "loss": 0.1419, + "step": 6923 + }, + { + "epoch": 4.9803992087754, + "grad_norm": 4.273125152073289, + "learning_rate": 1.517352092245011e-06, + "loss": 0.0498, + "step": 6924 + }, + { + "epoch": 4.981118503866211, + "grad_norm": 0.5411052420687827, + "learning_rate": 1.517011869876348e-06, + "loss": 0.0034, + "step": 6925 + }, + { + "epoch": 4.981837798957022, + "grad_norm": 6.923867124342787, + "learning_rate": 1.5166716466323947e-06, + "loss": 0.2091, + "step": 6926 + }, + { + "epoch": 4.982557094047833, + "grad_norm": 3.8323867575202533, + "learning_rate": 1.5163314225306558e-06, + "loss": 0.1297, + "step": 6927 + }, + { + "epoch": 4.983276389138644, + "grad_norm": 3.2506065964362856, + "learning_rate": 1.5159911975886368e-06, + "loss": 0.0658, + "step": 6928 + }, + { + "epoch": 4.9839956842294555, + "grad_norm": 4.723872117156507, + "learning_rate": 1.5156509718238423e-06, + "loss": 0.1375, + "step": 6929 + }, + { + "epoch": 4.984714979320266, + "grad_norm": 4.691252578741504, + "learning_rate": 1.5153107452537777e-06, + "loss": 0.0626, + "step": 6930 + }, + { + "epoch": 4.985434274411077, + "grad_norm": 5.29064719518989, + "learning_rate": 1.5149705178959487e-06, + "loss": 0.1432, + "step": 6931 + }, + { + "epoch": 4.986153569501888, + "grad_norm": 2.847715397734495, + "learning_rate": 1.5146302897678597e-06, + "loss": 0.0761, + "step": 6932 + }, + { + "epoch": 4.9868728645927, + "grad_norm": 2.6533257976625833, + "learning_rate": 1.5142900608870171e-06, + "loss": 0.0553, + "step": 6933 + }, + { + "epoch": 4.98759215968351, + "grad_norm": 1.6387824075135526, + "learning_rate": 1.5139498312709259e-06, + "loss": 0.0317, + "step": 6934 + }, + { + "epoch": 4.988311454774321, + "grad_norm": 2.7261466230295386, + "learning_rate": 1.5136096009370913e-06, + "loss": 0.0367, + "step": 6935 + }, + { + "epoch": 4.989030749865132, + "grad_norm": 0.3128878981333429, + "learning_rate": 1.5132693699030188e-06, + "loss": 0.0013, + "step": 6936 + }, + { + "epoch": 4.989750044955943, + "grad_norm": 2.963570099177867, + "learning_rate": 1.5129291381862134e-06, + "loss": 0.0546, + "step": 6937 + }, + { + "epoch": 4.990469340046754, + "grad_norm": 2.9431270469603614, + "learning_rate": 1.512588905804182e-06, + "loss": 0.0627, + "step": 6938 + }, + { + "epoch": 4.991188635137565, + "grad_norm": 3.325245101266144, + "learning_rate": 1.512248672774429e-06, + "loss": 0.0394, + "step": 6939 + }, + { + "epoch": 4.991907930228376, + "grad_norm": 1.7149744987059847, + "learning_rate": 1.5119084391144599e-06, + "loss": 0.0046, + "step": 6940 + }, + { + "epoch": 4.992627225319187, + "grad_norm": 2.844712210136346, + "learning_rate": 1.511568204841781e-06, + "loss": 0.0306, + "step": 6941 + }, + { + "epoch": 4.9933465204099985, + "grad_norm": 0.23153926956581566, + "learning_rate": 1.5112279699738977e-06, + "loss": 0.0007, + "step": 6942 + }, + { + "epoch": 4.994065815500809, + "grad_norm": 3.3271337219637553, + "learning_rate": 1.5108877345283156e-06, + "loss": 0.0751, + "step": 6943 + }, + { + "epoch": 4.994785110591621, + "grad_norm": 0.9894677743179612, + "learning_rate": 1.5105474985225407e-06, + "loss": 0.0097, + "step": 6944 + }, + { + "epoch": 4.995504405682431, + "grad_norm": 1.990112343955673, + "learning_rate": 1.510207261974078e-06, + "loss": 0.0703, + "step": 6945 + }, + { + "epoch": 4.996223700773243, + "grad_norm": 2.4344867081083894, + "learning_rate": 1.509867024900434e-06, + "loss": 0.0436, + "step": 6946 + }, + { + "epoch": 4.996942995864053, + "grad_norm": 0.4497056004414728, + "learning_rate": 1.5095267873191141e-06, + "loss": 0.0016, + "step": 6947 + }, + { + "epoch": 4.997662290954864, + "grad_norm": 0.15019978953055568, + "learning_rate": 1.509186549247624e-06, + "loss": 0.0002, + "step": 6948 + }, + { + "epoch": 4.998381586045675, + "grad_norm": 3.0390218462884624, + "learning_rate": 1.5088463107034705e-06, + "loss": 0.0931, + "step": 6949 + }, + { + "epoch": 4.999100881136486, + "grad_norm": 1.204731139332293, + "learning_rate": 1.5085060717041585e-06, + "loss": 0.0023, + "step": 6950 + }, + { + "epoch": 4.999820176227297, + "grad_norm": 2.380233033007584, + "learning_rate": 1.5081658322671944e-06, + "loss": 0.0448, + "step": 6951 + }, + { + "epoch": 5.000539471318108, + "grad_norm": 2.267254459147145, + "learning_rate": 1.5078255924100837e-06, + "loss": 0.0524, + "step": 6952 + }, + { + "epoch": 5.001258766408919, + "grad_norm": 0.01435540542976982, + "learning_rate": 1.5074853521503322e-06, + "loss": 0.0, + "step": 6953 + }, + { + "epoch": 5.00197806149973, + "grad_norm": 2.7854498493384257, + "learning_rate": 1.5071451115054464e-06, + "loss": 0.0643, + "step": 6954 + }, + { + "epoch": 5.0026973565905415, + "grad_norm": 1.5630510273106055, + "learning_rate": 1.5068048704929323e-06, + "loss": 0.0386, + "step": 6955 + }, + { + "epoch": 5.003416651681352, + "grad_norm": 2.6410703196290926, + "learning_rate": 1.5064646291302955e-06, + "loss": 0.0287, + "step": 6956 + }, + { + "epoch": 5.004135946772164, + "grad_norm": 3.560652224326323, + "learning_rate": 1.5061243874350422e-06, + "loss": 0.0231, + "step": 6957 + }, + { + "epoch": 5.004855241862974, + "grad_norm": 2.4723191797849804, + "learning_rate": 1.5057841454246784e-06, + "loss": 0.0589, + "step": 6958 + }, + { + "epoch": 5.005574536953786, + "grad_norm": 0.1952447908934331, + "learning_rate": 1.5054439031167103e-06, + "loss": 0.0006, + "step": 6959 + }, + { + "epoch": 5.006293832044596, + "grad_norm": 2.6189836499445462, + "learning_rate": 1.5051036605286436e-06, + "loss": 0.0485, + "step": 6960 + }, + { + "epoch": 5.007013127135408, + "grad_norm": 2.0331964369289004, + "learning_rate": 1.5047634176779846e-06, + "loss": 0.0437, + "step": 6961 + }, + { + "epoch": 5.007732422226218, + "grad_norm": 2.887945786766406, + "learning_rate": 1.5044231745822398e-06, + "loss": 0.0502, + "step": 6962 + }, + { + "epoch": 5.008451717317029, + "grad_norm": 1.6373058951924067, + "learning_rate": 1.5040829312589146e-06, + "loss": 0.0251, + "step": 6963 + }, + { + "epoch": 5.00917101240784, + "grad_norm": 2.3659059312321267, + "learning_rate": 1.5037426877255153e-06, + "loss": 0.0641, + "step": 6964 + }, + { + "epoch": 5.009890307498651, + "grad_norm": 1.4368556590719734, + "learning_rate": 1.5034024439995484e-06, + "loss": 0.0181, + "step": 6965 + }, + { + "epoch": 5.010609602589462, + "grad_norm": 0.6779271626435611, + "learning_rate": 1.5030622000985194e-06, + "loss": 0.007, + "step": 6966 + }, + { + "epoch": 5.011328897680273, + "grad_norm": 1.455814307083035, + "learning_rate": 1.5027219560399353e-06, + "loss": 0.0225, + "step": 6967 + }, + { + "epoch": 5.0120481927710845, + "grad_norm": 1.9357316390718564, + "learning_rate": 1.5023817118413018e-06, + "loss": 0.0255, + "step": 6968 + }, + { + "epoch": 5.012767487861895, + "grad_norm": 2.4549707658299327, + "learning_rate": 1.5020414675201245e-06, + "loss": 0.0398, + "step": 6969 + }, + { + "epoch": 5.0134867829527066, + "grad_norm": 5.679339483180636, + "learning_rate": 1.5017012230939104e-06, + "loss": 0.0447, + "step": 6970 + }, + { + "epoch": 5.014206078043517, + "grad_norm": 2.726237769128921, + "learning_rate": 1.5013609785801655e-06, + "loss": 0.0692, + "step": 6971 + }, + { + "epoch": 5.014925373134329, + "grad_norm": 0.774184672405304, + "learning_rate": 1.501020733996396e-06, + "loss": 0.0051, + "step": 6972 + }, + { + "epoch": 5.015644668225139, + "grad_norm": 0.014045320207681415, + "learning_rate": 1.500680489360108e-06, + "loss": 0.0001, + "step": 6973 + }, + { + "epoch": 5.016363963315951, + "grad_norm": 3.6348352149903302, + "learning_rate": 1.500340244688807e-06, + "loss": 0.0689, + "step": 6974 + }, + { + "epoch": 5.017083258406761, + "grad_norm": 1.894262044439209, + "learning_rate": 1.5e-06, + "loss": 0.0328, + "step": 6975 + }, + { + "epoch": 5.017802553497573, + "grad_norm": 2.8206312056582674, + "learning_rate": 1.4996597553111934e-06, + "loss": 0.0611, + "step": 6976 + }, + { + "epoch": 5.018521848588383, + "grad_norm": 3.382917913031027, + "learning_rate": 1.4993195106398928e-06, + "loss": 0.0364, + "step": 6977 + }, + { + "epoch": 5.019241143679195, + "grad_norm": 4.656130478671204, + "learning_rate": 1.4989792660036043e-06, + "loss": 0.0728, + "step": 6978 + }, + { + "epoch": 5.019960438770005, + "grad_norm": 0.3604638003849488, + "learning_rate": 1.4986390214198348e-06, + "loss": 0.0026, + "step": 6979 + }, + { + "epoch": 5.020679733860816, + "grad_norm": 4.3657680708086115, + "learning_rate": 1.4982987769060897e-06, + "loss": 0.0866, + "step": 6980 + }, + { + "epoch": 5.0213990289516275, + "grad_norm": 1.0875182534509769, + "learning_rate": 1.4979585324798757e-06, + "loss": 0.0089, + "step": 6981 + }, + { + "epoch": 5.022118324042438, + "grad_norm": 3.05704347554018, + "learning_rate": 1.497618288158699e-06, + "loss": 0.0632, + "step": 6982 + }, + { + "epoch": 5.0228376191332496, + "grad_norm": 1.497784974411615, + "learning_rate": 1.4972780439600645e-06, + "loss": 0.0049, + "step": 6983 + }, + { + "epoch": 5.02355691422406, + "grad_norm": 2.0341273785158545, + "learning_rate": 1.4969377999014805e-06, + "loss": 0.0131, + "step": 6984 + }, + { + "epoch": 5.024276209314872, + "grad_norm": 2.2148977745094185, + "learning_rate": 1.4965975560004517e-06, + "loss": 0.0048, + "step": 6985 + }, + { + "epoch": 5.024995504405682, + "grad_norm": 1.736879508382241, + "learning_rate": 1.496257312274485e-06, + "loss": 0.0248, + "step": 6986 + }, + { + "epoch": 5.025714799496494, + "grad_norm": 0.02261042373341707, + "learning_rate": 1.495917068741086e-06, + "loss": 0.0001, + "step": 6987 + }, + { + "epoch": 5.026434094587304, + "grad_norm": 2.332362902779914, + "learning_rate": 1.495576825417761e-06, + "loss": 0.0424, + "step": 6988 + }, + { + "epoch": 5.027153389678116, + "grad_norm": 3.6808881418334707, + "learning_rate": 1.4952365823220153e-06, + "loss": 0.0148, + "step": 6989 + }, + { + "epoch": 5.027872684768926, + "grad_norm": 0.008258489747737111, + "learning_rate": 1.4948963394713565e-06, + "loss": 0.0, + "step": 6990 + }, + { + "epoch": 5.028591979859738, + "grad_norm": 3.39321372490868, + "learning_rate": 1.49455609688329e-06, + "loss": 0.052, + "step": 6991 + }, + { + "epoch": 5.029311274950548, + "grad_norm": 0.008063404137574053, + "learning_rate": 1.4942158545753219e-06, + "loss": 0.0, + "step": 6992 + }, + { + "epoch": 5.03003057004136, + "grad_norm": 4.390947403936277, + "learning_rate": 1.4938756125649581e-06, + "loss": 0.0933, + "step": 6993 + }, + { + "epoch": 5.0307498651321705, + "grad_norm": 1.567814128710308, + "learning_rate": 1.4935353708697046e-06, + "loss": 0.0093, + "step": 6994 + }, + { + "epoch": 5.031469160222981, + "grad_norm": 5.057168189155978, + "learning_rate": 1.4931951295070678e-06, + "loss": 0.0835, + "step": 6995 + }, + { + "epoch": 5.0321884553137926, + "grad_norm": 1.075565212927923, + "learning_rate": 1.4928548884945537e-06, + "loss": 0.0063, + "step": 6996 + }, + { + "epoch": 5.032907750404603, + "grad_norm": 7.008734985598111, + "learning_rate": 1.492514647849668e-06, + "loss": 0.1877, + "step": 6997 + }, + { + "epoch": 5.033627045495415, + "grad_norm": 0.16857828199451916, + "learning_rate": 1.4921744075899168e-06, + "loss": 0.001, + "step": 6998 + }, + { + "epoch": 5.034346340586225, + "grad_norm": 0.026468428945644478, + "learning_rate": 1.4918341677328057e-06, + "loss": 0.0001, + "step": 6999 + }, + { + "epoch": 5.035065635677037, + "grad_norm": 5.886509764964726, + "learning_rate": 1.4914939282958417e-06, + "loss": 0.089, + "step": 7000 + }, + { + "epoch": 5.035784930767847, + "grad_norm": 4.136083540208941, + "learning_rate": 1.4911536892965298e-06, + "loss": 0.0716, + "step": 7001 + }, + { + "epoch": 5.036504225858659, + "grad_norm": 3.227100459356839, + "learning_rate": 1.490813450752376e-06, + "loss": 0.0438, + "step": 7002 + }, + { + "epoch": 5.037223520949469, + "grad_norm": 1.570293649173586, + "learning_rate": 1.4904732126808864e-06, + "loss": 0.0325, + "step": 7003 + }, + { + "epoch": 5.037942816040281, + "grad_norm": 1.8826060064133368, + "learning_rate": 1.490132975099566e-06, + "loss": 0.0336, + "step": 7004 + }, + { + "epoch": 5.038662111131091, + "grad_norm": 1.2101741513211464, + "learning_rate": 1.489792738025922e-06, + "loss": 0.0095, + "step": 7005 + }, + { + "epoch": 5.039381406221903, + "grad_norm": 3.6743500159758615, + "learning_rate": 1.4894525014774596e-06, + "loss": 0.0684, + "step": 7006 + }, + { + "epoch": 5.0401007013127135, + "grad_norm": 2.5977768472786322, + "learning_rate": 1.4891122654716845e-06, + "loss": 0.0633, + "step": 7007 + }, + { + "epoch": 5.040819996403525, + "grad_norm": 1.9637021062606859, + "learning_rate": 1.4887720300261024e-06, + "loss": 0.0325, + "step": 7008 + }, + { + "epoch": 5.0415392914943356, + "grad_norm": 2.606004135637719, + "learning_rate": 1.4884317951582192e-06, + "loss": 0.0254, + "step": 7009 + }, + { + "epoch": 5.042258586585146, + "grad_norm": 0.4152013694502947, + "learning_rate": 1.4880915608855402e-06, + "loss": 0.0007, + "step": 7010 + }, + { + "epoch": 5.042977881675958, + "grad_norm": 1.9080765064230647, + "learning_rate": 1.4877513272255713e-06, + "loss": 0.0212, + "step": 7011 + }, + { + "epoch": 5.043697176766768, + "grad_norm": 0.27286148755465844, + "learning_rate": 1.4874110941958185e-06, + "loss": 0.0011, + "step": 7012 + }, + { + "epoch": 5.04441647185758, + "grad_norm": 2.3772030866142075, + "learning_rate": 1.4870708618137867e-06, + "loss": 0.0473, + "step": 7013 + }, + { + "epoch": 5.04513576694839, + "grad_norm": 1.5515666133939856, + "learning_rate": 1.4867306300969817e-06, + "loss": 0.0278, + "step": 7014 + }, + { + "epoch": 5.045855062039202, + "grad_norm": 1.4745378811602783, + "learning_rate": 1.486390399062909e-06, + "loss": 0.0216, + "step": 7015 + }, + { + "epoch": 5.046574357130012, + "grad_norm": 5.570176517611281, + "learning_rate": 1.4860501687290744e-06, + "loss": 0.064, + "step": 7016 + }, + { + "epoch": 5.047293652220824, + "grad_norm": 4.1468453286546145, + "learning_rate": 1.4857099391129832e-06, + "loss": 0.0886, + "step": 7017 + }, + { + "epoch": 5.048012947311634, + "grad_norm": 0.1120097405400623, + "learning_rate": 1.4853697102321404e-06, + "loss": 0.0002, + "step": 7018 + }, + { + "epoch": 5.048732242402446, + "grad_norm": 3.3721540481919003, + "learning_rate": 1.4850294821040518e-06, + "loss": 0.0276, + "step": 7019 + }, + { + "epoch": 5.0494515374932565, + "grad_norm": 2.6209145903699995, + "learning_rate": 1.4846892547462224e-06, + "loss": 0.0421, + "step": 7020 + }, + { + "epoch": 5.050170832584068, + "grad_norm": 2.642128687866523, + "learning_rate": 1.484349028176158e-06, + "loss": 0.037, + "step": 7021 + }, + { + "epoch": 5.0508901276748785, + "grad_norm": 1.6603327613054624, + "learning_rate": 1.4840088024113635e-06, + "loss": 0.0182, + "step": 7022 + }, + { + "epoch": 5.05160942276569, + "grad_norm": 0.8923992158505517, + "learning_rate": 1.4836685774693447e-06, + "loss": 0.0108, + "step": 7023 + }, + { + "epoch": 5.052328717856501, + "grad_norm": 1.2289972657921946, + "learning_rate": 1.4833283533676058e-06, + "loss": 0.0096, + "step": 7024 + }, + { + "epoch": 5.053048012947311, + "grad_norm": 1.5243551078154565, + "learning_rate": 1.4829881301236517e-06, + "loss": 0.0171, + "step": 7025 + }, + { + "epoch": 5.053767308038123, + "grad_norm": 1.285697397112331, + "learning_rate": 1.482647907754989e-06, + "loss": 0.0022, + "step": 7026 + }, + { + "epoch": 5.054486603128933, + "grad_norm": 7.1722983039312425, + "learning_rate": 1.4823076862791218e-06, + "loss": 0.1071, + "step": 7027 + }, + { + "epoch": 5.055205898219745, + "grad_norm": 0.019765823150338233, + "learning_rate": 1.4819674657135554e-06, + "loss": 0.0001, + "step": 7028 + }, + { + "epoch": 5.055925193310555, + "grad_norm": 1.6243923278286518, + "learning_rate": 1.4816272460757943e-06, + "loss": 0.0082, + "step": 7029 + }, + { + "epoch": 5.056644488401367, + "grad_norm": 3.2150993192320665, + "learning_rate": 1.4812870273833436e-06, + "loss": 0.0541, + "step": 7030 + }, + { + "epoch": 5.057363783492177, + "grad_norm": 2.9023233157908788, + "learning_rate": 1.480946809653708e-06, + "loss": 0.0456, + "step": 7031 + }, + { + "epoch": 5.058083078582989, + "grad_norm": 3.931152715414752, + "learning_rate": 1.4806065929043928e-06, + "loss": 0.0378, + "step": 7032 + }, + { + "epoch": 5.0588023736737995, + "grad_norm": 0.009590555448437187, + "learning_rate": 1.4802663771529023e-06, + "loss": 0.0, + "step": 7033 + }, + { + "epoch": 5.059521668764611, + "grad_norm": 1.2895138827915276, + "learning_rate": 1.4799261624167415e-06, + "loss": 0.0168, + "step": 7034 + }, + { + "epoch": 5.0602409638554215, + "grad_norm": 3.4554724769780965, + "learning_rate": 1.4795859487134148e-06, + "loss": 0.0745, + "step": 7035 + }, + { + "epoch": 5.060960258946233, + "grad_norm": 4.831121311932715, + "learning_rate": 1.4792457360604268e-06, + "loss": 0.0592, + "step": 7036 + }, + { + "epoch": 5.061679554037044, + "grad_norm": 2.961641675105508, + "learning_rate": 1.4789055244752821e-06, + "loss": 0.0484, + "step": 7037 + }, + { + "epoch": 5.062398849127855, + "grad_norm": 3.4822168931103943, + "learning_rate": 1.4785653139754855e-06, + "loss": 0.0646, + "step": 7038 + }, + { + "epoch": 5.063118144218666, + "grad_norm": 0.05146292546644698, + "learning_rate": 1.4782251045785408e-06, + "loss": 0.0002, + "step": 7039 + }, + { + "epoch": 5.063837439309477, + "grad_norm": 3.3065010207587195, + "learning_rate": 1.477884896301953e-06, + "loss": 0.0388, + "step": 7040 + }, + { + "epoch": 5.064556734400288, + "grad_norm": 2.7914641902583748, + "learning_rate": 1.4775446891632252e-06, + "loss": 0.0497, + "step": 7041 + }, + { + "epoch": 5.065276029491098, + "grad_norm": 3.626127559632273, + "learning_rate": 1.4772044831798633e-06, + "loss": 0.0829, + "step": 7042 + }, + { + "epoch": 5.06599532458191, + "grad_norm": 1.034894280981372, + "learning_rate": 1.4768642783693708e-06, + "loss": 0.007, + "step": 7043 + }, + { + "epoch": 5.06671461967272, + "grad_norm": 2.8593898553284505, + "learning_rate": 1.4765240747492517e-06, + "loss": 0.0299, + "step": 7044 + }, + { + "epoch": 5.067433914763532, + "grad_norm": 3.804247395264018, + "learning_rate": 1.4761838723370102e-06, + "loss": 0.0826, + "step": 7045 + }, + { + "epoch": 5.0681532098543425, + "grad_norm": 3.281943276288062, + "learning_rate": 1.47584367115015e-06, + "loss": 0.0825, + "step": 7046 + }, + { + "epoch": 5.068872504945154, + "grad_norm": 0.24215584760407233, + "learning_rate": 1.4755034712061757e-06, + "loss": 0.0005, + "step": 7047 + }, + { + "epoch": 5.0695918000359645, + "grad_norm": 0.40346704817525686, + "learning_rate": 1.4751632725225908e-06, + "loss": 0.0014, + "step": 7048 + }, + { + "epoch": 5.070311095126776, + "grad_norm": 1.2741532503466162, + "learning_rate": 1.4748230751168988e-06, + "loss": 0.0193, + "step": 7049 + }, + { + "epoch": 5.071030390217587, + "grad_norm": 1.2037760401861202, + "learning_rate": 1.4744828790066042e-06, + "loss": 0.013, + "step": 7050 + }, + { + "epoch": 5.071749685308398, + "grad_norm": 0.9004766785618685, + "learning_rate": 1.47414268420921e-06, + "loss": 0.0007, + "step": 7051 + }, + { + "epoch": 5.072468980399209, + "grad_norm": 3.583247213680404, + "learning_rate": 1.4738024907422201e-06, + "loss": 0.0715, + "step": 7052 + }, + { + "epoch": 5.07318827549002, + "grad_norm": 1.7035436921219935, + "learning_rate": 1.473462298623138e-06, + "loss": 0.0157, + "step": 7053 + }, + { + "epoch": 5.073907570580831, + "grad_norm": 3.2461348830691334, + "learning_rate": 1.4731221078694674e-06, + "loss": 0.0454, + "step": 7054 + }, + { + "epoch": 5.074626865671641, + "grad_norm": 5.062411197411337, + "learning_rate": 1.4727819184987116e-06, + "loss": 0.0908, + "step": 7055 + }, + { + "epoch": 5.075346160762453, + "grad_norm": 0.6604260851046734, + "learning_rate": 1.4724417305283737e-06, + "loss": 0.0041, + "step": 7056 + }, + { + "epoch": 5.076065455853263, + "grad_norm": 2.6231347706906525, + "learning_rate": 1.4721015439759565e-06, + "loss": 0.0344, + "step": 7057 + }, + { + "epoch": 5.076784750944075, + "grad_norm": 1.9866585767101637, + "learning_rate": 1.4717613588589645e-06, + "loss": 0.0137, + "step": 7058 + }, + { + "epoch": 5.0775040460348855, + "grad_norm": 3.551028592552635, + "learning_rate": 1.4714211751949006e-06, + "loss": 0.0267, + "step": 7059 + }, + { + "epoch": 5.078223341125697, + "grad_norm": 3.0267883168748497, + "learning_rate": 1.4710809930012671e-06, + "loss": 0.0593, + "step": 7060 + }, + { + "epoch": 5.0789426362165075, + "grad_norm": 2.748074542923553, + "learning_rate": 1.4707408122955673e-06, + "loss": 0.0651, + "step": 7061 + }, + { + "epoch": 5.079661931307319, + "grad_norm": 1.1281701090007539, + "learning_rate": 1.470400633095303e-06, + "loss": 0.0056, + "step": 7062 + }, + { + "epoch": 5.08038122639813, + "grad_norm": 6.6490185275997735, + "learning_rate": 1.4700604554179791e-06, + "loss": 0.1723, + "step": 7063 + }, + { + "epoch": 5.081100521488941, + "grad_norm": 2.7471562731494386, + "learning_rate": 1.4697202792810975e-06, + "loss": 0.0276, + "step": 7064 + }, + { + "epoch": 5.081819816579752, + "grad_norm": 0.03965276487233567, + "learning_rate": 1.4693801047021603e-06, + "loss": 0.0001, + "step": 7065 + }, + { + "epoch": 5.082539111670563, + "grad_norm": 2.0637790048802533, + "learning_rate": 1.4690399316986706e-06, + "loss": 0.0034, + "step": 7066 + }, + { + "epoch": 5.083258406761374, + "grad_norm": 1.9073480564378755, + "learning_rate": 1.4686997602881305e-06, + "loss": 0.028, + "step": 7067 + }, + { + "epoch": 5.083977701852185, + "grad_norm": 2.775500314146843, + "learning_rate": 1.468359590488043e-06, + "loss": 0.0335, + "step": 7068 + }, + { + "epoch": 5.084696996942996, + "grad_norm": 4.712073263826327, + "learning_rate": 1.4680194223159102e-06, + "loss": 0.0742, + "step": 7069 + }, + { + "epoch": 5.085416292033807, + "grad_norm": 5.204237486818348, + "learning_rate": 1.467679255789234e-06, + "loss": 0.114, + "step": 7070 + }, + { + "epoch": 5.086135587124618, + "grad_norm": 1.597006800907216, + "learning_rate": 1.467339090925517e-06, + "loss": 0.0126, + "step": 7071 + }, + { + "epoch": 5.0868548822154285, + "grad_norm": 3.160525371188695, + "learning_rate": 1.4669989277422608e-06, + "loss": 0.0279, + "step": 7072 + }, + { + "epoch": 5.08757417730624, + "grad_norm": 0.5117282530087784, + "learning_rate": 1.4666587662569678e-06, + "loss": 0.0015, + "step": 7073 + }, + { + "epoch": 5.0882934723970505, + "grad_norm": 0.10657452470694416, + "learning_rate": 1.46631860648714e-06, + "loss": 0.0005, + "step": 7074 + }, + { + "epoch": 5.089012767487862, + "grad_norm": 4.1229395503173665, + "learning_rate": 1.465978448450279e-06, + "loss": 0.0472, + "step": 7075 + }, + { + "epoch": 5.089732062578673, + "grad_norm": 2.6067315294798976, + "learning_rate": 1.4656382921638865e-06, + "loss": 0.0495, + "step": 7076 + }, + { + "epoch": 5.090451357669484, + "grad_norm": 1.1701063874747382, + "learning_rate": 1.4652981376454642e-06, + "loss": 0.015, + "step": 7077 + }, + { + "epoch": 5.091170652760295, + "grad_norm": 1.4568572856624313, + "learning_rate": 1.464957984912513e-06, + "loss": 0.03, + "step": 7078 + }, + { + "epoch": 5.091889947851106, + "grad_norm": 0.13021231629414803, + "learning_rate": 1.464617833982536e-06, + "loss": 0.0005, + "step": 7079 + }, + { + "epoch": 5.092609242941917, + "grad_norm": 4.368900002837487, + "learning_rate": 1.464277684873033e-06, + "loss": 0.0629, + "step": 7080 + }, + { + "epoch": 5.093328538032728, + "grad_norm": 2.1033750990382667, + "learning_rate": 1.463937537601506e-06, + "loss": 0.0323, + "step": 7081 + }, + { + "epoch": 5.094047833123539, + "grad_norm": 4.325439243877494, + "learning_rate": 1.4635973921854565e-06, + "loss": 0.0915, + "step": 7082 + }, + { + "epoch": 5.09476712821435, + "grad_norm": 3.822989948984249, + "learning_rate": 1.4632572486423838e-06, + "loss": 0.0795, + "step": 7083 + }, + { + "epoch": 5.095486423305161, + "grad_norm": 1.0237786640097009, + "learning_rate": 1.462917106989791e-06, + "loss": 0.0017, + "step": 7084 + }, + { + "epoch": 5.096205718395972, + "grad_norm": 1.356174239338846, + "learning_rate": 1.462576967245178e-06, + "loss": 0.0079, + "step": 7085 + }, + { + "epoch": 5.096925013486783, + "grad_norm": 2.6908393707884577, + "learning_rate": 1.462236829426046e-06, + "loss": 0.0343, + "step": 7086 + }, + { + "epoch": 5.0976443085775935, + "grad_norm": 0.8894534544567909, + "learning_rate": 1.4618966935498952e-06, + "loss": 0.0082, + "step": 7087 + }, + { + "epoch": 5.098363603668405, + "grad_norm": 2.9587166616721325, + "learning_rate": 1.4615565596342263e-06, + "loss": 0.0258, + "step": 7088 + }, + { + "epoch": 5.099082898759216, + "grad_norm": 1.1257984906335152, + "learning_rate": 1.4612164276965398e-06, + "loss": 0.0044, + "step": 7089 + }, + { + "epoch": 5.099802193850027, + "grad_norm": 3.069298143192928, + "learning_rate": 1.4608762977543366e-06, + "loss": 0.0603, + "step": 7090 + }, + { + "epoch": 5.100521488940838, + "grad_norm": 2.6214662466109897, + "learning_rate": 1.4605361698251164e-06, + "loss": 0.0404, + "step": 7091 + }, + { + "epoch": 5.101240784031649, + "grad_norm": 1.7232643837024697, + "learning_rate": 1.4601960439263793e-06, + "loss": 0.022, + "step": 7092 + }, + { + "epoch": 5.10196007912246, + "grad_norm": 0.25472706446810084, + "learning_rate": 1.4598559200756254e-06, + "loss": 0.0019, + "step": 7093 + }, + { + "epoch": 5.102679374213271, + "grad_norm": 2.377732178466331, + "learning_rate": 1.4595157982903552e-06, + "loss": 0.032, + "step": 7094 + }, + { + "epoch": 5.103398669304082, + "grad_norm": 4.056201496952641, + "learning_rate": 1.4591756785880687e-06, + "loss": 0.0189, + "step": 7095 + }, + { + "epoch": 5.104117964394893, + "grad_norm": 0.03708984494105273, + "learning_rate": 1.458835560986265e-06, + "loss": 0.0001, + "step": 7096 + }, + { + "epoch": 5.104837259485704, + "grad_norm": 3.0742965619942, + "learning_rate": 1.4584954455024438e-06, + "loss": 0.0545, + "step": 7097 + }, + { + "epoch": 5.105556554576515, + "grad_norm": 3.658115828954782, + "learning_rate": 1.4581553321541048e-06, + "loss": 0.0422, + "step": 7098 + }, + { + "epoch": 5.106275849667326, + "grad_norm": 1.459179348539147, + "learning_rate": 1.4578152209587468e-06, + "loss": 0.0199, + "step": 7099 + }, + { + "epoch": 5.106995144758137, + "grad_norm": 0.3357075892883036, + "learning_rate": 1.4574751119338702e-06, + "loss": 0.0019, + "step": 7100 + }, + { + "epoch": 5.107714439848948, + "grad_norm": 1.0325186829961464, + "learning_rate": 1.457135005096974e-06, + "loss": 0.0205, + "step": 7101 + }, + { + "epoch": 5.108433734939759, + "grad_norm": 0.5659235778379595, + "learning_rate": 1.456794900465557e-06, + "loss": 0.0046, + "step": 7102 + }, + { + "epoch": 5.10915303003057, + "grad_norm": 6.076199957256797, + "learning_rate": 1.456454798057118e-06, + "loss": 0.1675, + "step": 7103 + }, + { + "epoch": 5.109872325121381, + "grad_norm": 2.3160519066840584, + "learning_rate": 1.4561146978891559e-06, + "loss": 0.0377, + "step": 7104 + }, + { + "epoch": 5.110591620212192, + "grad_norm": 1.100580428761019, + "learning_rate": 1.45577459997917e-06, + "loss": 0.0084, + "step": 7105 + }, + { + "epoch": 5.111310915303003, + "grad_norm": 3.6164294843404634, + "learning_rate": 1.4554345043446586e-06, + "loss": 0.0452, + "step": 7106 + }, + { + "epoch": 5.112030210393814, + "grad_norm": 0.5131341006230276, + "learning_rate": 1.4550944110031203e-06, + "loss": 0.0006, + "step": 7107 + }, + { + "epoch": 5.112749505484625, + "grad_norm": 5.7479289974262535, + "learning_rate": 1.454754319972053e-06, + "loss": 0.0869, + "step": 7108 + }, + { + "epoch": 5.113468800575436, + "grad_norm": 2.700003967537174, + "learning_rate": 1.4544142312689553e-06, + "loss": 0.0438, + "step": 7109 + }, + { + "epoch": 5.114188095666247, + "grad_norm": 2.148360191199612, + "learning_rate": 1.4540741449113255e-06, + "loss": 0.0296, + "step": 7110 + }, + { + "epoch": 5.114907390757058, + "grad_norm": 1.6836504391897729, + "learning_rate": 1.4537340609166619e-06, + "loss": 0.0314, + "step": 7111 + }, + { + "epoch": 5.115626685847869, + "grad_norm": 3.476639679789797, + "learning_rate": 1.453393979302462e-06, + "loss": 0.0504, + "step": 7112 + }, + { + "epoch": 5.11634598093868, + "grad_norm": 2.4527627469002433, + "learning_rate": 1.4530539000862235e-06, + "loss": 0.0282, + "step": 7113 + }, + { + "epoch": 5.117065276029491, + "grad_norm": 3.305672985008273, + "learning_rate": 1.4527138232854437e-06, + "loss": 0.0242, + "step": 7114 + }, + { + "epoch": 5.1177845711203025, + "grad_norm": 2.9301726994597614, + "learning_rate": 1.4523737489176214e-06, + "loss": 0.0059, + "step": 7115 + }, + { + "epoch": 5.118503866211113, + "grad_norm": 6.39430179269705, + "learning_rate": 1.4520336770002534e-06, + "loss": 0.0538, + "step": 7116 + }, + { + "epoch": 5.119223161301925, + "grad_norm": 1.4899924051804005, + "learning_rate": 1.4516936075508368e-06, + "loss": 0.0196, + "step": 7117 + }, + { + "epoch": 5.119942456392735, + "grad_norm": 1.380829611351476, + "learning_rate": 1.4513535405868691e-06, + "loss": 0.0178, + "step": 7118 + }, + { + "epoch": 5.120661751483546, + "grad_norm": 3.939571313043218, + "learning_rate": 1.451013476125847e-06, + "loss": 0.0354, + "step": 7119 + }, + { + "epoch": 5.121381046574357, + "grad_norm": 1.0981739840508575, + "learning_rate": 1.4506734141852667e-06, + "loss": 0.0035, + "step": 7120 + }, + { + "epoch": 5.122100341665168, + "grad_norm": 2.7964137852430357, + "learning_rate": 1.4503333547826267e-06, + "loss": 0.03, + "step": 7121 + }, + { + "epoch": 5.122819636755979, + "grad_norm": 3.1255444409518964, + "learning_rate": 1.4499932979354228e-06, + "loss": 0.0895, + "step": 7122 + }, + { + "epoch": 5.12353893184679, + "grad_norm": 5.669732971629444, + "learning_rate": 1.4496532436611513e-06, + "loss": 0.1374, + "step": 7123 + }, + { + "epoch": 5.124258226937601, + "grad_norm": 4.215046214241308, + "learning_rate": 1.4493131919773089e-06, + "loss": 0.0671, + "step": 7124 + }, + { + "epoch": 5.124977522028412, + "grad_norm": 4.266358035281217, + "learning_rate": 1.4489731429013915e-06, + "loss": 0.0748, + "step": 7125 + }, + { + "epoch": 5.125696817119223, + "grad_norm": 0.6428896439566887, + "learning_rate": 1.4486330964508955e-06, + "loss": 0.0047, + "step": 7126 + }, + { + "epoch": 5.126416112210034, + "grad_norm": 8.414129809296195, + "learning_rate": 1.448293052643317e-06, + "loss": 0.1058, + "step": 7127 + }, + { + "epoch": 5.1271354073008455, + "grad_norm": 3.3184404639120686, + "learning_rate": 1.447953011496152e-06, + "loss": 0.0511, + "step": 7128 + }, + { + "epoch": 5.127854702391656, + "grad_norm": 1.6891596454361857, + "learning_rate": 1.4476129730268957e-06, + "loss": 0.0198, + "step": 7129 + }, + { + "epoch": 5.128573997482468, + "grad_norm": 1.6128589887260774, + "learning_rate": 1.4472729372530433e-06, + "loss": 0.0063, + "step": 7130 + }, + { + "epoch": 5.129293292573278, + "grad_norm": 2.010860642485462, + "learning_rate": 1.4469329041920918e-06, + "loss": 0.0221, + "step": 7131 + }, + { + "epoch": 5.130012587664089, + "grad_norm": 5.956197795233131, + "learning_rate": 1.4465928738615351e-06, + "loss": 0.122, + "step": 7132 + }, + { + "epoch": 5.1307318827549, + "grad_norm": 3.216652165842596, + "learning_rate": 1.446252846278869e-06, + "loss": 0.0627, + "step": 7133 + }, + { + "epoch": 5.131451177845711, + "grad_norm": 3.5441822480314147, + "learning_rate": 1.445912821461588e-06, + "loss": 0.0364, + "step": 7134 + }, + { + "epoch": 5.132170472936522, + "grad_norm": 6.295289978519549, + "learning_rate": 1.445572799427187e-06, + "loss": 0.1605, + "step": 7135 + }, + { + "epoch": 5.132889768027333, + "grad_norm": 0.6129352726491087, + "learning_rate": 1.4452327801931613e-06, + "loss": 0.0008, + "step": 7136 + }, + { + "epoch": 5.133609063118144, + "grad_norm": 3.5797402375657326, + "learning_rate": 1.4448927637770054e-06, + "loss": 0.0626, + "step": 7137 + }, + { + "epoch": 5.134328358208955, + "grad_norm": 2.613409625839271, + "learning_rate": 1.4445527501962135e-06, + "loss": 0.0305, + "step": 7138 + }, + { + "epoch": 5.135047653299766, + "grad_norm": 0.029485900240609056, + "learning_rate": 1.44421273946828e-06, + "loss": 0.0001, + "step": 7139 + }, + { + "epoch": 5.135766948390577, + "grad_norm": 2.0917183517695412, + "learning_rate": 1.443872731610699e-06, + "loss": 0.0231, + "step": 7140 + }, + { + "epoch": 5.1364862434813885, + "grad_norm": 0.6928133758431365, + "learning_rate": 1.443532726640964e-06, + "loss": 0.0083, + "step": 7141 + }, + { + "epoch": 5.137205538572199, + "grad_norm": 2.6744687468516526, + "learning_rate": 1.4431927245765697e-06, + "loss": 0.0318, + "step": 7142 + }, + { + "epoch": 5.137924833663011, + "grad_norm": 7.5566629458437165, + "learning_rate": 1.4428527254350098e-06, + "loss": 0.0778, + "step": 7143 + }, + { + "epoch": 5.138644128753821, + "grad_norm": 0.7100712383345502, + "learning_rate": 1.4425127292337771e-06, + "loss": 0.0048, + "step": 7144 + }, + { + "epoch": 5.139363423844633, + "grad_norm": 0.04400691260608603, + "learning_rate": 1.4421727359903656e-06, + "loss": 0.0001, + "step": 7145 + }, + { + "epoch": 5.140082718935443, + "grad_norm": 4.932640718146009, + "learning_rate": 1.441832745722268e-06, + "loss": 0.0142, + "step": 7146 + }, + { + "epoch": 5.140802014026255, + "grad_norm": 0.9449596510347695, + "learning_rate": 1.4414927584469777e-06, + "loss": 0.0042, + "step": 7147 + }, + { + "epoch": 5.141521309117065, + "grad_norm": 2.5406653066565177, + "learning_rate": 1.441152774181988e-06, + "loss": 0.0263, + "step": 7148 + }, + { + "epoch": 5.142240604207876, + "grad_norm": 2.896095735800872, + "learning_rate": 1.4408127929447914e-06, + "loss": 0.0553, + "step": 7149 + }, + { + "epoch": 5.142959899298687, + "grad_norm": 1.1600150331437333, + "learning_rate": 1.4404728147528804e-06, + "loss": 0.0093, + "step": 7150 + }, + { + "epoch": 5.143679194389498, + "grad_norm": 3.5862202357932675, + "learning_rate": 1.440132839623747e-06, + "loss": 0.063, + "step": 7151 + }, + { + "epoch": 5.144398489480309, + "grad_norm": 1.76065264224485, + "learning_rate": 1.4397928675748849e-06, + "loss": 0.0309, + "step": 7152 + }, + { + "epoch": 5.14511778457112, + "grad_norm": 2.5410614337103477, + "learning_rate": 1.439452898623785e-06, + "loss": 0.0364, + "step": 7153 + }, + { + "epoch": 5.1458370796619315, + "grad_norm": 2.826891253565452, + "learning_rate": 1.4391129327879406e-06, + "loss": 0.0481, + "step": 7154 + }, + { + "epoch": 5.146556374752742, + "grad_norm": 1.7339694638781733, + "learning_rate": 1.438772970084842e-06, + "loss": 0.032, + "step": 7155 + }, + { + "epoch": 5.147275669843554, + "grad_norm": 0.8383850294409325, + "learning_rate": 1.4384330105319811e-06, + "loss": 0.0079, + "step": 7156 + }, + { + "epoch": 5.147994964934364, + "grad_norm": 1.31069109258662, + "learning_rate": 1.4380930541468504e-06, + "loss": 0.0044, + "step": 7157 + }, + { + "epoch": 5.148714260025176, + "grad_norm": 2.355349592980351, + "learning_rate": 1.4377531009469408e-06, + "loss": 0.026, + "step": 7158 + }, + { + "epoch": 5.149433555115986, + "grad_norm": 5.980449302371705, + "learning_rate": 1.4374131509497434e-06, + "loss": 0.095, + "step": 7159 + }, + { + "epoch": 5.150152850206798, + "grad_norm": 4.826054033042511, + "learning_rate": 1.4370732041727494e-06, + "loss": 0.1186, + "step": 7160 + }, + { + "epoch": 5.150872145297608, + "grad_norm": 2.9093728114902957, + "learning_rate": 1.4367332606334495e-06, + "loss": 0.0505, + "step": 7161 + }, + { + "epoch": 5.15159144038842, + "grad_norm": 3.2019725002856725, + "learning_rate": 1.4363933203493338e-06, + "loss": 0.0287, + "step": 7162 + }, + { + "epoch": 5.15231073547923, + "grad_norm": 2.936169213361136, + "learning_rate": 1.436053383337894e-06, + "loss": 0.0456, + "step": 7163 + }, + { + "epoch": 5.153030030570041, + "grad_norm": 2.924062140091857, + "learning_rate": 1.43571344961662e-06, + "loss": 0.0476, + "step": 7164 + }, + { + "epoch": 5.153749325660852, + "grad_norm": 2.813945134371245, + "learning_rate": 1.4353735192030014e-06, + "loss": 0.0424, + "step": 7165 + }, + { + "epoch": 5.154468620751663, + "grad_norm": 4.097911578832152, + "learning_rate": 1.4350335921145291e-06, + "loss": 0.0821, + "step": 7166 + }, + { + "epoch": 5.1551879158424745, + "grad_norm": 0.4358178001578339, + "learning_rate": 1.434693668368692e-06, + "loss": 0.0056, + "step": 7167 + }, + { + "epoch": 5.155907210933285, + "grad_norm": 0.7876615415333963, + "learning_rate": 1.4343537479829807e-06, + "loss": 0.0043, + "step": 7168 + }, + { + "epoch": 5.156626506024097, + "grad_norm": 3.6314138819341246, + "learning_rate": 1.4340138309748845e-06, + "loss": 0.0325, + "step": 7169 + }, + { + "epoch": 5.157345801114907, + "grad_norm": 1.667859770902186, + "learning_rate": 1.4336739173618921e-06, + "loss": 0.0266, + "step": 7170 + }, + { + "epoch": 5.158065096205719, + "grad_norm": 3.559695781049528, + "learning_rate": 1.4333340071614933e-06, + "loss": 0.0919, + "step": 7171 + }, + { + "epoch": 5.158784391296529, + "grad_norm": 1.869034767735519, + "learning_rate": 1.4329941003911762e-06, + "loss": 0.0329, + "step": 7172 + }, + { + "epoch": 5.159503686387341, + "grad_norm": 0.0155359761066338, + "learning_rate": 1.4326541970684312e-06, + "loss": 0.0, + "step": 7173 + }, + { + "epoch": 5.160222981478151, + "grad_norm": 4.703830635688674, + "learning_rate": 1.4323142972107457e-06, + "loss": 0.0317, + "step": 7174 + }, + { + "epoch": 5.160942276568963, + "grad_norm": 8.578889328518157, + "learning_rate": 1.4319744008356085e-06, + "loss": 0.1012, + "step": 7175 + }, + { + "epoch": 5.161661571659773, + "grad_norm": 0.8754264425062591, + "learning_rate": 1.4316345079605077e-06, + "loss": 0.0074, + "step": 7176 + }, + { + "epoch": 5.162380866750585, + "grad_norm": 2.0087011814053035, + "learning_rate": 1.431294618602931e-06, + "loss": 0.0339, + "step": 7177 + }, + { + "epoch": 5.163100161841395, + "grad_norm": 0.33913496271771804, + "learning_rate": 1.4309547327803675e-06, + "loss": 0.001, + "step": 7178 + }, + { + "epoch": 5.163819456932206, + "grad_norm": 3.354413613541638, + "learning_rate": 1.430614850510304e-06, + "loss": 0.0852, + "step": 7179 + }, + { + "epoch": 5.1645387520230175, + "grad_norm": 3.871688569961813, + "learning_rate": 1.4302749718102281e-06, + "loss": 0.0447, + "step": 7180 + }, + { + "epoch": 5.165258047113828, + "grad_norm": 4.391231933137038, + "learning_rate": 1.4299350966976277e-06, + "loss": 0.0337, + "step": 7181 + }, + { + "epoch": 5.1659773422046396, + "grad_norm": 5.168584566957976, + "learning_rate": 1.429595225189989e-06, + "loss": 0.0434, + "step": 7182 + }, + { + "epoch": 5.16669663729545, + "grad_norm": 3.084353819245071, + "learning_rate": 1.4292553573047997e-06, + "loss": 0.0552, + "step": 7183 + }, + { + "epoch": 5.167415932386262, + "grad_norm": 2.154199626033384, + "learning_rate": 1.4289154930595466e-06, + "loss": 0.0208, + "step": 7184 + }, + { + "epoch": 5.168135227477072, + "grad_norm": 1.7586056093665672, + "learning_rate": 1.4285756324717163e-06, + "loss": 0.036, + "step": 7185 + }, + { + "epoch": 5.168854522567884, + "grad_norm": 6.274515909796575, + "learning_rate": 1.4282357755587948e-06, + "loss": 0.1192, + "step": 7186 + }, + { + "epoch": 5.169573817658694, + "grad_norm": 0.44841270219740786, + "learning_rate": 1.427895922338269e-06, + "loss": 0.001, + "step": 7187 + }, + { + "epoch": 5.170293112749506, + "grad_norm": 1.4756422667524427, + "learning_rate": 1.4275560728276236e-06, + "loss": 0.0102, + "step": 7188 + }, + { + "epoch": 5.171012407840316, + "grad_norm": 0.5697262965376998, + "learning_rate": 1.4272162270443464e-06, + "loss": 0.0032, + "step": 7189 + }, + { + "epoch": 5.171731702931128, + "grad_norm": 6.136678498828114, + "learning_rate": 1.426876385005922e-06, + "loss": 0.1217, + "step": 7190 + }, + { + "epoch": 5.172450998021938, + "grad_norm": 6.22084010536719, + "learning_rate": 1.4265365467298358e-06, + "loss": 0.1246, + "step": 7191 + }, + { + "epoch": 5.17317029311275, + "grad_norm": 4.5648197304523785, + "learning_rate": 1.4261967122335735e-06, + "loss": 0.1055, + "step": 7192 + }, + { + "epoch": 5.1738895882035605, + "grad_norm": 5.337441616035062, + "learning_rate": 1.425856881534619e-06, + "loss": 0.0485, + "step": 7193 + }, + { + "epoch": 5.174608883294372, + "grad_norm": 2.2130824023019056, + "learning_rate": 1.4255170546504588e-06, + "loss": 0.0453, + "step": 7194 + }, + { + "epoch": 5.1753281783851826, + "grad_norm": 3.565728315666002, + "learning_rate": 1.4251772315985768e-06, + "loss": 0.0771, + "step": 7195 + }, + { + "epoch": 5.176047473475993, + "grad_norm": 1.9551091836821965, + "learning_rate": 1.4248374123964577e-06, + "loss": 0.0073, + "step": 7196 + }, + { + "epoch": 5.176766768566805, + "grad_norm": 0.7035823662961502, + "learning_rate": 1.4244975970615856e-06, + "loss": 0.0102, + "step": 7197 + }, + { + "epoch": 5.177486063657615, + "grad_norm": 2.416668565231576, + "learning_rate": 1.4241577856114444e-06, + "loss": 0.028, + "step": 7198 + }, + { + "epoch": 5.178205358748427, + "grad_norm": 4.278193670372044, + "learning_rate": 1.4238179780635186e-06, + "loss": 0.0757, + "step": 7199 + }, + { + "epoch": 5.178924653839237, + "grad_norm": 3.64493004058581, + "learning_rate": 1.4234781744352915e-06, + "loss": 0.0559, + "step": 7200 + }, + { + "epoch": 5.179643948930049, + "grad_norm": 3.1987854522173835, + "learning_rate": 1.4231383747442468e-06, + "loss": 0.0616, + "step": 7201 + }, + { + "epoch": 5.180363244020859, + "grad_norm": 2.2495439247717077, + "learning_rate": 1.4227985790078674e-06, + "loss": 0.0451, + "step": 7202 + }, + { + "epoch": 5.181082539111671, + "grad_norm": 2.382791635572287, + "learning_rate": 1.4224587872436368e-06, + "loss": 0.0335, + "step": 7203 + }, + { + "epoch": 5.181801834202481, + "grad_norm": 2.353497620067541, + "learning_rate": 1.4221189994690374e-06, + "loss": 0.0243, + "step": 7204 + }, + { + "epoch": 5.182521129293293, + "grad_norm": 1.0221062326201884, + "learning_rate": 1.4217792157015525e-06, + "loss": 0.0048, + "step": 7205 + }, + { + "epoch": 5.1832404243841035, + "grad_norm": 0.34865766091341693, + "learning_rate": 1.4214394359586643e-06, + "loss": 0.0011, + "step": 7206 + }, + { + "epoch": 5.183959719474915, + "grad_norm": 4.080356282147448, + "learning_rate": 1.421099660257855e-06, + "loss": 0.0654, + "step": 7207 + }, + { + "epoch": 5.1846790145657256, + "grad_norm": 0.01987476226232673, + "learning_rate": 1.4207598886166067e-06, + "loss": 0.0001, + "step": 7208 + }, + { + "epoch": 5.185398309656536, + "grad_norm": 1.9978217054562848, + "learning_rate": 1.4204201210524004e-06, + "loss": 0.0305, + "step": 7209 + }, + { + "epoch": 5.186117604747348, + "grad_norm": 1.6964806635167067, + "learning_rate": 1.4200803575827192e-06, + "loss": 0.0072, + "step": 7210 + }, + { + "epoch": 5.186836899838158, + "grad_norm": 1.7506996373088237, + "learning_rate": 1.4197405982250444e-06, + "loss": 0.0181, + "step": 7211 + }, + { + "epoch": 5.18755619492897, + "grad_norm": 2.6536578350977997, + "learning_rate": 1.4194008429968561e-06, + "loss": 0.0409, + "step": 7212 + }, + { + "epoch": 5.18827549001978, + "grad_norm": 0.1707261990837736, + "learning_rate": 1.4190610919156366e-06, + "loss": 0.0004, + "step": 7213 + }, + { + "epoch": 5.188994785110592, + "grad_norm": 1.0975445260621872, + "learning_rate": 1.4187213449988649e-06, + "loss": 0.0106, + "step": 7214 + }, + { + "epoch": 5.189714080201402, + "grad_norm": 2.271903300140239, + "learning_rate": 1.4183816022640236e-06, + "loss": 0.0315, + "step": 7215 + }, + { + "epoch": 5.190433375292214, + "grad_norm": 3.589342614081717, + "learning_rate": 1.4180418637285918e-06, + "loss": 0.0481, + "step": 7216 + }, + { + "epoch": 5.191152670383024, + "grad_norm": 0.2339473496847589, + "learning_rate": 1.4177021294100501e-06, + "loss": 0.0008, + "step": 7217 + }, + { + "epoch": 5.191871965473836, + "grad_norm": 0.15844253237006414, + "learning_rate": 1.4173623993258782e-06, + "loss": 0.0006, + "step": 7218 + }, + { + "epoch": 5.1925912605646465, + "grad_norm": 0.16687298672408776, + "learning_rate": 1.4170226734935559e-06, + "loss": 0.0004, + "step": 7219 + }, + { + "epoch": 5.193310555655458, + "grad_norm": 3.162021658057057, + "learning_rate": 1.4166829519305627e-06, + "loss": 0.0795, + "step": 7220 + }, + { + "epoch": 5.1940298507462686, + "grad_norm": 3.076908804021646, + "learning_rate": 1.416343234654378e-06, + "loss": 0.0077, + "step": 7221 + }, + { + "epoch": 5.19474914583708, + "grad_norm": 4.669690127997388, + "learning_rate": 1.4160035216824807e-06, + "loss": 0.0955, + "step": 7222 + }, + { + "epoch": 5.195468440927891, + "grad_norm": 4.2039755750266545, + "learning_rate": 1.4156638130323499e-06, + "loss": 0.0584, + "step": 7223 + }, + { + "epoch": 5.196187736018702, + "grad_norm": 3.3904332450186883, + "learning_rate": 1.4153241087214637e-06, + "loss": 0.0832, + "step": 7224 + }, + { + "epoch": 5.196907031109513, + "grad_norm": 2.5958552089894007, + "learning_rate": 1.4149844087673004e-06, + "loss": 0.0319, + "step": 7225 + }, + { + "epoch": 5.197626326200323, + "grad_norm": 3.280972915256917, + "learning_rate": 1.414644713187339e-06, + "loss": 0.0426, + "step": 7226 + }, + { + "epoch": 5.198345621291135, + "grad_norm": 3.584878152237225, + "learning_rate": 1.414305021999057e-06, + "loss": 0.0838, + "step": 7227 + }, + { + "epoch": 5.199064916381945, + "grad_norm": 2.611463612279108, + "learning_rate": 1.4139653352199318e-06, + "loss": 0.0411, + "step": 7228 + }, + { + "epoch": 5.199784211472757, + "grad_norm": 1.006402112701204, + "learning_rate": 1.4136256528674413e-06, + "loss": 0.0082, + "step": 7229 + }, + { + "epoch": 5.200503506563567, + "grad_norm": 1.5432634220833203, + "learning_rate": 1.413285974959062e-06, + "loss": 0.0117, + "step": 7230 + }, + { + "epoch": 5.201222801654379, + "grad_norm": 3.1623849002224795, + "learning_rate": 1.412946301512272e-06, + "loss": 0.095, + "step": 7231 + }, + { + "epoch": 5.2019420967451895, + "grad_norm": 0.025549759508927315, + "learning_rate": 1.4126066325445475e-06, + "loss": 0.0001, + "step": 7232 + }, + { + "epoch": 5.202661391836001, + "grad_norm": 3.332272499676463, + "learning_rate": 1.4122669680733654e-06, + "loss": 0.0745, + "step": 7233 + }, + { + "epoch": 5.2033806869268115, + "grad_norm": 1.0678971449879893, + "learning_rate": 1.4119273081162016e-06, + "loss": 0.0106, + "step": 7234 + }, + { + "epoch": 5.204099982017623, + "grad_norm": 3.0294974096147427, + "learning_rate": 1.4115876526905324e-06, + "loss": 0.0568, + "step": 7235 + }, + { + "epoch": 5.204819277108434, + "grad_norm": 2.5203868892299766, + "learning_rate": 1.4112480018138338e-06, + "loss": 0.0541, + "step": 7236 + }, + { + "epoch": 5.205538572199245, + "grad_norm": 1.3481361717180005, + "learning_rate": 1.4109083555035815e-06, + "loss": 0.0151, + "step": 7237 + }, + { + "epoch": 5.206257867290056, + "grad_norm": 3.4022645812018903, + "learning_rate": 1.4105687137772504e-06, + "loss": 0.042, + "step": 7238 + }, + { + "epoch": 5.206977162380867, + "grad_norm": 3.159206220168551, + "learning_rate": 1.4102290766523163e-06, + "loss": 0.0669, + "step": 7239 + }, + { + "epoch": 5.207696457471678, + "grad_norm": 1.650210702517907, + "learning_rate": 1.409889444146253e-06, + "loss": 0.0164, + "step": 7240 + }, + { + "epoch": 5.208415752562488, + "grad_norm": 1.7971135946428314, + "learning_rate": 1.4095498162765367e-06, + "loss": 0.0101, + "step": 7241 + }, + { + "epoch": 5.2091350476533, + "grad_norm": 1.4006550760366963, + "learning_rate": 1.409210193060641e-06, + "loss": 0.0068, + "step": 7242 + }, + { + "epoch": 5.20985434274411, + "grad_norm": 1.0701062820188385, + "learning_rate": 1.4088705745160402e-06, + "loss": 0.0174, + "step": 7243 + }, + { + "epoch": 5.210573637834922, + "grad_norm": 2.1581534526633988, + "learning_rate": 1.4085309606602084e-06, + "loss": 0.0372, + "step": 7244 + }, + { + "epoch": 5.2112929329257325, + "grad_norm": 3.847573286689069, + "learning_rate": 1.4081913515106186e-06, + "loss": 0.049, + "step": 7245 + }, + { + "epoch": 5.212012228016544, + "grad_norm": 0.9475948338039364, + "learning_rate": 1.4078517470847453e-06, + "loss": 0.0066, + "step": 7246 + }, + { + "epoch": 5.2127315231073545, + "grad_norm": 2.0061817168487153, + "learning_rate": 1.4075121474000617e-06, + "loss": 0.0034, + "step": 7247 + }, + { + "epoch": 5.213450818198166, + "grad_norm": 2.1960734389475722, + "learning_rate": 1.4071725524740403e-06, + "loss": 0.0276, + "step": 7248 + }, + { + "epoch": 5.214170113288977, + "grad_norm": 9.873915777504358, + "learning_rate": 1.4068329623241544e-06, + "loss": 0.0736, + "step": 7249 + }, + { + "epoch": 5.214889408379788, + "grad_norm": 2.1847932203943334, + "learning_rate": 1.4064933769678759e-06, + "loss": 0.0427, + "step": 7250 + }, + { + "epoch": 5.215608703470599, + "grad_norm": 7.12503668431323, + "learning_rate": 1.4061537964226765e-06, + "loss": 0.1117, + "step": 7251 + }, + { + "epoch": 5.21632799856141, + "grad_norm": 5.43639709805573, + "learning_rate": 1.40581422070603e-06, + "loss": 0.1019, + "step": 7252 + }, + { + "epoch": 5.217047293652221, + "grad_norm": 3.003617697370464, + "learning_rate": 1.405474649835407e-06, + "loss": 0.0614, + "step": 7253 + }, + { + "epoch": 5.217766588743032, + "grad_norm": 2.823414172640912, + "learning_rate": 1.405135083828279e-06, + "loss": 0.0572, + "step": 7254 + }, + { + "epoch": 5.218485883833843, + "grad_norm": 3.228911142852904, + "learning_rate": 1.4047955227021176e-06, + "loss": 0.0553, + "step": 7255 + }, + { + "epoch": 5.219205178924653, + "grad_norm": 3.0185998168962627, + "learning_rate": 1.4044559664743935e-06, + "loss": 0.0462, + "step": 7256 + }, + { + "epoch": 5.219924474015465, + "grad_norm": 2.336173185741833, + "learning_rate": 1.4041164151625778e-06, + "loss": 0.0253, + "step": 7257 + }, + { + "epoch": 5.2206437691062755, + "grad_norm": 6.1760570245109445, + "learning_rate": 1.403776868784141e-06, + "loss": 0.0422, + "step": 7258 + }, + { + "epoch": 5.221363064197087, + "grad_norm": 6.806435880609753, + "learning_rate": 1.403437327356553e-06, + "loss": 0.1298, + "step": 7259 + }, + { + "epoch": 5.2220823592878975, + "grad_norm": 1.1299634567965575, + "learning_rate": 1.4030977908972842e-06, + "loss": 0.0062, + "step": 7260 + }, + { + "epoch": 5.222801654378709, + "grad_norm": 1.7094617941886807, + "learning_rate": 1.4027582594238037e-06, + "loss": 0.0087, + "step": 7261 + }, + { + "epoch": 5.22352094946952, + "grad_norm": 5.16734664550083, + "learning_rate": 1.4024187329535817e-06, + "loss": 0.1108, + "step": 7262 + }, + { + "epoch": 5.224240244560331, + "grad_norm": 1.6631170039162453, + "learning_rate": 1.4020792115040875e-06, + "loss": 0.0222, + "step": 7263 + }, + { + "epoch": 5.224959539651142, + "grad_norm": 2.483967164910691, + "learning_rate": 1.4017396950927897e-06, + "loss": 0.0351, + "step": 7264 + }, + { + "epoch": 5.225678834741953, + "grad_norm": 2.037607587279582, + "learning_rate": 1.4014001837371569e-06, + "loss": 0.0066, + "step": 7265 + }, + { + "epoch": 5.226398129832764, + "grad_norm": 0.21872939380650266, + "learning_rate": 1.401060677454657e-06, + "loss": 0.0012, + "step": 7266 + }, + { + "epoch": 5.227117424923575, + "grad_norm": 4.134077166460227, + "learning_rate": 1.40072117626276e-06, + "loss": 0.055, + "step": 7267 + }, + { + "epoch": 5.227836720014386, + "grad_norm": 5.630271533279505, + "learning_rate": 1.4003816801789323e-06, + "loss": 0.1015, + "step": 7268 + }, + { + "epoch": 5.228556015105197, + "grad_norm": 1.994115059338867, + "learning_rate": 1.4000421892206423e-06, + "loss": 0.0305, + "step": 7269 + }, + { + "epoch": 5.229275310196008, + "grad_norm": 2.36189846185108, + "learning_rate": 1.3997027034053572e-06, + "loss": 0.0384, + "step": 7270 + }, + { + "epoch": 5.229994605286819, + "grad_norm": 0.17073308769698906, + "learning_rate": 1.3993632227505437e-06, + "loss": 0.0004, + "step": 7271 + }, + { + "epoch": 5.23071390037763, + "grad_norm": 3.4453964034552103, + "learning_rate": 1.399023747273669e-06, + "loss": 0.0168, + "step": 7272 + }, + { + "epoch": 5.2314331954684405, + "grad_norm": 0.7080265151552322, + "learning_rate": 1.3986842769922002e-06, + "loss": 0.0036, + "step": 7273 + }, + { + "epoch": 5.232152490559252, + "grad_norm": 0.7939378553936798, + "learning_rate": 1.398344811923603e-06, + "loss": 0.0016, + "step": 7274 + }, + { + "epoch": 5.232871785650063, + "grad_norm": 0.10295085954673631, + "learning_rate": 1.3980053520853438e-06, + "loss": 0.0002, + "step": 7275 + }, + { + "epoch": 5.233591080740874, + "grad_norm": 3.102920023580569, + "learning_rate": 1.3976658974948881e-06, + "loss": 0.0358, + "step": 7276 + }, + { + "epoch": 5.234310375831685, + "grad_norm": 0.6336349555085077, + "learning_rate": 1.3973264481697015e-06, + "loss": 0.001, + "step": 7277 + }, + { + "epoch": 5.235029670922496, + "grad_norm": 4.716387675350153, + "learning_rate": 1.3969870041272496e-06, + "loss": 0.062, + "step": 7278 + }, + { + "epoch": 5.235748966013307, + "grad_norm": 0.8686285800363539, + "learning_rate": 1.3966475653849972e-06, + "loss": 0.0135, + "step": 7279 + }, + { + "epoch": 5.236468261104118, + "grad_norm": 1.07139134742929, + "learning_rate": 1.396308131960409e-06, + "loss": 0.0173, + "step": 7280 + }, + { + "epoch": 5.237187556194929, + "grad_norm": 2.434549102614779, + "learning_rate": 1.3959687038709495e-06, + "loss": 0.023, + "step": 7281 + }, + { + "epoch": 5.23790685128574, + "grad_norm": 0.8671583084824102, + "learning_rate": 1.395629281134082e-06, + "loss": 0.0085, + "step": 7282 + }, + { + "epoch": 5.238626146376551, + "grad_norm": 4.6065199645404125, + "learning_rate": 1.3952898637672719e-06, + "loss": 0.1208, + "step": 7283 + }, + { + "epoch": 5.239345441467362, + "grad_norm": 1.5093537908959085, + "learning_rate": 1.3949504517879824e-06, + "loss": 0.0046, + "step": 7284 + }, + { + "epoch": 5.240064736558173, + "grad_norm": 2.8508044894912787, + "learning_rate": 1.3946110452136762e-06, + "loss": 0.0345, + "step": 7285 + }, + { + "epoch": 5.2407840316489835, + "grad_norm": 3.7253621618771695, + "learning_rate": 1.394271644061817e-06, + "loss": 0.095, + "step": 7286 + }, + { + "epoch": 5.241503326739795, + "grad_norm": 2.8156049180656595, + "learning_rate": 1.3939322483498665e-06, + "loss": 0.0496, + "step": 7287 + }, + { + "epoch": 5.242222621830606, + "grad_norm": 2.405260947516223, + "learning_rate": 1.3935928580952887e-06, + "loss": 0.0352, + "step": 7288 + }, + { + "epoch": 5.242941916921417, + "grad_norm": 2.0167300137548696, + "learning_rate": 1.393253473315545e-06, + "loss": 0.0561, + "step": 7289 + }, + { + "epoch": 5.243661212012228, + "grad_norm": 0.4806546753728555, + "learning_rate": 1.3929140940280977e-06, + "loss": 0.0009, + "step": 7290 + }, + { + "epoch": 5.244380507103039, + "grad_norm": 0.02661159954914292, + "learning_rate": 1.3925747202504081e-06, + "loss": 0.0002, + "step": 7291 + }, + { + "epoch": 5.24509980219385, + "grad_norm": 0.08804544477795936, + "learning_rate": 1.3922353519999376e-06, + "loss": 0.0002, + "step": 7292 + }, + { + "epoch": 5.245819097284661, + "grad_norm": 3.6449213154036895, + "learning_rate": 1.3918959892941474e-06, + "loss": 0.127, + "step": 7293 + }, + { + "epoch": 5.246538392375472, + "grad_norm": 1.8540257210969018, + "learning_rate": 1.3915566321504984e-06, + "loss": 0.0294, + "step": 7294 + }, + { + "epoch": 5.247257687466283, + "grad_norm": 4.033722352413505, + "learning_rate": 1.3912172805864512e-06, + "loss": 0.0923, + "step": 7295 + }, + { + "epoch": 5.247976982557094, + "grad_norm": 1.6679384912494009, + "learning_rate": 1.3908779346194658e-06, + "loss": 0.0163, + "step": 7296 + }, + { + "epoch": 5.248696277647905, + "grad_norm": 1.1754194550527266, + "learning_rate": 1.3905385942670025e-06, + "loss": 0.0137, + "step": 7297 + }, + { + "epoch": 5.249415572738716, + "grad_norm": 0.9038850391213998, + "learning_rate": 1.3901992595465199e-06, + "loss": 0.0045, + "step": 7298 + }, + { + "epoch": 5.250134867829527, + "grad_norm": 3.9057187159820606, + "learning_rate": 1.3898599304754787e-06, + "loss": 0.044, + "step": 7299 + }, + { + "epoch": 5.250854162920338, + "grad_norm": 4.351082280902411, + "learning_rate": 1.3895206070713374e-06, + "loss": 0.0656, + "step": 7300 + }, + { + "epoch": 5.2515734580111495, + "grad_norm": 2.266108579304811, + "learning_rate": 1.3891812893515546e-06, + "loss": 0.0169, + "step": 7301 + }, + { + "epoch": 5.25229275310196, + "grad_norm": 1.7633782901002162, + "learning_rate": 1.3888419773335892e-06, + "loss": 0.022, + "step": 7302 + }, + { + "epoch": 5.253012048192771, + "grad_norm": 0.03038005959690293, + "learning_rate": 1.3885026710348986e-06, + "loss": 0.0001, + "step": 7303 + }, + { + "epoch": 5.253731343283582, + "grad_norm": 0.7201032675110464, + "learning_rate": 1.388163370472942e-06, + "loss": 0.002, + "step": 7304 + }, + { + "epoch": 5.254450638374393, + "grad_norm": 0.9409681243401774, + "learning_rate": 1.3878240756651763e-06, + "loss": 0.008, + "step": 7305 + }, + { + "epoch": 5.255169933465204, + "grad_norm": 2.436955463502817, + "learning_rate": 1.3874847866290587e-06, + "loss": 0.0378, + "step": 7306 + }, + { + "epoch": 5.255889228556015, + "grad_norm": 1.0551681925788194, + "learning_rate": 1.3871455033820462e-06, + "loss": 0.0078, + "step": 7307 + }, + { + "epoch": 5.256608523646826, + "grad_norm": 3.167512250970498, + "learning_rate": 1.3868062259415955e-06, + "loss": 0.0385, + "step": 7308 + }, + { + "epoch": 5.257327818737637, + "grad_norm": 3.67785664990662, + "learning_rate": 1.3864669543251634e-06, + "loss": 0.0346, + "step": 7309 + }, + { + "epoch": 5.258047113828448, + "grad_norm": 0.26820137756498974, + "learning_rate": 1.3861276885502061e-06, + "loss": 0.0003, + "step": 7310 + }, + { + "epoch": 5.258766408919259, + "grad_norm": 4.633817599675061, + "learning_rate": 1.385788428634179e-06, + "loss": 0.1132, + "step": 7311 + }, + { + "epoch": 5.25948570401007, + "grad_norm": 2.3906882512191663, + "learning_rate": 1.3854491745945375e-06, + "loss": 0.046, + "step": 7312 + }, + { + "epoch": 5.260204999100881, + "grad_norm": 5.5169800822987325, + "learning_rate": 1.3851099264487374e-06, + "loss": 0.0821, + "step": 7313 + }, + { + "epoch": 5.2609242941916925, + "grad_norm": 2.974172248912101, + "learning_rate": 1.3847706842142327e-06, + "loss": 0.043, + "step": 7314 + }, + { + "epoch": 5.261643589282503, + "grad_norm": 0.049749566932954965, + "learning_rate": 1.384431447908479e-06, + "loss": 0.0002, + "step": 7315 + }, + { + "epoch": 5.262362884373315, + "grad_norm": 2.3895202136343143, + "learning_rate": 1.3840922175489301e-06, + "loss": 0.0384, + "step": 7316 + }, + { + "epoch": 5.263082179464125, + "grad_norm": 2.9012241900365225, + "learning_rate": 1.38375299315304e-06, + "loss": 0.0104, + "step": 7317 + }, + { + "epoch": 5.263801474554937, + "grad_norm": 5.559663550166, + "learning_rate": 1.3834137747382626e-06, + "loss": 0.113, + "step": 7318 + }, + { + "epoch": 5.264520769645747, + "grad_norm": 0.894155780640653, + "learning_rate": 1.3830745623220506e-06, + "loss": 0.003, + "step": 7319 + }, + { + "epoch": 5.265240064736558, + "grad_norm": 0.2436162741553856, + "learning_rate": 1.3827353559218585e-06, + "loss": 0.0005, + "step": 7320 + }, + { + "epoch": 5.265959359827369, + "grad_norm": 2.8987401515929783, + "learning_rate": 1.3823961555551378e-06, + "loss": 0.0377, + "step": 7321 + }, + { + "epoch": 5.26667865491818, + "grad_norm": 5.587602948222889, + "learning_rate": 1.3820569612393416e-06, + "loss": 0.0977, + "step": 7322 + }, + { + "epoch": 5.267397950008991, + "grad_norm": 1.442695806450937, + "learning_rate": 1.3817177729919215e-06, + "loss": 0.0151, + "step": 7323 + }, + { + "epoch": 5.268117245099802, + "grad_norm": 3.1609945798601236, + "learning_rate": 1.3813785908303291e-06, + "loss": 0.064, + "step": 7324 + }, + { + "epoch": 5.268836540190613, + "grad_norm": 3.509996098405957, + "learning_rate": 1.381039414772017e-06, + "loss": 0.0339, + "step": 7325 + }, + { + "epoch": 5.269555835281424, + "grad_norm": 3.886025424872802, + "learning_rate": 1.3807002448344359e-06, + "loss": 0.0966, + "step": 7326 + }, + { + "epoch": 5.2702751303722355, + "grad_norm": 3.042164562881901, + "learning_rate": 1.3803610810350364e-06, + "loss": 0.0589, + "step": 7327 + }, + { + "epoch": 5.270994425463046, + "grad_norm": 0.04470411043932871, + "learning_rate": 1.3800219233912696e-06, + "loss": 0.0002, + "step": 7328 + }, + { + "epoch": 5.271713720553858, + "grad_norm": 0.05113513644372662, + "learning_rate": 1.379682771920585e-06, + "loss": 0.0002, + "step": 7329 + }, + { + "epoch": 5.272433015644668, + "grad_norm": 2.6269694304428173, + "learning_rate": 1.3793436266404335e-06, + "loss": 0.0292, + "step": 7330 + }, + { + "epoch": 5.27315231073548, + "grad_norm": 0.5200992805856504, + "learning_rate": 1.379004487568264e-06, + "loss": 0.0008, + "step": 7331 + }, + { + "epoch": 5.27387160582629, + "grad_norm": 2.6083309842227558, + "learning_rate": 1.378665354721526e-06, + "loss": 0.0393, + "step": 7332 + }, + { + "epoch": 5.274590900917101, + "grad_norm": 2.0571457032492906, + "learning_rate": 1.3783262281176686e-06, + "loss": 0.0357, + "step": 7333 + }, + { + "epoch": 5.275310196007912, + "grad_norm": 3.949010965685134, + "learning_rate": 1.3779871077741404e-06, + "loss": 0.09, + "step": 7334 + }, + { + "epoch": 5.276029491098723, + "grad_norm": 5.826945034510109, + "learning_rate": 1.3776479937083894e-06, + "loss": 0.0778, + "step": 7335 + }, + { + "epoch": 5.276748786189534, + "grad_norm": 2.8293794881197343, + "learning_rate": 1.3773088859378642e-06, + "loss": 0.0365, + "step": 7336 + }, + { + "epoch": 5.277468081280345, + "grad_norm": 2.156259096842191, + "learning_rate": 1.376969784480012e-06, + "loss": 0.0386, + "step": 7337 + }, + { + "epoch": 5.278187376371156, + "grad_norm": 2.8953127054406482, + "learning_rate": 1.3766306893522805e-06, + "loss": 0.0386, + "step": 7338 + }, + { + "epoch": 5.278906671461967, + "grad_norm": 4.808952080165297, + "learning_rate": 1.3762916005721166e-06, + "loss": 0.1447, + "step": 7339 + }, + { + "epoch": 5.2796259665527785, + "grad_norm": 2.742865646955717, + "learning_rate": 1.3759525181569664e-06, + "loss": 0.0468, + "step": 7340 + }, + { + "epoch": 5.280345261643589, + "grad_norm": 1.0352984783415804, + "learning_rate": 1.3756134421242774e-06, + "loss": 0.0057, + "step": 7341 + }, + { + "epoch": 5.281064556734401, + "grad_norm": 0.44545233303695436, + "learning_rate": 1.3752743724914953e-06, + "loss": 0.0029, + "step": 7342 + }, + { + "epoch": 5.281783851825211, + "grad_norm": 3.002640301829756, + "learning_rate": 1.374935309276066e-06, + "loss": 0.0518, + "step": 7343 + }, + { + "epoch": 5.282503146916023, + "grad_norm": 6.015126007150464, + "learning_rate": 1.3745962524954341e-06, + "loss": 0.1336, + "step": 7344 + }, + { + "epoch": 5.283222442006833, + "grad_norm": 2.116306281369816, + "learning_rate": 1.3742572021670447e-06, + "loss": 0.0277, + "step": 7345 + }, + { + "epoch": 5.283941737097645, + "grad_norm": 3.2859656976148806, + "learning_rate": 1.3739181583083436e-06, + "loss": 0.0449, + "step": 7346 + }, + { + "epoch": 5.284661032188455, + "grad_norm": 3.853990772313854, + "learning_rate": 1.3735791209367745e-06, + "loss": 0.0384, + "step": 7347 + }, + { + "epoch": 5.285380327279267, + "grad_norm": 2.396385551803708, + "learning_rate": 1.3732400900697817e-06, + "loss": 0.0272, + "step": 7348 + }, + { + "epoch": 5.286099622370077, + "grad_norm": 3.3930304690179507, + "learning_rate": 1.3729010657248086e-06, + "loss": 0.0882, + "step": 7349 + }, + { + "epoch": 5.286818917460888, + "grad_norm": 3.497403948245251, + "learning_rate": 1.3725620479192987e-06, + "loss": 0.073, + "step": 7350 + }, + { + "epoch": 5.287538212551699, + "grad_norm": 3.0432443101620485, + "learning_rate": 1.3722230366706952e-06, + "loss": 0.0051, + "step": 7351 + }, + { + "epoch": 5.28825750764251, + "grad_norm": 2.7353792060430018, + "learning_rate": 1.3718840319964408e-06, + "loss": 0.0273, + "step": 7352 + }, + { + "epoch": 5.2889768027333215, + "grad_norm": 0.9963772450694115, + "learning_rate": 1.371545033913978e-06, + "loss": 0.0185, + "step": 7353 + }, + { + "epoch": 5.289696097824132, + "grad_norm": 0.09465936310596733, + "learning_rate": 1.3712060424407487e-06, + "loss": 0.0003, + "step": 7354 + }, + { + "epoch": 5.290415392914944, + "grad_norm": 3.6970788829717667, + "learning_rate": 1.3708670575941946e-06, + "loss": 0.0613, + "step": 7355 + }, + { + "epoch": 5.291134688005754, + "grad_norm": 0.6904985358256687, + "learning_rate": 1.3705280793917565e-06, + "loss": 0.0073, + "step": 7356 + }, + { + "epoch": 5.291853983096566, + "grad_norm": 1.840721622808998, + "learning_rate": 1.3701891078508768e-06, + "loss": 0.0266, + "step": 7357 + }, + { + "epoch": 5.292573278187376, + "grad_norm": 4.50740061276203, + "learning_rate": 1.3698501429889948e-06, + "loss": 0.0618, + "step": 7358 + }, + { + "epoch": 5.293292573278188, + "grad_norm": 2.8509275880235134, + "learning_rate": 1.3695111848235518e-06, + "loss": 0.0241, + "step": 7359 + }, + { + "epoch": 5.294011868368998, + "grad_norm": 0.27149257113349784, + "learning_rate": 1.3691722333719874e-06, + "loss": 0.0021, + "step": 7360 + }, + { + "epoch": 5.29473116345981, + "grad_norm": 4.574544316283811, + "learning_rate": 1.3688332886517404e-06, + "loss": 0.114, + "step": 7361 + }, + { + "epoch": 5.29545045855062, + "grad_norm": 6.259319077335234, + "learning_rate": 1.3684943506802517e-06, + "loss": 0.0564, + "step": 7362 + }, + { + "epoch": 5.296169753641431, + "grad_norm": 0.71547554520606, + "learning_rate": 1.3681554194749596e-06, + "loss": 0.0063, + "step": 7363 + }, + { + "epoch": 5.296889048732242, + "grad_norm": 2.256746313777551, + "learning_rate": 1.3678164950533026e-06, + "loss": 0.037, + "step": 7364 + }, + { + "epoch": 5.297608343823053, + "grad_norm": 2.465486035491169, + "learning_rate": 1.367477577432719e-06, + "loss": 0.0374, + "step": 7365 + }, + { + "epoch": 5.2983276389138645, + "grad_norm": 1.8671134525241873, + "learning_rate": 1.3671386666306462e-06, + "loss": 0.0043, + "step": 7366 + }, + { + "epoch": 5.299046934004675, + "grad_norm": 2.5263937716207376, + "learning_rate": 1.3667997626645225e-06, + "loss": 0.0412, + "step": 7367 + }, + { + "epoch": 5.299766229095487, + "grad_norm": 1.5617461760233204, + "learning_rate": 1.366460865551785e-06, + "loss": 0.0154, + "step": 7368 + }, + { + "epoch": 5.300485524186297, + "grad_norm": 6.473534641320791, + "learning_rate": 1.3661219753098702e-06, + "loss": 0.1362, + "step": 7369 + }, + { + "epoch": 5.301204819277109, + "grad_norm": 4.5666433572552805, + "learning_rate": 1.3657830919562152e-06, + "loss": 0.0411, + "step": 7370 + }, + { + "epoch": 5.301924114367919, + "grad_norm": 0.3128917206110903, + "learning_rate": 1.365444215508255e-06, + "loss": 0.0004, + "step": 7371 + }, + { + "epoch": 5.302643409458731, + "grad_norm": 5.237381455890651, + "learning_rate": 1.3651053459834265e-06, + "loss": 0.0543, + "step": 7372 + }, + { + "epoch": 5.303362704549541, + "grad_norm": 0.044711280618026386, + "learning_rate": 1.3647664833991648e-06, + "loss": 0.0002, + "step": 7373 + }, + { + "epoch": 5.304081999640353, + "grad_norm": 0.8513679011131594, + "learning_rate": 1.364427627772905e-06, + "loss": 0.0083, + "step": 7374 + }, + { + "epoch": 5.304801294731163, + "grad_norm": 3.351829766781728, + "learning_rate": 1.3640887791220819e-06, + "loss": 0.0331, + "step": 7375 + }, + { + "epoch": 5.305520589821975, + "grad_norm": 2.9558749863381477, + "learning_rate": 1.3637499374641293e-06, + "loss": 0.0598, + "step": 7376 + }, + { + "epoch": 5.306239884912785, + "grad_norm": 0.047635873091766735, + "learning_rate": 1.363411102816481e-06, + "loss": 0.0003, + "step": 7377 + }, + { + "epoch": 5.306959180003597, + "grad_norm": 2.269110598714633, + "learning_rate": 1.3630722751965722e-06, + "loss": 0.0215, + "step": 7378 + }, + { + "epoch": 5.3076784750944075, + "grad_norm": 2.3125824797012537, + "learning_rate": 1.3627334546218354e-06, + "loss": 0.0307, + "step": 7379 + }, + { + "epoch": 5.308397770185218, + "grad_norm": 3.497016753481696, + "learning_rate": 1.362394641109703e-06, + "loss": 0.0763, + "step": 7380 + }, + { + "epoch": 5.30911706527603, + "grad_norm": 1.5365563450923145, + "learning_rate": 1.3620558346776079e-06, + "loss": 0.0178, + "step": 7381 + }, + { + "epoch": 5.30983636036684, + "grad_norm": 2.178467571901136, + "learning_rate": 1.3617170353429813e-06, + "loss": 0.0347, + "step": 7382 + }, + { + "epoch": 5.310555655457652, + "grad_norm": 2.4872539120879233, + "learning_rate": 1.361378243123257e-06, + "loss": 0.0799, + "step": 7383 + }, + { + "epoch": 5.311274950548462, + "grad_norm": 3.3756528513892334, + "learning_rate": 1.3610394580358652e-06, + "loss": 0.0711, + "step": 7384 + }, + { + "epoch": 5.311994245639274, + "grad_norm": 2.882380154232042, + "learning_rate": 1.3607006800982375e-06, + "loss": 0.0499, + "step": 7385 + }, + { + "epoch": 5.312713540730084, + "grad_norm": 2.1171676934566164, + "learning_rate": 1.360361909327804e-06, + "loss": 0.0546, + "step": 7386 + }, + { + "epoch": 5.313432835820896, + "grad_norm": 0.7489669050472886, + "learning_rate": 1.3600231457419952e-06, + "loss": 0.0033, + "step": 7387 + }, + { + "epoch": 5.314152130911706, + "grad_norm": 0.21289825516752708, + "learning_rate": 1.3596843893582417e-06, + "loss": 0.0004, + "step": 7388 + }, + { + "epoch": 5.314871426002518, + "grad_norm": 3.9796238993144173, + "learning_rate": 1.3593456401939727e-06, + "loss": 0.0591, + "step": 7389 + }, + { + "epoch": 5.315590721093328, + "grad_norm": 0.2556325547374903, + "learning_rate": 1.3590068982666172e-06, + "loss": 0.0003, + "step": 7390 + }, + { + "epoch": 5.31631001618414, + "grad_norm": 0.49443689382737355, + "learning_rate": 1.3586681635936045e-06, + "loss": 0.0035, + "step": 7391 + }, + { + "epoch": 5.3170293112749505, + "grad_norm": 4.134704731921691, + "learning_rate": 1.3583294361923625e-06, + "loss": 0.0813, + "step": 7392 + }, + { + "epoch": 5.317748606365762, + "grad_norm": 3.624335554750656, + "learning_rate": 1.35799071608032e-06, + "loss": 0.0708, + "step": 7393 + }, + { + "epoch": 5.3184679014565726, + "grad_norm": 6.152085985579055, + "learning_rate": 1.3576520032749045e-06, + "loss": 0.091, + "step": 7394 + }, + { + "epoch": 5.319187196547384, + "grad_norm": 0.08536217344221173, + "learning_rate": 1.3573132977935432e-06, + "loss": 0.0002, + "step": 7395 + }, + { + "epoch": 5.319906491638195, + "grad_norm": 3.2455198883028453, + "learning_rate": 1.3569745996536633e-06, + "loss": 0.0475, + "step": 7396 + }, + { + "epoch": 5.320625786729005, + "grad_norm": 1.8859126002181437, + "learning_rate": 1.3566359088726914e-06, + "loss": 0.0185, + "step": 7397 + }, + { + "epoch": 5.321345081819817, + "grad_norm": 5.184887336453816, + "learning_rate": 1.3562972254680529e-06, + "loss": 0.0882, + "step": 7398 + }, + { + "epoch": 5.322064376910627, + "grad_norm": 3.652074000327136, + "learning_rate": 1.355958549457175e-06, + "loss": 0.0681, + "step": 7399 + }, + { + "epoch": 5.322783672001439, + "grad_norm": 0.022006868777262576, + "learning_rate": 1.3556198808574827e-06, + "loss": 0.0001, + "step": 7400 + }, + { + "epoch": 5.323502967092249, + "grad_norm": 0.005357752497003767, + "learning_rate": 1.355281219686401e-06, + "loss": 0.0, + "step": 7401 + }, + { + "epoch": 5.324222262183061, + "grad_norm": 1.5618577167955383, + "learning_rate": 1.354942565961355e-06, + "loss": 0.0208, + "step": 7402 + }, + { + "epoch": 5.324941557273871, + "grad_norm": 0.007096775554909222, + "learning_rate": 1.3546039196997674e-06, + "loss": 0.0, + "step": 7403 + }, + { + "epoch": 5.325660852364683, + "grad_norm": 3.9772916042238577, + "learning_rate": 1.3542652809190641e-06, + "loss": 0.0876, + "step": 7404 + }, + { + "epoch": 5.3263801474554935, + "grad_norm": 4.197670756026596, + "learning_rate": 1.353926649636668e-06, + "loss": 0.0147, + "step": 7405 + }, + { + "epoch": 5.327099442546305, + "grad_norm": 2.781141640375021, + "learning_rate": 1.353588025870002e-06, + "loss": 0.0296, + "step": 7406 + }, + { + "epoch": 5.3278187376371156, + "grad_norm": 2.505636770965489, + "learning_rate": 1.3532494096364894e-06, + "loss": 0.0456, + "step": 7407 + }, + { + "epoch": 5.328538032727927, + "grad_norm": 3.5295516397020585, + "learning_rate": 1.3529108009535516e-06, + "loss": 0.063, + "step": 7408 + }, + { + "epoch": 5.329257327818738, + "grad_norm": 1.9168363362853391, + "learning_rate": 1.3525721998386118e-06, + "loss": 0.0215, + "step": 7409 + }, + { + "epoch": 5.329976622909548, + "grad_norm": 1.8524292535043556, + "learning_rate": 1.3522336063090913e-06, + "loss": 0.0295, + "step": 7410 + }, + { + "epoch": 5.33069591800036, + "grad_norm": 2.3618645391967745, + "learning_rate": 1.3518950203824109e-06, + "loss": 0.0499, + "step": 7411 + }, + { + "epoch": 5.33141521309117, + "grad_norm": 1.1524247476825695, + "learning_rate": 1.3515564420759916e-06, + "loss": 0.017, + "step": 7412 + }, + { + "epoch": 5.332134508181982, + "grad_norm": 3.4128900947661696, + "learning_rate": 1.3512178714072537e-06, + "loss": 0.0899, + "step": 7413 + }, + { + "epoch": 5.332853803272792, + "grad_norm": 5.573829972207637, + "learning_rate": 1.3508793083936179e-06, + "loss": 0.0815, + "step": 7414 + }, + { + "epoch": 5.333573098363604, + "grad_norm": 2.690711625868066, + "learning_rate": 1.3505407530525037e-06, + "loss": 0.0443, + "step": 7415 + }, + { + "epoch": 5.334292393454414, + "grad_norm": 4.075162174829484, + "learning_rate": 1.35020220540133e-06, + "loss": 0.0758, + "step": 7416 + }, + { + "epoch": 5.335011688545226, + "grad_norm": 1.5007511438675336, + "learning_rate": 1.3498636654575157e-06, + "loss": 0.024, + "step": 7417 + }, + { + "epoch": 5.3357309836360365, + "grad_norm": 2.725701995524049, + "learning_rate": 1.3495251332384787e-06, + "loss": 0.051, + "step": 7418 + }, + { + "epoch": 5.336450278726848, + "grad_norm": 5.868439768764539, + "learning_rate": 1.3491866087616383e-06, + "loss": 0.0952, + "step": 7419 + }, + { + "epoch": 5.3371695738176586, + "grad_norm": 8.822216270204818, + "learning_rate": 1.348848092044412e-06, + "loss": 0.0159, + "step": 7420 + }, + { + "epoch": 5.33788886890847, + "grad_norm": 0.06884758421379188, + "learning_rate": 1.3485095831042162e-06, + "loss": 0.0002, + "step": 7421 + }, + { + "epoch": 5.338608163999281, + "grad_norm": 4.3491551979403535, + "learning_rate": 1.3481710819584685e-06, + "loss": 0.0648, + "step": 7422 + }, + { + "epoch": 5.339327459090092, + "grad_norm": 2.2554290242347754, + "learning_rate": 1.3478325886245852e-06, + "loss": 0.0327, + "step": 7423 + }, + { + "epoch": 5.340046754180903, + "grad_norm": 2.097280143782767, + "learning_rate": 1.347494103119982e-06, + "loss": 0.0125, + "step": 7424 + }, + { + "epoch": 5.340766049271714, + "grad_norm": 2.436383146317189, + "learning_rate": 1.3471556254620752e-06, + "loss": 0.0374, + "step": 7425 + }, + { + "epoch": 5.341485344362525, + "grad_norm": 3.10137631801775, + "learning_rate": 1.3468171556682798e-06, + "loss": 0.0379, + "step": 7426 + }, + { + "epoch": 5.342204639453335, + "grad_norm": 2.2826221550096975, + "learning_rate": 1.3464786937560106e-06, + "loss": 0.0235, + "step": 7427 + }, + { + "epoch": 5.342923934544147, + "grad_norm": 2.9368844892899757, + "learning_rate": 1.3461402397426822e-06, + "loss": 0.0457, + "step": 7428 + }, + { + "epoch": 5.343643229634957, + "grad_norm": 0.10403070364154932, + "learning_rate": 1.3458017936457082e-06, + "loss": 0.0002, + "step": 7429 + }, + { + "epoch": 5.344362524725769, + "grad_norm": 3.55369168296378, + "learning_rate": 1.3454633554825028e-06, + "loss": 0.0604, + "step": 7430 + }, + { + "epoch": 5.3450818198165795, + "grad_norm": 3.163394047894213, + "learning_rate": 1.3451249252704792e-06, + "loss": 0.0514, + "step": 7431 + }, + { + "epoch": 5.345801114907391, + "grad_norm": 2.528430127643923, + "learning_rate": 1.34478650302705e-06, + "loss": 0.0318, + "step": 7432 + }, + { + "epoch": 5.3465204099982016, + "grad_norm": 0.14896109007356648, + "learning_rate": 1.3444480887696276e-06, + "loss": 0.0004, + "step": 7433 + }, + { + "epoch": 5.347239705089013, + "grad_norm": 3.310387108743183, + "learning_rate": 1.3441096825156236e-06, + "loss": 0.0319, + "step": 7434 + }, + { + "epoch": 5.347959000179824, + "grad_norm": 3.9745746546419802, + "learning_rate": 1.3437712842824507e-06, + "loss": 0.0794, + "step": 7435 + }, + { + "epoch": 5.348678295270635, + "grad_norm": 0.18461487169916307, + "learning_rate": 1.3434328940875197e-06, + "loss": 0.0009, + "step": 7436 + }, + { + "epoch": 5.349397590361446, + "grad_norm": 1.8243994298630977, + "learning_rate": 1.3430945119482409e-06, + "loss": 0.0235, + "step": 7437 + }, + { + "epoch": 5.350116885452257, + "grad_norm": 0.11467752657687232, + "learning_rate": 1.3427561378820253e-06, + "loss": 0.0002, + "step": 7438 + }, + { + "epoch": 5.350836180543068, + "grad_norm": 0.9308511810202094, + "learning_rate": 1.3424177719062812e-06, + "loss": 0.0125, + "step": 7439 + }, + { + "epoch": 5.351555475633878, + "grad_norm": 2.9751563374074066, + "learning_rate": 1.3420794140384203e-06, + "loss": 0.0255, + "step": 7440 + }, + { + "epoch": 5.35227477072469, + "grad_norm": 0.2318394698946671, + "learning_rate": 1.3417410642958505e-06, + "loss": 0.0005, + "step": 7441 + }, + { + "epoch": 5.3529940658155, + "grad_norm": 1.8280914338244785, + "learning_rate": 1.341402722695981e-06, + "loss": 0.0388, + "step": 7442 + }, + { + "epoch": 5.353713360906312, + "grad_norm": 4.788308224894148, + "learning_rate": 1.3410643892562197e-06, + "loss": 0.0797, + "step": 7443 + }, + { + "epoch": 5.3544326559971225, + "grad_norm": 0.9552633255241214, + "learning_rate": 1.3407260639939747e-06, + "loss": 0.0055, + "step": 7444 + }, + { + "epoch": 5.355151951087934, + "grad_norm": 1.253474172981717, + "learning_rate": 1.340387746926653e-06, + "loss": 0.017, + "step": 7445 + }, + { + "epoch": 5.3558712461787445, + "grad_norm": 1.6625539804538199, + "learning_rate": 1.3400494380716622e-06, + "loss": 0.0182, + "step": 7446 + }, + { + "epoch": 5.356590541269556, + "grad_norm": 0.00909631310824595, + "learning_rate": 1.3397111374464085e-06, + "loss": 0.0001, + "step": 7447 + }, + { + "epoch": 5.357309836360367, + "grad_norm": 2.7208388240716626, + "learning_rate": 1.339372845068298e-06, + "loss": 0.0226, + "step": 7448 + }, + { + "epoch": 5.358029131451178, + "grad_norm": 6.915472159987167, + "learning_rate": 1.339034560954737e-06, + "loss": 0.1075, + "step": 7449 + }, + { + "epoch": 5.358748426541989, + "grad_norm": 0.08346980465621069, + "learning_rate": 1.3386962851231295e-06, + "loss": 0.0002, + "step": 7450 + }, + { + "epoch": 5.3594677216328, + "grad_norm": 2.319474711944378, + "learning_rate": 1.3383580175908824e-06, + "loss": 0.0602, + "step": 7451 + }, + { + "epoch": 5.360187016723611, + "grad_norm": 0.00976056347919367, + "learning_rate": 1.3380197583753985e-06, + "loss": 0.0, + "step": 7452 + }, + { + "epoch": 5.360906311814422, + "grad_norm": 0.009808001724680384, + "learning_rate": 1.3376815074940826e-06, + "loss": 0.0, + "step": 7453 + }, + { + "epoch": 5.361625606905233, + "grad_norm": 1.7684223084985775, + "learning_rate": 1.337343264964338e-06, + "loss": 0.0369, + "step": 7454 + }, + { + "epoch": 5.362344901996044, + "grad_norm": 2.2308292358504747, + "learning_rate": 1.3370050308035673e-06, + "loss": 0.0152, + "step": 7455 + }, + { + "epoch": 5.363064197086855, + "grad_norm": 0.07507522102638309, + "learning_rate": 1.3366668050291746e-06, + "loss": 0.0003, + "step": 7456 + }, + { + "epoch": 5.3637834921776655, + "grad_norm": 4.348400368162575, + "learning_rate": 1.3363285876585614e-06, + "loss": 0.0947, + "step": 7457 + }, + { + "epoch": 5.364502787268477, + "grad_norm": 1.6149332061123665, + "learning_rate": 1.3359903787091297e-06, + "loss": 0.0146, + "step": 7458 + }, + { + "epoch": 5.3652220823592875, + "grad_norm": 2.999273790053224, + "learning_rate": 1.335652178198281e-06, + "loss": 0.0647, + "step": 7459 + }, + { + "epoch": 5.365941377450099, + "grad_norm": 7.545266639272397, + "learning_rate": 1.3353139861434159e-06, + "loss": 0.2474, + "step": 7460 + }, + { + "epoch": 5.36666067254091, + "grad_norm": 6.65439893252641, + "learning_rate": 1.3349758025619357e-06, + "loss": 0.2751, + "step": 7461 + }, + { + "epoch": 5.367379967631721, + "grad_norm": 3.0143671019902447, + "learning_rate": 1.33463762747124e-06, + "loss": 0.0634, + "step": 7462 + }, + { + "epoch": 5.368099262722532, + "grad_norm": 3.811642101517247, + "learning_rate": 1.3342994608887286e-06, + "loss": 0.0908, + "step": 7463 + }, + { + "epoch": 5.368818557813343, + "grad_norm": 0.21989668303421817, + "learning_rate": 1.333961302831801e-06, + "loss": 0.0006, + "step": 7464 + }, + { + "epoch": 5.369537852904154, + "grad_norm": 0.031050614008504264, + "learning_rate": 1.3336231533178557e-06, + "loss": 0.0001, + "step": 7465 + }, + { + "epoch": 5.370257147994965, + "grad_norm": 1.2425139951456419, + "learning_rate": 1.3332850123642911e-06, + "loss": 0.0171, + "step": 7466 + }, + { + "epoch": 5.370976443085776, + "grad_norm": 4.066965218546766, + "learning_rate": 1.3329468799885053e-06, + "loss": 0.0417, + "step": 7467 + }, + { + "epoch": 5.371695738176587, + "grad_norm": 0.4105521097742785, + "learning_rate": 1.3326087562078961e-06, + "loss": 0.0009, + "step": 7468 + }, + { + "epoch": 5.372415033267398, + "grad_norm": 4.1511885308648155, + "learning_rate": 1.33227064103986e-06, + "loss": 0.0705, + "step": 7469 + }, + { + "epoch": 5.373134328358209, + "grad_norm": 1.2008577098037356, + "learning_rate": 1.331932534501794e-06, + "loss": 0.0028, + "step": 7470 + }, + { + "epoch": 5.37385362344902, + "grad_norm": 3.023336178684661, + "learning_rate": 1.3315944366110934e-06, + "loss": 0.0119, + "step": 7471 + }, + { + "epoch": 5.374572918539831, + "grad_norm": 3.320378708573247, + "learning_rate": 1.3312563473851552e-06, + "loss": 0.0812, + "step": 7472 + }, + { + "epoch": 5.375292213630642, + "grad_norm": 4.103660329342572, + "learning_rate": 1.3309182668413744e-06, + "loss": 0.0628, + "step": 7473 + }, + { + "epoch": 5.376011508721453, + "grad_norm": 1.9991440202589386, + "learning_rate": 1.3305801949971458e-06, + "loss": 0.0261, + "step": 7474 + }, + { + "epoch": 5.376730803812264, + "grad_norm": 3.8561165827885335, + "learning_rate": 1.330242131869863e-06, + "loss": 0.0813, + "step": 7475 + }, + { + "epoch": 5.377450098903075, + "grad_norm": 0.4896540021914113, + "learning_rate": 1.3299040774769202e-06, + "loss": 0.0035, + "step": 7476 + }, + { + "epoch": 5.378169393993886, + "grad_norm": 3.501628325240536, + "learning_rate": 1.3295660318357118e-06, + "loss": 0.0348, + "step": 7477 + }, + { + "epoch": 5.378888689084697, + "grad_norm": 0.15540708756364402, + "learning_rate": 1.3292279949636302e-06, + "loss": 0.0006, + "step": 7478 + }, + { + "epoch": 5.379607984175508, + "grad_norm": 2.0082806252797556, + "learning_rate": 1.3288899668780681e-06, + "loss": 0.0101, + "step": 7479 + }, + { + "epoch": 5.380327279266319, + "grad_norm": 1.855318137445232, + "learning_rate": 1.3285519475964176e-06, + "loss": 0.0328, + "step": 7480 + }, + { + "epoch": 5.38104657435713, + "grad_norm": 1.5361822802020118, + "learning_rate": 1.3282139371360698e-06, + "loss": 0.0155, + "step": 7481 + }, + { + "epoch": 5.381765869447941, + "grad_norm": 1.3341942741540846, + "learning_rate": 1.3278759355144172e-06, + "loss": 0.0169, + "step": 7482 + }, + { + "epoch": 5.382485164538752, + "grad_norm": 1.1734736817403506, + "learning_rate": 1.3275379427488495e-06, + "loss": 0.0079, + "step": 7483 + }, + { + "epoch": 5.383204459629563, + "grad_norm": 4.639039487769261, + "learning_rate": 1.3271999588567576e-06, + "loss": 0.0832, + "step": 7484 + }, + { + "epoch": 5.383923754720374, + "grad_norm": 2.3669189542895888, + "learning_rate": 1.326861983855531e-06, + "loss": 0.0231, + "step": 7485 + }, + { + "epoch": 5.384643049811185, + "grad_norm": 3.9106269227511716, + "learning_rate": 1.3265240177625595e-06, + "loss": 0.0504, + "step": 7486 + }, + { + "epoch": 5.385362344901996, + "grad_norm": 0.6648852374251532, + "learning_rate": 1.3261860605952314e-06, + "loss": 0.001, + "step": 7487 + }, + { + "epoch": 5.386081639992807, + "grad_norm": 2.1631858601468155, + "learning_rate": 1.3258481123709357e-06, + "loss": 0.0178, + "step": 7488 + }, + { + "epoch": 5.386800935083618, + "grad_norm": 4.263296737161368, + "learning_rate": 1.3255101731070607e-06, + "loss": 0.04, + "step": 7489 + }, + { + "epoch": 5.387520230174429, + "grad_norm": 3.12541946522176, + "learning_rate": 1.3251722428209934e-06, + "loss": 0.0657, + "step": 7490 + }, + { + "epoch": 5.38823952526524, + "grad_norm": 0.15592625098416918, + "learning_rate": 1.3248343215301212e-06, + "loss": 0.0002, + "step": 7491 + }, + { + "epoch": 5.388958820356051, + "grad_norm": 2.919073355504835, + "learning_rate": 1.3244964092518299e-06, + "loss": 0.0404, + "step": 7492 + }, + { + "epoch": 5.389678115446862, + "grad_norm": 0.9306622967216757, + "learning_rate": 1.3241585060035071e-06, + "loss": 0.015, + "step": 7493 + }, + { + "epoch": 5.390397410537673, + "grad_norm": 3.3168841154292634, + "learning_rate": 1.323820611802538e-06, + "loss": 0.0471, + "step": 7494 + }, + { + "epoch": 5.391116705628484, + "grad_norm": 2.129982688228415, + "learning_rate": 1.3234827266663074e-06, + "loss": 0.0424, + "step": 7495 + }, + { + "epoch": 5.391836000719295, + "grad_norm": 1.7944169247544588, + "learning_rate": 1.3231448506122006e-06, + "loss": 0.019, + "step": 7496 + }, + { + "epoch": 5.392555295810106, + "grad_norm": 1.5228399769448717, + "learning_rate": 1.3228069836576013e-06, + "loss": 0.0167, + "step": 7497 + }, + { + "epoch": 5.393274590900917, + "grad_norm": 1.2279089130107501, + "learning_rate": 1.3224691258198942e-06, + "loss": 0.0054, + "step": 7498 + }, + { + "epoch": 5.393993885991728, + "grad_norm": 0.1373783101365158, + "learning_rate": 1.322131277116462e-06, + "loss": 0.0006, + "step": 7499 + }, + { + "epoch": 5.3947131810825395, + "grad_norm": 1.5149689327000213, + "learning_rate": 1.3217934375646879e-06, + "loss": 0.003, + "step": 7500 + }, + { + "epoch": 5.39543247617335, + "grad_norm": 0.5681217452368451, + "learning_rate": 1.3214556071819543e-06, + "loss": 0.0047, + "step": 7501 + }, + { + "epoch": 5.396151771264162, + "grad_norm": 4.976825309955332, + "learning_rate": 1.3211177859856427e-06, + "loss": 0.1161, + "step": 7502 + }, + { + "epoch": 5.396871066354972, + "grad_norm": 3.98095037175492, + "learning_rate": 1.3207799739931355e-06, + "loss": 0.065, + "step": 7503 + }, + { + "epoch": 5.397590361445783, + "grad_norm": 3.458753318094638, + "learning_rate": 1.3204421712218132e-06, + "loss": 0.0386, + "step": 7504 + }, + { + "epoch": 5.398309656536594, + "grad_norm": 2.2952346325961055, + "learning_rate": 1.3201043776890564e-06, + "loss": 0.0316, + "step": 7505 + }, + { + "epoch": 5.399028951627405, + "grad_norm": 0.1177932868482289, + "learning_rate": 1.3197665934122451e-06, + "loss": 0.0003, + "step": 7506 + }, + { + "epoch": 5.399748246718216, + "grad_norm": 0.026337892025551812, + "learning_rate": 1.3194288184087592e-06, + "loss": 0.0001, + "step": 7507 + }, + { + "epoch": 5.400467541809027, + "grad_norm": 4.320575453237553, + "learning_rate": 1.3190910526959766e-06, + "loss": 0.0868, + "step": 7508 + }, + { + "epoch": 5.401186836899838, + "grad_norm": 1.4495175228877692, + "learning_rate": 1.3187532962912779e-06, + "loss": 0.0218, + "step": 7509 + }, + { + "epoch": 5.401906131990649, + "grad_norm": 0.12908400735589654, + "learning_rate": 1.3184155492120404e-06, + "loss": 0.0003, + "step": 7510 + }, + { + "epoch": 5.40262542708146, + "grad_norm": 2.1952026075513302, + "learning_rate": 1.3180778114756412e-06, + "loss": 0.0328, + "step": 7511 + }, + { + "epoch": 5.403344722172271, + "grad_norm": 3.4129632906669283, + "learning_rate": 1.3177400830994581e-06, + "loss": 0.06, + "step": 7512 + }, + { + "epoch": 5.4040640172630825, + "grad_norm": 1.929878696873473, + "learning_rate": 1.317402364100867e-06, + "loss": 0.0059, + "step": 7513 + }, + { + "epoch": 5.404783312353893, + "grad_norm": 1.3847611207340984, + "learning_rate": 1.3170646544972456e-06, + "loss": 0.0147, + "step": 7514 + }, + { + "epoch": 5.405502607444705, + "grad_norm": 3.984508710573086, + "learning_rate": 1.3167269543059688e-06, + "loss": 0.0365, + "step": 7515 + }, + { + "epoch": 5.406221902535515, + "grad_norm": 2.5770397177067297, + "learning_rate": 1.3163892635444117e-06, + "loss": 0.0344, + "step": 7516 + }, + { + "epoch": 5.406941197626326, + "grad_norm": 0.8403415813089086, + "learning_rate": 1.3160515822299494e-06, + "loss": 0.0069, + "step": 7517 + }, + { + "epoch": 5.407660492717137, + "grad_norm": 3.910052501776252, + "learning_rate": 1.315713910379956e-06, + "loss": 0.0688, + "step": 7518 + }, + { + "epoch": 5.408379787807948, + "grad_norm": 4.032116843434403, + "learning_rate": 1.3153762480118057e-06, + "loss": 0.0874, + "step": 7519 + }, + { + "epoch": 5.409099082898759, + "grad_norm": 4.647863857604363, + "learning_rate": 1.3150385951428714e-06, + "loss": 0.0866, + "step": 7520 + }, + { + "epoch": 5.40981837798957, + "grad_norm": 0.13505572878409697, + "learning_rate": 1.314700951790526e-06, + "loss": 0.0004, + "step": 7521 + }, + { + "epoch": 5.410537673080381, + "grad_norm": 4.382278789764294, + "learning_rate": 1.3143633179721421e-06, + "loss": 0.07, + "step": 7522 + }, + { + "epoch": 5.411256968171192, + "grad_norm": 9.338973184114566, + "learning_rate": 1.314025693705091e-06, + "loss": 0.1144, + "step": 7523 + }, + { + "epoch": 5.411976263262003, + "grad_norm": 0.21336052666509522, + "learning_rate": 1.3136880790067446e-06, + "loss": 0.0004, + "step": 7524 + }, + { + "epoch": 5.412695558352814, + "grad_norm": 1.6163703756372352, + "learning_rate": 1.3133504738944738e-06, + "loss": 0.0204, + "step": 7525 + }, + { + "epoch": 5.4134148534436255, + "grad_norm": 2.2285090155246574, + "learning_rate": 1.3130128783856485e-06, + "loss": 0.0315, + "step": 7526 + }, + { + "epoch": 5.414134148534436, + "grad_norm": 0.3295317115840085, + "learning_rate": 1.3126752924976392e-06, + "loss": 0.0008, + "step": 7527 + }, + { + "epoch": 5.414853443625248, + "grad_norm": 3.998093532181488, + "learning_rate": 1.3123377162478147e-06, + "loss": 0.0142, + "step": 7528 + }, + { + "epoch": 5.415572738716058, + "grad_norm": 4.953540926286474, + "learning_rate": 1.3120001496535434e-06, + "loss": 0.0454, + "step": 7529 + }, + { + "epoch": 5.41629203380687, + "grad_norm": 4.7476457809576225, + "learning_rate": 1.3116625927321953e-06, + "loss": 0.1446, + "step": 7530 + }, + { + "epoch": 5.41701132889768, + "grad_norm": 4.930912630661696, + "learning_rate": 1.3113250455011373e-06, + "loss": 0.0808, + "step": 7531 + }, + { + "epoch": 5.417730623988492, + "grad_norm": 1.3546760436538818, + "learning_rate": 1.3109875079777368e-06, + "loss": 0.0169, + "step": 7532 + }, + { + "epoch": 5.418449919079302, + "grad_norm": 3.576294279375639, + "learning_rate": 1.3106499801793614e-06, + "loss": 0.0538, + "step": 7533 + }, + { + "epoch": 5.419169214170113, + "grad_norm": 3.2544538345124456, + "learning_rate": 1.310312462123376e-06, + "loss": 0.0628, + "step": 7534 + }, + { + "epoch": 5.419888509260924, + "grad_norm": 1.4478554278805353, + "learning_rate": 1.3099749538271478e-06, + "loss": 0.0237, + "step": 7535 + }, + { + "epoch": 5.420607804351735, + "grad_norm": 3.8247920730294056, + "learning_rate": 1.309637455308042e-06, + "loss": 0.0691, + "step": 7536 + }, + { + "epoch": 5.421327099442546, + "grad_norm": 3.960748293715007, + "learning_rate": 1.3092999665834232e-06, + "loss": 0.0462, + "step": 7537 + }, + { + "epoch": 5.422046394533357, + "grad_norm": 5.750345157269577, + "learning_rate": 1.3089624876706558e-06, + "loss": 0.1132, + "step": 7538 + }, + { + "epoch": 5.4227656896241685, + "grad_norm": 2.274279373816701, + "learning_rate": 1.3086250185871035e-06, + "loss": 0.0356, + "step": 7539 + }, + { + "epoch": 5.423484984714979, + "grad_norm": 2.438443012366515, + "learning_rate": 1.3082875593501303e-06, + "loss": 0.0511, + "step": 7540 + }, + { + "epoch": 5.424204279805791, + "grad_norm": 1.26708068941434, + "learning_rate": 1.3079501099770986e-06, + "loss": 0.0049, + "step": 7541 + }, + { + "epoch": 5.424923574896601, + "grad_norm": 0.7480833940093767, + "learning_rate": 1.307612670485371e-06, + "loss": 0.0019, + "step": 7542 + }, + { + "epoch": 5.425642869987413, + "grad_norm": 3.2082116582266162, + "learning_rate": 1.3072752408923091e-06, + "loss": 0.0472, + "step": 7543 + }, + { + "epoch": 5.426362165078223, + "grad_norm": 1.9928755228531352, + "learning_rate": 1.3069378212152739e-06, + "loss": 0.0166, + "step": 7544 + }, + { + "epoch": 5.427081460169035, + "grad_norm": 2.968165251081598, + "learning_rate": 1.3066004114716273e-06, + "loss": 0.0867, + "step": 7545 + }, + { + "epoch": 5.427800755259845, + "grad_norm": 2.6080057349919326, + "learning_rate": 1.3062630116787293e-06, + "loss": 0.0515, + "step": 7546 + }, + { + "epoch": 5.428520050350657, + "grad_norm": 4.05290531744028, + "learning_rate": 1.3059256218539392e-06, + "loss": 0.1014, + "step": 7547 + }, + { + "epoch": 5.429239345441467, + "grad_norm": 2.035837327436559, + "learning_rate": 1.3055882420146166e-06, + "loss": 0.005, + "step": 7548 + }, + { + "epoch": 5.429958640532279, + "grad_norm": 3.8578209861563253, + "learning_rate": 1.3052508721781201e-06, + "loss": 0.071, + "step": 7549 + }, + { + "epoch": 5.430677935623089, + "grad_norm": 2.2066391955631692, + "learning_rate": 1.3049135123618072e-06, + "loss": 0.0208, + "step": 7550 + }, + { + "epoch": 5.4313972307139, + "grad_norm": 3.414765640735232, + "learning_rate": 1.3045761625830377e-06, + "loss": 0.0757, + "step": 7551 + }, + { + "epoch": 5.4321165258047115, + "grad_norm": 2.2295652532049264, + "learning_rate": 1.3042388228591675e-06, + "loss": 0.0374, + "step": 7552 + }, + { + "epoch": 5.432835820895522, + "grad_norm": 3.6032040098302995, + "learning_rate": 1.3039014932075535e-06, + "loss": 0.0344, + "step": 7553 + }, + { + "epoch": 5.433555115986334, + "grad_norm": 3.536864955553012, + "learning_rate": 1.3035641736455518e-06, + "loss": 0.0464, + "step": 7554 + }, + { + "epoch": 5.434274411077144, + "grad_norm": 2.953911479188811, + "learning_rate": 1.303226864190518e-06, + "loss": 0.0798, + "step": 7555 + }, + { + "epoch": 5.434993706167956, + "grad_norm": 4.06343942543462, + "learning_rate": 1.3028895648598078e-06, + "loss": 0.0798, + "step": 7556 + }, + { + "epoch": 5.435713001258766, + "grad_norm": 2.2081538489712984, + "learning_rate": 1.3025522756707754e-06, + "loss": 0.0334, + "step": 7557 + }, + { + "epoch": 5.436432296349578, + "grad_norm": 3.661499091735498, + "learning_rate": 1.3022149966407751e-06, + "loss": 0.0501, + "step": 7558 + }, + { + "epoch": 5.437151591440388, + "grad_norm": 4.005395549267602, + "learning_rate": 1.3018777277871604e-06, + "loss": 0.0709, + "step": 7559 + }, + { + "epoch": 5.4378708865312, + "grad_norm": 2.0106580195392, + "learning_rate": 1.301540469127284e-06, + "loss": 0.006, + "step": 7560 + }, + { + "epoch": 5.43859018162201, + "grad_norm": 0.2524008902434002, + "learning_rate": 1.3012032206784994e-06, + "loss": 0.0018, + "step": 7561 + }, + { + "epoch": 5.439309476712822, + "grad_norm": 2.2506327149006005, + "learning_rate": 1.300865982458158e-06, + "loss": 0.0292, + "step": 7562 + }, + { + "epoch": 5.440028771803632, + "grad_norm": 1.3941933020465018, + "learning_rate": 1.3005287544836111e-06, + "loss": 0.0043, + "step": 7563 + }, + { + "epoch": 5.440748066894443, + "grad_norm": 2.7649676194225656, + "learning_rate": 1.3001915367722101e-06, + "loss": 0.027, + "step": 7564 + }, + { + "epoch": 5.4414673619852545, + "grad_norm": 1.3942772185485761, + "learning_rate": 1.2998543293413048e-06, + "loss": 0.0221, + "step": 7565 + }, + { + "epoch": 5.442186657076065, + "grad_norm": 3.634697503221275, + "learning_rate": 1.2995171322082461e-06, + "loss": 0.067, + "step": 7566 + }, + { + "epoch": 5.442905952166877, + "grad_norm": 3.3656602212637807, + "learning_rate": 1.2991799453903829e-06, + "loss": 0.0532, + "step": 7567 + }, + { + "epoch": 5.443625247257687, + "grad_norm": 2.3238324946879008, + "learning_rate": 1.2988427689050641e-06, + "loss": 0.0162, + "step": 7568 + }, + { + "epoch": 5.444344542348499, + "grad_norm": 2.544887183137854, + "learning_rate": 1.2985056027696382e-06, + "loss": 0.0388, + "step": 7569 + }, + { + "epoch": 5.445063837439309, + "grad_norm": 1.6907091857886878, + "learning_rate": 1.2981684470014517e-06, + "loss": 0.0244, + "step": 7570 + }, + { + "epoch": 5.445783132530121, + "grad_norm": 0.6096827350588276, + "learning_rate": 1.2978313016178536e-06, + "loss": 0.0037, + "step": 7571 + }, + { + "epoch": 5.446502427620931, + "grad_norm": 4.0468912089554, + "learning_rate": 1.2974941666361902e-06, + "loss": 0.0364, + "step": 7572 + }, + { + "epoch": 5.447221722711743, + "grad_norm": 4.270455137726797, + "learning_rate": 1.2971570420738071e-06, + "loss": 0.0922, + "step": 7573 + }, + { + "epoch": 5.447941017802553, + "grad_norm": 3.1071226284493676, + "learning_rate": 1.2968199279480505e-06, + "loss": 0.0829, + "step": 7574 + }, + { + "epoch": 5.448660312893365, + "grad_norm": 5.683063252771741, + "learning_rate": 1.296482824276265e-06, + "loss": 0.0796, + "step": 7575 + }, + { + "epoch": 5.449379607984175, + "grad_norm": 1.975499815908523, + "learning_rate": 1.2961457310757954e-06, + "loss": 0.0158, + "step": 7576 + }, + { + "epoch": 5.450098903074987, + "grad_norm": 0.5960023935735871, + "learning_rate": 1.295808648363986e-06, + "loss": 0.003, + "step": 7577 + }, + { + "epoch": 5.4508181981657975, + "grad_norm": 0.08700871687112798, + "learning_rate": 1.29547157615818e-06, + "loss": 0.0002, + "step": 7578 + }, + { + "epoch": 5.451537493256609, + "grad_norm": 2.8315632685945227, + "learning_rate": 1.2951345144757205e-06, + "loss": 0.0386, + "step": 7579 + }, + { + "epoch": 5.45225678834742, + "grad_norm": 1.3468005682598276, + "learning_rate": 1.29479746333395e-06, + "loss": 0.013, + "step": 7580 + }, + { + "epoch": 5.45297608343823, + "grad_norm": 1.0672543625232753, + "learning_rate": 1.2944604227502095e-06, + "loss": 0.0059, + "step": 7581 + }, + { + "epoch": 5.453695378529042, + "grad_norm": 1.458082547982284, + "learning_rate": 1.294123392741842e-06, + "loss": 0.007, + "step": 7582 + }, + { + "epoch": 5.454414673619852, + "grad_norm": 2.3517832775058714, + "learning_rate": 1.293786373326187e-06, + "loss": 0.0525, + "step": 7583 + }, + { + "epoch": 5.455133968710664, + "grad_norm": 3.8592419377004235, + "learning_rate": 1.293449364520585e-06, + "loss": 0.02, + "step": 7584 + }, + { + "epoch": 5.455853263801474, + "grad_norm": 2.6022526609431678, + "learning_rate": 1.293112366342376e-06, + "loss": 0.0125, + "step": 7585 + }, + { + "epoch": 5.456572558892286, + "grad_norm": 1.3487167765777979, + "learning_rate": 1.292775378808898e-06, + "loss": 0.0234, + "step": 7586 + }, + { + "epoch": 5.457291853983096, + "grad_norm": 5.565259794187252, + "learning_rate": 1.2924384019374916e-06, + "loss": 0.122, + "step": 7587 + }, + { + "epoch": 5.458011149073908, + "grad_norm": 1.348896264135799, + "learning_rate": 1.2921014357454935e-06, + "loss": 0.011, + "step": 7588 + }, + { + "epoch": 5.458730444164718, + "grad_norm": 4.023299624133011, + "learning_rate": 1.2917644802502415e-06, + "loss": 0.0319, + "step": 7589 + }, + { + "epoch": 5.45944973925553, + "grad_norm": 1.4999207566228159, + "learning_rate": 1.2914275354690725e-06, + "loss": 0.0079, + "step": 7590 + }, + { + "epoch": 5.4601690343463405, + "grad_norm": 0.905363692666937, + "learning_rate": 1.2910906014193227e-06, + "loss": 0.0025, + "step": 7591 + }, + { + "epoch": 5.460888329437152, + "grad_norm": 1.4218245977077772, + "learning_rate": 1.2907536781183285e-06, + "loss": 0.0265, + "step": 7592 + }, + { + "epoch": 5.461607624527963, + "grad_norm": 2.01656250953325, + "learning_rate": 1.290416765583425e-06, + "loss": 0.0139, + "step": 7593 + }, + { + "epoch": 5.462326919618773, + "grad_norm": 3.9869939660725504, + "learning_rate": 1.2900798638319465e-06, + "loss": 0.0683, + "step": 7594 + }, + { + "epoch": 5.463046214709585, + "grad_norm": 1.3059054396064422, + "learning_rate": 1.2897429728812278e-06, + "loss": 0.0028, + "step": 7595 + }, + { + "epoch": 5.463765509800395, + "grad_norm": 2.36729315622929, + "learning_rate": 1.2894060927486022e-06, + "loss": 0.0504, + "step": 7596 + }, + { + "epoch": 5.464484804891207, + "grad_norm": 3.2516332331259483, + "learning_rate": 1.2890692234514024e-06, + "loss": 0.0579, + "step": 7597 + }, + { + "epoch": 5.465204099982017, + "grad_norm": 8.46881572071739, + "learning_rate": 1.2887323650069618e-06, + "loss": 0.0923, + "step": 7598 + }, + { + "epoch": 5.465923395072829, + "grad_norm": 0.03357431493190687, + "learning_rate": 1.2883955174326117e-06, + "loss": 0.0001, + "step": 7599 + }, + { + "epoch": 5.466642690163639, + "grad_norm": 4.261604417182827, + "learning_rate": 1.2880586807456836e-06, + "loss": 0.0262, + "step": 7600 + }, + { + "epoch": 5.467361985254451, + "grad_norm": 3.9391865692453782, + "learning_rate": 1.2877218549635087e-06, + "loss": 0.1054, + "step": 7601 + }, + { + "epoch": 5.468081280345261, + "grad_norm": 3.214041606953252, + "learning_rate": 1.2873850401034163e-06, + "loss": 0.0383, + "step": 7602 + }, + { + "epoch": 5.468800575436073, + "grad_norm": 0.03839917439888741, + "learning_rate": 1.2870482361827374e-06, + "loss": 0.0002, + "step": 7603 + }, + { + "epoch": 5.4695198705268835, + "grad_norm": 1.4735251755701253, + "learning_rate": 1.2867114432188005e-06, + "loss": 0.026, + "step": 7604 + }, + { + "epoch": 5.470239165617695, + "grad_norm": 0.6442930836628655, + "learning_rate": 1.286374661228935e-06, + "loss": 0.0023, + "step": 7605 + }, + { + "epoch": 5.4709584607085056, + "grad_norm": 2.9030939902810466, + "learning_rate": 1.2860378902304674e-06, + "loss": 0.0323, + "step": 7606 + }, + { + "epoch": 5.471677755799317, + "grad_norm": 0.30351634585818166, + "learning_rate": 1.2857011302407254e-06, + "loss": 0.0019, + "step": 7607 + }, + { + "epoch": 5.472397050890128, + "grad_norm": 0.23794727326772758, + "learning_rate": 1.285364381277037e-06, + "loss": 0.0004, + "step": 7608 + }, + { + "epoch": 5.473116345980939, + "grad_norm": 4.41014704280344, + "learning_rate": 1.2850276433567283e-06, + "loss": 0.1303, + "step": 7609 + }, + { + "epoch": 5.47383564107175, + "grad_norm": 2.970183079576433, + "learning_rate": 1.2846909164971243e-06, + "loss": 0.0391, + "step": 7610 + }, + { + "epoch": 5.47455493616256, + "grad_norm": 2.1665462094521706, + "learning_rate": 1.2843542007155508e-06, + "loss": 0.034, + "step": 7611 + }, + { + "epoch": 5.475274231253372, + "grad_norm": 0.8693936244681728, + "learning_rate": 1.2840174960293322e-06, + "loss": 0.0033, + "step": 7612 + }, + { + "epoch": 5.475993526344182, + "grad_norm": 7.6688224577132065, + "learning_rate": 1.2836808024557926e-06, + "loss": 0.0577, + "step": 7613 + }, + { + "epoch": 5.476712821434994, + "grad_norm": 0.24235073845129812, + "learning_rate": 1.2833441200122553e-06, + "loss": 0.0015, + "step": 7614 + }, + { + "epoch": 5.477432116525804, + "grad_norm": 4.1064719315557126, + "learning_rate": 1.2830074487160435e-06, + "loss": 0.1174, + "step": 7615 + }, + { + "epoch": 5.478151411616616, + "grad_norm": 3.459628247753235, + "learning_rate": 1.2826707885844796e-06, + "loss": 0.0474, + "step": 7616 + }, + { + "epoch": 5.4788707067074265, + "grad_norm": 0.17488906874960744, + "learning_rate": 1.2823341396348847e-06, + "loss": 0.001, + "step": 7617 + }, + { + "epoch": 5.479590001798238, + "grad_norm": 2.2630222203097627, + "learning_rate": 1.2819975018845805e-06, + "loss": 0.0247, + "step": 7618 + }, + { + "epoch": 5.4803092968890486, + "grad_norm": 0.1255270968657423, + "learning_rate": 1.2816608753508876e-06, + "loss": 0.0002, + "step": 7619 + }, + { + "epoch": 5.48102859197986, + "grad_norm": 1.8240539793839008, + "learning_rate": 1.281324260051126e-06, + "loss": 0.0088, + "step": 7620 + }, + { + "epoch": 5.481747887070671, + "grad_norm": 1.8819454751199975, + "learning_rate": 1.2809876560026152e-06, + "loss": 0.0204, + "step": 7621 + }, + { + "epoch": 5.482467182161482, + "grad_norm": 4.079871767863255, + "learning_rate": 1.2806510632226739e-06, + "loss": 0.0846, + "step": 7622 + }, + { + "epoch": 5.483186477252293, + "grad_norm": 1.6152296710818272, + "learning_rate": 1.2803144817286198e-06, + "loss": 0.0199, + "step": 7623 + }, + { + "epoch": 5.483905772343104, + "grad_norm": 0.15134692728986116, + "learning_rate": 1.279977911537772e-06, + "loss": 0.0003, + "step": 7624 + }, + { + "epoch": 5.484625067433915, + "grad_norm": 3.567284167834874, + "learning_rate": 1.2796413526674466e-06, + "loss": 0.0363, + "step": 7625 + }, + { + "epoch": 5.485344362524726, + "grad_norm": 5.6844464801373595, + "learning_rate": 1.2793048051349607e-06, + "loss": 0.0796, + "step": 7626 + }, + { + "epoch": 5.486063657615537, + "grad_norm": 2.534637065510339, + "learning_rate": 1.2789682689576299e-06, + "loss": 0.0338, + "step": 7627 + }, + { + "epoch": 5.486782952706347, + "grad_norm": 2.3735032964600764, + "learning_rate": 1.2786317441527693e-06, + "loss": 0.038, + "step": 7628 + }, + { + "epoch": 5.487502247797159, + "grad_norm": 1.317781480969631, + "learning_rate": 1.2782952307376945e-06, + "loss": 0.0041, + "step": 7629 + }, + { + "epoch": 5.4882215428879695, + "grad_norm": 1.742969396315307, + "learning_rate": 1.2779587287297192e-06, + "loss": 0.01, + "step": 7630 + }, + { + "epoch": 5.488940837978781, + "grad_norm": 4.050404381312407, + "learning_rate": 1.277622238146157e-06, + "loss": 0.0656, + "step": 7631 + }, + { + "epoch": 5.4896601330695916, + "grad_norm": 4.460367889663098, + "learning_rate": 1.2772857590043213e-06, + "loss": 0.0637, + "step": 7632 + }, + { + "epoch": 5.490379428160403, + "grad_norm": 0.18872372659434508, + "learning_rate": 1.2769492913215237e-06, + "loss": 0.0004, + "step": 7633 + }, + { + "epoch": 5.491098723251214, + "grad_norm": 3.772690941995044, + "learning_rate": 1.2766128351150771e-06, + "loss": 0.0776, + "step": 7634 + }, + { + "epoch": 5.491818018342025, + "grad_norm": 0.786848311605422, + "learning_rate": 1.2762763904022922e-06, + "loss": 0.0033, + "step": 7635 + }, + { + "epoch": 5.492537313432836, + "grad_norm": 0.04754309543230001, + "learning_rate": 1.2759399572004797e-06, + "loss": 0.0002, + "step": 7636 + }, + { + "epoch": 5.493256608523647, + "grad_norm": 0.9507801144909905, + "learning_rate": 1.2756035355269497e-06, + "loss": 0.0051, + "step": 7637 + }, + { + "epoch": 5.493975903614458, + "grad_norm": 1.1165103782891814, + "learning_rate": 1.2752671253990118e-06, + "loss": 0.0207, + "step": 7638 + }, + { + "epoch": 5.494695198705269, + "grad_norm": 3.733519715668466, + "learning_rate": 1.274930726833974e-06, + "loss": 0.0454, + "step": 7639 + }, + { + "epoch": 5.49541449379608, + "grad_norm": 5.8990037818235646, + "learning_rate": 1.2745943398491462e-06, + "loss": 0.0827, + "step": 7640 + }, + { + "epoch": 5.49613378888689, + "grad_norm": 2.696257766099016, + "learning_rate": 1.2742579644618356e-06, + "loss": 0.0343, + "step": 7641 + }, + { + "epoch": 5.496853083977702, + "grad_norm": 2.9644671267345095, + "learning_rate": 1.2739216006893489e-06, + "loss": 0.0431, + "step": 7642 + }, + { + "epoch": 5.4975723790685125, + "grad_norm": 0.0677733562331387, + "learning_rate": 1.2735852485489924e-06, + "loss": 0.0002, + "step": 7643 + }, + { + "epoch": 5.498291674159324, + "grad_norm": 0.5874833024041378, + "learning_rate": 1.2732489080580716e-06, + "loss": 0.0043, + "step": 7644 + }, + { + "epoch": 5.4990109692501346, + "grad_norm": 1.6287683635980479, + "learning_rate": 1.2729125792338935e-06, + "loss": 0.0182, + "step": 7645 + }, + { + "epoch": 5.499730264340946, + "grad_norm": 6.1826826351214805, + "learning_rate": 1.2725762620937616e-06, + "loss": 0.1767, + "step": 7646 + }, + { + "epoch": 5.500449559431757, + "grad_norm": 0.47477329064806606, + "learning_rate": 1.2722399566549803e-06, + "loss": 0.0017, + "step": 7647 + }, + { + "epoch": 5.501168854522568, + "grad_norm": 3.0149664475590376, + "learning_rate": 1.2719036629348531e-06, + "loss": 0.0347, + "step": 7648 + }, + { + "epoch": 5.501888149613379, + "grad_norm": 1.5654070505309692, + "learning_rate": 1.2715673809506825e-06, + "loss": 0.0264, + "step": 7649 + }, + { + "epoch": 5.50260744470419, + "grad_norm": 1.7191614243483848, + "learning_rate": 1.2712311107197715e-06, + "loss": 0.0122, + "step": 7650 + }, + { + "epoch": 5.503326739795001, + "grad_norm": 0.049972644228537345, + "learning_rate": 1.2708948522594216e-06, + "loss": 0.0002, + "step": 7651 + }, + { + "epoch": 5.504046034885812, + "grad_norm": 4.1054794660213165, + "learning_rate": 1.2705586055869333e-06, + "loss": 0.0605, + "step": 7652 + }, + { + "epoch": 5.504765329976623, + "grad_norm": 1.2710048518261252, + "learning_rate": 1.2702223707196078e-06, + "loss": 0.0014, + "step": 7653 + }, + { + "epoch": 5.505484625067434, + "grad_norm": 1.3945600084053698, + "learning_rate": 1.2698861476747443e-06, + "loss": 0.0137, + "step": 7654 + }, + { + "epoch": 5.506203920158245, + "grad_norm": 1.8181633169904745, + "learning_rate": 1.2695499364696429e-06, + "loss": 0.0305, + "step": 7655 + }, + { + "epoch": 5.506923215249056, + "grad_norm": 6.014228688624087, + "learning_rate": 1.2692137371216018e-06, + "loss": 0.0613, + "step": 7656 + }, + { + "epoch": 5.507642510339867, + "grad_norm": 4.27477178107093, + "learning_rate": 1.2688775496479188e-06, + "loss": 0.0909, + "step": 7657 + }, + { + "epoch": 5.5083618054306775, + "grad_norm": 1.254198222085139, + "learning_rate": 1.268541374065892e-06, + "loss": 0.0241, + "step": 7658 + }, + { + "epoch": 5.509081100521489, + "grad_norm": 1.0895671028134222, + "learning_rate": 1.2682052103928175e-06, + "loss": 0.0063, + "step": 7659 + }, + { + "epoch": 5.5098003956123, + "grad_norm": 1.0664022360444922, + "learning_rate": 1.2678690586459911e-06, + "loss": 0.0081, + "step": 7660 + }, + { + "epoch": 5.510519690703111, + "grad_norm": 0.3710684729445661, + "learning_rate": 1.26753291884271e-06, + "loss": 0.0015, + "step": 7661 + }, + { + "epoch": 5.511238985793922, + "grad_norm": 3.3197699039962822, + "learning_rate": 1.2671967910002683e-06, + "loss": 0.0353, + "step": 7662 + }, + { + "epoch": 5.511958280884733, + "grad_norm": 3.6328679728664874, + "learning_rate": 1.2668606751359603e-06, + "loss": 0.0677, + "step": 7663 + }, + { + "epoch": 5.512677575975544, + "grad_norm": 1.0425945829518677, + "learning_rate": 1.2665245712670802e-06, + "loss": 0.0137, + "step": 7664 + }, + { + "epoch": 5.513396871066355, + "grad_norm": 1.8671841168492003, + "learning_rate": 1.2661884794109197e-06, + "loss": 0.0154, + "step": 7665 + }, + { + "epoch": 5.514116166157166, + "grad_norm": 3.3723899035954044, + "learning_rate": 1.2658523995847732e-06, + "loss": 0.0518, + "step": 7666 + }, + { + "epoch": 5.514835461247977, + "grad_norm": 2.6733676990564232, + "learning_rate": 1.2655163318059318e-06, + "loss": 0.0367, + "step": 7667 + }, + { + "epoch": 5.515554756338788, + "grad_norm": 4.0257842559984836, + "learning_rate": 1.2651802760916864e-06, + "loss": 0.0815, + "step": 7668 + }, + { + "epoch": 5.516274051429599, + "grad_norm": 0.30850374942491526, + "learning_rate": 1.2648442324593283e-06, + "loss": 0.0012, + "step": 7669 + }, + { + "epoch": 5.51699334652041, + "grad_norm": 6.78379618873102, + "learning_rate": 1.264508200926147e-06, + "loss": 0.0815, + "step": 7670 + }, + { + "epoch": 5.5177126416112205, + "grad_norm": 2.95235914950292, + "learning_rate": 1.2641721815094322e-06, + "loss": 0.0138, + "step": 7671 + }, + { + "epoch": 5.518431936702032, + "grad_norm": 5.40869070998929, + "learning_rate": 1.2638361742264727e-06, + "loss": 0.0919, + "step": 7672 + }, + { + "epoch": 5.5191512317928435, + "grad_norm": 1.4907094385873145, + "learning_rate": 1.2635001790945568e-06, + "loss": 0.0222, + "step": 7673 + }, + { + "epoch": 5.519870526883654, + "grad_norm": 8.720607419152712, + "learning_rate": 1.2631641961309716e-06, + "loss": 0.1241, + "step": 7674 + }, + { + "epoch": 5.520589821974465, + "grad_norm": 1.1502842916975393, + "learning_rate": 1.2628282253530037e-06, + "loss": 0.0131, + "step": 7675 + }, + { + "epoch": 5.521309117065276, + "grad_norm": 5.7949652879058196, + "learning_rate": 1.2624922667779406e-06, + "loss": 0.0559, + "step": 7676 + }, + { + "epoch": 5.522028412156087, + "grad_norm": 0.031245404243390673, + "learning_rate": 1.2621563204230679e-06, + "loss": 0.0001, + "step": 7677 + }, + { + "epoch": 5.522747707246898, + "grad_norm": 2.747290404614942, + "learning_rate": 1.2618203863056694e-06, + "loss": 0.0136, + "step": 7678 + }, + { + "epoch": 5.523467002337709, + "grad_norm": 1.9059260183475646, + "learning_rate": 1.26148446444303e-06, + "loss": 0.0359, + "step": 7679 + }, + { + "epoch": 5.52418629742852, + "grad_norm": 0.26783344099838396, + "learning_rate": 1.261148554852434e-06, + "loss": 0.0012, + "step": 7680 + }, + { + "epoch": 5.524905592519331, + "grad_norm": 2.606039116852023, + "learning_rate": 1.2608126575511632e-06, + "loss": 0.0379, + "step": 7681 + }, + { + "epoch": 5.525624887610142, + "grad_norm": 7.513064727101199, + "learning_rate": 1.2604767725565018e-06, + "loss": 0.1088, + "step": 7682 + }, + { + "epoch": 5.526344182700953, + "grad_norm": 2.6299687439747834, + "learning_rate": 1.260140899885731e-06, + "loss": 0.0306, + "step": 7683 + }, + { + "epoch": 5.527063477791764, + "grad_norm": 1.3527969102096447, + "learning_rate": 1.259805039556132e-06, + "loss": 0.0092, + "step": 7684 + }, + { + "epoch": 5.527782772882575, + "grad_norm": 0.029008492440217987, + "learning_rate": 1.2594691915849852e-06, + "loss": 0.0001, + "step": 7685 + }, + { + "epoch": 5.5285020679733865, + "grad_norm": 3.3588767737646354, + "learning_rate": 1.2591333559895705e-06, + "loss": 0.0433, + "step": 7686 + }, + { + "epoch": 5.529221363064197, + "grad_norm": 4.8642395834619485, + "learning_rate": 1.2587975327871677e-06, + "loss": 0.1007, + "step": 7687 + }, + { + "epoch": 5.529940658155008, + "grad_norm": 2.529469493301085, + "learning_rate": 1.2584617219950552e-06, + "loss": 0.0513, + "step": 7688 + }, + { + "epoch": 5.530659953245819, + "grad_norm": 4.71907895649444, + "learning_rate": 1.2581259236305112e-06, + "loss": 0.048, + "step": 7689 + }, + { + "epoch": 5.53137924833663, + "grad_norm": 1.2673951313857732, + "learning_rate": 1.2577901377108133e-06, + "loss": 0.0103, + "step": 7690 + }, + { + "epoch": 5.532098543427441, + "grad_norm": 0.18364485011558515, + "learning_rate": 1.2574543642532373e-06, + "loss": 0.0014, + "step": 7691 + }, + { + "epoch": 5.532817838518252, + "grad_norm": 0.8220915375286836, + "learning_rate": 1.2571186032750606e-06, + "loss": 0.0015, + "step": 7692 + }, + { + "epoch": 5.533537133609063, + "grad_norm": 1.9796032902737728, + "learning_rate": 1.2567828547935579e-06, + "loss": 0.0273, + "step": 7693 + }, + { + "epoch": 5.534256428699874, + "grad_norm": 1.9225383786515406, + "learning_rate": 1.2564471188260043e-06, + "loss": 0.0243, + "step": 7694 + }, + { + "epoch": 5.534975723790685, + "grad_norm": 5.538238270605186, + "learning_rate": 1.2561113953896742e-06, + "loss": 0.0742, + "step": 7695 + }, + { + "epoch": 5.535695018881496, + "grad_norm": 2.4707444378615206, + "learning_rate": 1.25577568450184e-06, + "loss": 0.0355, + "step": 7696 + }, + { + "epoch": 5.536414313972307, + "grad_norm": 3.3936080374093485, + "learning_rate": 1.2554399861797765e-06, + "loss": 0.0417, + "step": 7697 + }, + { + "epoch": 5.537133609063118, + "grad_norm": 4.004763072973091, + "learning_rate": 1.2551043004407549e-06, + "loss": 0.0765, + "step": 7698 + }, + { + "epoch": 5.5378529041539295, + "grad_norm": 6.727941131875739, + "learning_rate": 1.254768627302047e-06, + "loss": 0.066, + "step": 7699 + }, + { + "epoch": 5.53857219924474, + "grad_norm": 2.0482264448957492, + "learning_rate": 1.2544329667809239e-06, + "loss": 0.0436, + "step": 7700 + }, + { + "epoch": 5.539291494335551, + "grad_norm": 2.8723777432815445, + "learning_rate": 1.2540973188946555e-06, + "loss": 0.0303, + "step": 7701 + }, + { + "epoch": 5.540010789426362, + "grad_norm": 1.106682251565676, + "learning_rate": 1.253761683660511e-06, + "loss": 0.0072, + "step": 7702 + }, + { + "epoch": 5.540730084517174, + "grad_norm": 1.8331184177289166, + "learning_rate": 1.2534260610957608e-06, + "loss": 0.0241, + "step": 7703 + }, + { + "epoch": 5.541449379607984, + "grad_norm": 2.1919769704960315, + "learning_rate": 1.2530904512176725e-06, + "loss": 0.028, + "step": 7704 + }, + { + "epoch": 5.542168674698795, + "grad_norm": 2.101038281437633, + "learning_rate": 1.252754854043514e-06, + "loss": 0.0167, + "step": 7705 + }, + { + "epoch": 5.542887969789606, + "grad_norm": 0.061042950240325286, + "learning_rate": 1.2524192695905526e-06, + "loss": 0.0002, + "step": 7706 + }, + { + "epoch": 5.543607264880417, + "grad_norm": 2.129094817104068, + "learning_rate": 1.2520836978760537e-06, + "loss": 0.0265, + "step": 7707 + }, + { + "epoch": 5.544326559971228, + "grad_norm": 2.3158289016641915, + "learning_rate": 1.251748138917284e-06, + "loss": 0.0262, + "step": 7708 + }, + { + "epoch": 5.545045855062039, + "grad_norm": 3.2965145401280243, + "learning_rate": 1.2514125927315086e-06, + "loss": 0.0692, + "step": 7709 + }, + { + "epoch": 5.54576515015285, + "grad_norm": 3.762131425557469, + "learning_rate": 1.2510770593359918e-06, + "loss": 0.0797, + "step": 7710 + }, + { + "epoch": 5.546484445243661, + "grad_norm": 2.300625731602827, + "learning_rate": 1.2507415387479969e-06, + "loss": 0.0319, + "step": 7711 + }, + { + "epoch": 5.5472037403344725, + "grad_norm": 0.9023276766739263, + "learning_rate": 1.2504060309847867e-06, + "loss": 0.0166, + "step": 7712 + }, + { + "epoch": 5.547923035425283, + "grad_norm": 4.294629562508708, + "learning_rate": 1.2500705360636257e-06, + "loss": 0.0836, + "step": 7713 + }, + { + "epoch": 5.548642330516095, + "grad_norm": 4.464330016648919, + "learning_rate": 1.2497350540017739e-06, + "loss": 0.0583, + "step": 7714 + }, + { + "epoch": 5.549361625606905, + "grad_norm": 2.3364189381171943, + "learning_rate": 1.249399584816493e-06, + "loss": 0.0252, + "step": 7715 + }, + { + "epoch": 5.550080920697717, + "grad_norm": 3.0337283382757767, + "learning_rate": 1.249064128525043e-06, + "loss": 0.0485, + "step": 7716 + }, + { + "epoch": 5.550800215788527, + "grad_norm": 2.7372658317089784, + "learning_rate": 1.2487286851446838e-06, + "loss": 0.0421, + "step": 7717 + }, + { + "epoch": 5.551519510879338, + "grad_norm": 2.304698898296957, + "learning_rate": 1.2483932546926755e-06, + "loss": 0.0277, + "step": 7718 + }, + { + "epoch": 5.552238805970149, + "grad_norm": 0.035223373915042606, + "learning_rate": 1.2480578371862757e-06, + "loss": 0.0001, + "step": 7719 + }, + { + "epoch": 5.55295810106096, + "grad_norm": 8.36610933773056, + "learning_rate": 1.2477224326427426e-06, + "loss": 0.008, + "step": 7720 + }, + { + "epoch": 5.553677396151771, + "grad_norm": 3.1868050758009847, + "learning_rate": 1.247387041079333e-06, + "loss": 0.0698, + "step": 7721 + }, + { + "epoch": 5.554396691242582, + "grad_norm": 0.9863339177378369, + "learning_rate": 1.2470516625133037e-06, + "loss": 0.0097, + "step": 7722 + }, + { + "epoch": 5.555115986333393, + "grad_norm": 0.1623928845049415, + "learning_rate": 1.2467162969619102e-06, + "loss": 0.0002, + "step": 7723 + }, + { + "epoch": 5.555835281424204, + "grad_norm": 2.850173499446205, + "learning_rate": 1.2463809444424082e-06, + "loss": 0.0238, + "step": 7724 + }, + { + "epoch": 5.5565545765150155, + "grad_norm": 1.2344061449852353, + "learning_rate": 1.2460456049720517e-06, + "loss": 0.0095, + "step": 7725 + }, + { + "epoch": 5.557273871605826, + "grad_norm": 4.076258188357139, + "learning_rate": 1.2457102785680946e-06, + "loss": 0.0367, + "step": 7726 + }, + { + "epoch": 5.557993166696638, + "grad_norm": 1.2108509286467086, + "learning_rate": 1.24537496524779e-06, + "loss": 0.0071, + "step": 7727 + }, + { + "epoch": 5.558712461787448, + "grad_norm": 4.278470680598555, + "learning_rate": 1.2450396650283904e-06, + "loss": 0.0676, + "step": 7728 + }, + { + "epoch": 5.55943175687826, + "grad_norm": 1.0600518890246151, + "learning_rate": 1.2447043779271477e-06, + "loss": 0.0133, + "step": 7729 + }, + { + "epoch": 5.56015105196907, + "grad_norm": 3.5690319427940738, + "learning_rate": 1.2443691039613127e-06, + "loss": 0.0487, + "step": 7730 + }, + { + "epoch": 5.560870347059882, + "grad_norm": 0.045304588145897565, + "learning_rate": 1.2440338431481362e-06, + "loss": 0.0002, + "step": 7731 + }, + { + "epoch": 5.561589642150692, + "grad_norm": 3.3324292136090423, + "learning_rate": 1.243698595504868e-06, + "loss": 0.036, + "step": 7732 + }, + { + "epoch": 5.562308937241504, + "grad_norm": 2.29055823469417, + "learning_rate": 1.2433633610487563e-06, + "loss": 0.0278, + "step": 7733 + }, + { + "epoch": 5.563028232332314, + "grad_norm": 2.4542283120533197, + "learning_rate": 1.2430281397970507e-06, + "loss": 0.0061, + "step": 7734 + }, + { + "epoch": 5.563747527423125, + "grad_norm": 1.9916811108617176, + "learning_rate": 1.2426929317669986e-06, + "loss": 0.0031, + "step": 7735 + }, + { + "epoch": 5.564466822513936, + "grad_norm": 3.7348297433778517, + "learning_rate": 1.242357736975847e-06, + "loss": 0.114, + "step": 7736 + }, + { + "epoch": 5.565186117604747, + "grad_norm": 0.0023656234529079788, + "learning_rate": 1.2420225554408418e-06, + "loss": 0.0, + "step": 7737 + }, + { + "epoch": 5.5659054126955585, + "grad_norm": 5.142713949250958, + "learning_rate": 1.2416873871792281e-06, + "loss": 0.086, + "step": 7738 + }, + { + "epoch": 5.566624707786369, + "grad_norm": 1.6205090860644837, + "learning_rate": 1.2413522322082528e-06, + "loss": 0.0175, + "step": 7739 + }, + { + "epoch": 5.567344002877181, + "grad_norm": 2.858752839629277, + "learning_rate": 1.241017090545159e-06, + "loss": 0.0197, + "step": 7740 + }, + { + "epoch": 5.568063297967991, + "grad_norm": 0.44378107965324165, + "learning_rate": 1.2406819622071907e-06, + "loss": 0.0038, + "step": 7741 + }, + { + "epoch": 5.568782593058803, + "grad_norm": 5.061138868918315, + "learning_rate": 1.2403468472115905e-06, + "loss": 0.095, + "step": 7742 + }, + { + "epoch": 5.569501888149613, + "grad_norm": 1.2610238137013263, + "learning_rate": 1.2400117455756007e-06, + "loss": 0.0128, + "step": 7743 + }, + { + "epoch": 5.570221183240425, + "grad_norm": 0.017333546500308473, + "learning_rate": 1.239676657316463e-06, + "loss": 0.0001, + "step": 7744 + }, + { + "epoch": 5.570940478331235, + "grad_norm": 4.250929851715818, + "learning_rate": 1.2393415824514186e-06, + "loss": 0.0664, + "step": 7745 + }, + { + "epoch": 5.571659773422047, + "grad_norm": 3.0681692448939377, + "learning_rate": 1.239006520997707e-06, + "loss": 0.0478, + "step": 7746 + }, + { + "epoch": 5.572379068512857, + "grad_norm": 0.10197686341742196, + "learning_rate": 1.2386714729725685e-06, + "loss": 0.0006, + "step": 7747 + }, + { + "epoch": 5.573098363603668, + "grad_norm": 0.23231470881067712, + "learning_rate": 1.238336438393241e-06, + "loss": 0.0018, + "step": 7748 + }, + { + "epoch": 5.573817658694479, + "grad_norm": 0.5070270410654102, + "learning_rate": 1.238001417276963e-06, + "loss": 0.0057, + "step": 7749 + }, + { + "epoch": 5.574536953785291, + "grad_norm": 2.9965844501290655, + "learning_rate": 1.2376664096409723e-06, + "loss": 0.0212, + "step": 7750 + }, + { + "epoch": 5.5752562488761015, + "grad_norm": 1.0409398751860375, + "learning_rate": 1.2373314155025054e-06, + "loss": 0.0077, + "step": 7751 + }, + { + "epoch": 5.575975543966912, + "grad_norm": 0.26609387378443816, + "learning_rate": 1.2369964348787981e-06, + "loss": 0.0018, + "step": 7752 + }, + { + "epoch": 5.576694839057724, + "grad_norm": 2.9304794491621675, + "learning_rate": 1.236661467787086e-06, + "loss": 0.0472, + "step": 7753 + }, + { + "epoch": 5.577414134148534, + "grad_norm": 5.685656845308165, + "learning_rate": 1.236326514244603e-06, + "loss": 0.054, + "step": 7754 + }, + { + "epoch": 5.578133429239346, + "grad_norm": 3.02517360895136, + "learning_rate": 1.2359915742685844e-06, + "loss": 0.0597, + "step": 7755 + }, + { + "epoch": 5.578852724330156, + "grad_norm": 2.7860054457063215, + "learning_rate": 1.2356566478762628e-06, + "loss": 0.0372, + "step": 7756 + }, + { + "epoch": 5.579572019420968, + "grad_norm": 2.2438356639150845, + "learning_rate": 1.2353217350848706e-06, + "loss": 0.0426, + "step": 7757 + }, + { + "epoch": 5.580291314511778, + "grad_norm": 0.12780000778052653, + "learning_rate": 1.2349868359116398e-06, + "loss": 0.0003, + "step": 7758 + }, + { + "epoch": 5.58101060960259, + "grad_norm": 2.1806043239970605, + "learning_rate": 1.2346519503738013e-06, + "loss": 0.0202, + "step": 7759 + }, + { + "epoch": 5.5817299046934, + "grad_norm": 1.821949698419662, + "learning_rate": 1.2343170784885859e-06, + "loss": 0.0451, + "step": 7760 + }, + { + "epoch": 5.582449199784212, + "grad_norm": 2.017113447805738, + "learning_rate": 1.2339822202732232e-06, + "loss": 0.028, + "step": 7761 + }, + { + "epoch": 5.583168494875022, + "grad_norm": 3.1881064849045755, + "learning_rate": 1.2336473757449423e-06, + "loss": 0.0676, + "step": 7762 + }, + { + "epoch": 5.583887789965834, + "grad_norm": 2.930914669346216, + "learning_rate": 1.2333125449209714e-06, + "loss": 0.0828, + "step": 7763 + }, + { + "epoch": 5.5846070850566445, + "grad_norm": 0.985150023699614, + "learning_rate": 1.232977727818538e-06, + "loss": 0.0113, + "step": 7764 + }, + { + "epoch": 5.585326380147455, + "grad_norm": 6.397353534351081, + "learning_rate": 1.2326429244548697e-06, + "loss": 0.033, + "step": 7765 + }, + { + "epoch": 5.586045675238267, + "grad_norm": 2.978873998963986, + "learning_rate": 1.232308134847192e-06, + "loss": 0.029, + "step": 7766 + }, + { + "epoch": 5.586764970329077, + "grad_norm": 1.6711389643120598, + "learning_rate": 1.231973359012731e-06, + "loss": 0.0119, + "step": 7767 + }, + { + "epoch": 5.587484265419889, + "grad_norm": 3.0201603909572086, + "learning_rate": 1.231638596968711e-06, + "loss": 0.0541, + "step": 7768 + }, + { + "epoch": 5.588203560510699, + "grad_norm": 4.228007628587333, + "learning_rate": 1.231303848732356e-06, + "loss": 0.0683, + "step": 7769 + }, + { + "epoch": 5.588922855601511, + "grad_norm": 1.908088128108533, + "learning_rate": 1.2309691143208895e-06, + "loss": 0.0157, + "step": 7770 + }, + { + "epoch": 5.589642150692321, + "grad_norm": 0.05391512471896949, + "learning_rate": 1.2306343937515347e-06, + "loss": 0.0002, + "step": 7771 + }, + { + "epoch": 5.590361445783133, + "grad_norm": 2.9048107590604832, + "learning_rate": 1.2302996870415136e-06, + "loss": 0.034, + "step": 7772 + }, + { + "epoch": 5.591080740873943, + "grad_norm": 0.2824985130339726, + "learning_rate": 1.229964994208047e-06, + "loss": 0.0005, + "step": 7773 + }, + { + "epoch": 5.591800035964755, + "grad_norm": 1.7846304301372102, + "learning_rate": 1.2296303152683549e-06, + "loss": 0.0462, + "step": 7774 + }, + { + "epoch": 5.592519331055565, + "grad_norm": 3.676272293549527, + "learning_rate": 1.2292956502396575e-06, + "loss": 0.0417, + "step": 7775 + }, + { + "epoch": 5.593238626146377, + "grad_norm": 0.054071061568353854, + "learning_rate": 1.2289609991391746e-06, + "loss": 0.0001, + "step": 7776 + }, + { + "epoch": 5.5939579212371875, + "grad_norm": 1.5443169939253327, + "learning_rate": 1.228626361984124e-06, + "loss": 0.0243, + "step": 7777 + }, + { + "epoch": 5.594677216327998, + "grad_norm": 0.5944437695190288, + "learning_rate": 1.2282917387917234e-06, + "loss": 0.0033, + "step": 7778 + }, + { + "epoch": 5.59539651141881, + "grad_norm": 2.1957611223678617, + "learning_rate": 1.22795712957919e-06, + "loss": 0.0352, + "step": 7779 + }, + { + "epoch": 5.596115806509621, + "grad_norm": 1.3357577277984263, + "learning_rate": 1.2276225343637396e-06, + "loss": 0.0086, + "step": 7780 + }, + { + "epoch": 5.596835101600432, + "grad_norm": 4.36645449410245, + "learning_rate": 1.227287953162588e-06, + "loss": 0.0745, + "step": 7781 + }, + { + "epoch": 5.597554396691242, + "grad_norm": 2.153027225273113, + "learning_rate": 1.22695338599295e-06, + "loss": 0.0327, + "step": 7782 + }, + { + "epoch": 5.598273691782054, + "grad_norm": 3.1007463560702635, + "learning_rate": 1.2266188328720399e-06, + "loss": 0.0744, + "step": 7783 + }, + { + "epoch": 5.598992986872864, + "grad_norm": 2.470972422058394, + "learning_rate": 1.2262842938170705e-06, + "loss": 0.0486, + "step": 7784 + }, + { + "epoch": 5.599712281963676, + "grad_norm": 6.00350927540214, + "learning_rate": 1.2259497688452545e-06, + "loss": 0.0964, + "step": 7785 + }, + { + "epoch": 5.600431577054486, + "grad_norm": 0.07709068357387217, + "learning_rate": 1.225615257973804e-06, + "loss": 0.0003, + "step": 7786 + }, + { + "epoch": 5.601150872145298, + "grad_norm": 2.719187039681179, + "learning_rate": 1.2252807612199303e-06, + "loss": 0.0676, + "step": 7787 + }, + { + "epoch": 5.601870167236108, + "grad_norm": 0.5591773579029653, + "learning_rate": 1.224946278600844e-06, + "loss": 0.0056, + "step": 7788 + }, + { + "epoch": 5.60258946232692, + "grad_norm": 1.4918587776176475, + "learning_rate": 1.224611810133754e-06, + "loss": 0.0235, + "step": 7789 + }, + { + "epoch": 5.6033087574177305, + "grad_norm": 2.60154952622805, + "learning_rate": 1.22427735583587e-06, + "loss": 0.03, + "step": 7790 + }, + { + "epoch": 5.604028052508542, + "grad_norm": 2.129158349041342, + "learning_rate": 1.2239429157243997e-06, + "loss": 0.0421, + "step": 7791 + }, + { + "epoch": 5.604747347599353, + "grad_norm": 1.673896239549214, + "learning_rate": 1.2236084898165512e-06, + "loss": 0.0219, + "step": 7792 + }, + { + "epoch": 5.605466642690164, + "grad_norm": 1.3700421685077258, + "learning_rate": 1.2232740781295314e-06, + "loss": 0.0276, + "step": 7793 + }, + { + "epoch": 5.606185937780975, + "grad_norm": 3.835095876205065, + "learning_rate": 1.222939680680546e-06, + "loss": 0.0621, + "step": 7794 + }, + { + "epoch": 5.606905232871785, + "grad_norm": 2.03574260326345, + "learning_rate": 1.2226052974868004e-06, + "loss": 0.03, + "step": 7795 + }, + { + "epoch": 5.607624527962597, + "grad_norm": 0.9606438074539635, + "learning_rate": 1.2222709285654986e-06, + "loss": 0.0023, + "step": 7796 + }, + { + "epoch": 5.608343823053407, + "grad_norm": 1.1288150139594153, + "learning_rate": 1.2219365739338455e-06, + "loss": 0.0142, + "step": 7797 + }, + { + "epoch": 5.609063118144219, + "grad_norm": 2.5095939032408925, + "learning_rate": 1.221602233609044e-06, + "loss": 0.0387, + "step": 7798 + }, + { + "epoch": 5.609782413235029, + "grad_norm": 1.6407840270016865, + "learning_rate": 1.221267907608296e-06, + "loss": 0.031, + "step": 7799 + }, + { + "epoch": 5.610501708325841, + "grad_norm": 3.024618164134947, + "learning_rate": 1.2209335959488038e-06, + "loss": 0.0438, + "step": 7800 + }, + { + "epoch": 5.611221003416651, + "grad_norm": 3.2454440037254333, + "learning_rate": 1.2205992986477674e-06, + "loss": 0.0525, + "step": 7801 + }, + { + "epoch": 5.611940298507463, + "grad_norm": 2.7674774547037595, + "learning_rate": 1.220265015722388e-06, + "loss": 0.0576, + "step": 7802 + }, + { + "epoch": 5.6126595935982735, + "grad_norm": 1.5567394859225523, + "learning_rate": 1.2199307471898643e-06, + "loss": 0.0309, + "step": 7803 + }, + { + "epoch": 5.613378888689085, + "grad_norm": 2.076148888728428, + "learning_rate": 1.2195964930673957e-06, + "loss": 0.0233, + "step": 7804 + }, + { + "epoch": 5.614098183779896, + "grad_norm": 2.715789277669573, + "learning_rate": 1.2192622533721796e-06, + "loss": 0.027, + "step": 7805 + }, + { + "epoch": 5.614817478870707, + "grad_norm": 3.0412446246940257, + "learning_rate": 1.2189280281214128e-06, + "loss": 0.0627, + "step": 7806 + }, + { + "epoch": 5.615536773961518, + "grad_norm": 2.074976963171042, + "learning_rate": 1.2185938173322933e-06, + "loss": 0.0176, + "step": 7807 + }, + { + "epoch": 5.616256069052329, + "grad_norm": 1.622412784318078, + "learning_rate": 1.2182596210220155e-06, + "loss": 0.0114, + "step": 7808 + }, + { + "epoch": 5.61697536414314, + "grad_norm": 1.4577036748497243, + "learning_rate": 1.2179254392077747e-06, + "loss": 0.0222, + "step": 7809 + }, + { + "epoch": 5.617694659233951, + "grad_norm": 2.401919320195346, + "learning_rate": 1.2175912719067655e-06, + "loss": 0.0062, + "step": 7810 + }, + { + "epoch": 5.618413954324762, + "grad_norm": 0.03481321065580161, + "learning_rate": 1.2172571191361808e-06, + "loss": 0.0001, + "step": 7811 + }, + { + "epoch": 5.619133249415572, + "grad_norm": 3.609835270189955, + "learning_rate": 1.2169229809132132e-06, + "loss": 0.0182, + "step": 7812 + }, + { + "epoch": 5.619852544506384, + "grad_norm": 3.2245543577057734, + "learning_rate": 1.2165888572550559e-06, + "loss": 0.0647, + "step": 7813 + }, + { + "epoch": 5.620571839597194, + "grad_norm": 2.314889164035302, + "learning_rate": 1.2162547481788994e-06, + "loss": 0.064, + "step": 7814 + }, + { + "epoch": 5.621291134688006, + "grad_norm": 2.8859619867091904, + "learning_rate": 1.215920653701934e-06, + "loss": 0.0386, + "step": 7815 + }, + { + "epoch": 5.6220104297788165, + "grad_norm": 0.3553185280211063, + "learning_rate": 1.21558657384135e-06, + "loss": 0.0013, + "step": 7816 + }, + { + "epoch": 5.622729724869628, + "grad_norm": 0.8893236737506396, + "learning_rate": 1.2152525086143351e-06, + "loss": 0.0067, + "step": 7817 + }, + { + "epoch": 5.6234490199604386, + "grad_norm": 2.366035137932432, + "learning_rate": 1.2149184580380793e-06, + "loss": 0.033, + "step": 7818 + }, + { + "epoch": 5.62416831505125, + "grad_norm": 4.275846592536599, + "learning_rate": 1.2145844221297693e-06, + "loss": 0.059, + "step": 7819 + }, + { + "epoch": 5.624887610142061, + "grad_norm": 3.474848327023783, + "learning_rate": 1.2142504009065915e-06, + "loss": 0.057, + "step": 7820 + }, + { + "epoch": 5.625606905232872, + "grad_norm": 1.527989328234736, + "learning_rate": 1.2139163943857323e-06, + "loss": 0.0139, + "step": 7821 + }, + { + "epoch": 5.626326200323683, + "grad_norm": 3.4205016025888026, + "learning_rate": 1.2135824025843765e-06, + "loss": 0.0648, + "step": 7822 + }, + { + "epoch": 5.627045495414494, + "grad_norm": 0.6729139871861205, + "learning_rate": 1.2132484255197094e-06, + "loss": 0.0007, + "step": 7823 + }, + { + "epoch": 5.627764790505305, + "grad_norm": 0.0033415353193787804, + "learning_rate": 1.212914463208914e-06, + "loss": 0.0, + "step": 7824 + }, + { + "epoch": 5.628484085596115, + "grad_norm": 8.217337478841063, + "learning_rate": 1.2125805156691736e-06, + "loss": 0.1562, + "step": 7825 + }, + { + "epoch": 5.629203380686927, + "grad_norm": 5.207323242978408, + "learning_rate": 1.21224658291767e-06, + "loss": 0.0658, + "step": 7826 + }, + { + "epoch": 5.629922675777738, + "grad_norm": 2.0237714855965145, + "learning_rate": 1.2119126649715843e-06, + "loss": 0.0294, + "step": 7827 + }, + { + "epoch": 5.630641970868549, + "grad_norm": 0.341479326329603, + "learning_rate": 1.2115787618480984e-06, + "loss": 0.0024, + "step": 7828 + }, + { + "epoch": 5.6313612659593595, + "grad_norm": 1.8469252039724493, + "learning_rate": 1.2112448735643914e-06, + "loss": 0.0195, + "step": 7829 + }, + { + "epoch": 5.632080561050171, + "grad_norm": 1.1924590972084832, + "learning_rate": 1.2109110001376427e-06, + "loss": 0.0138, + "step": 7830 + }, + { + "epoch": 5.6327998561409816, + "grad_norm": 0.5370736376739041, + "learning_rate": 1.2105771415850305e-06, + "loss": 0.0035, + "step": 7831 + }, + { + "epoch": 5.633519151231793, + "grad_norm": 1.7160955084309892, + "learning_rate": 1.2102432979237325e-06, + "loss": 0.0272, + "step": 7832 + }, + { + "epoch": 5.634238446322604, + "grad_norm": 2.950284882898954, + "learning_rate": 1.2099094691709247e-06, + "loss": 0.0406, + "step": 7833 + }, + { + "epoch": 5.634957741413415, + "grad_norm": 3.123942924408305, + "learning_rate": 1.2095756553437845e-06, + "loss": 0.0462, + "step": 7834 + }, + { + "epoch": 5.635677036504226, + "grad_norm": 2.9093752357617397, + "learning_rate": 1.2092418564594868e-06, + "loss": 0.0478, + "step": 7835 + }, + { + "epoch": 5.636396331595037, + "grad_norm": 4.685395310852276, + "learning_rate": 1.2089080725352059e-06, + "loss": 0.0545, + "step": 7836 + }, + { + "epoch": 5.637115626685848, + "grad_norm": 4.022252011705735, + "learning_rate": 1.2085743035881157e-06, + "loss": 0.0916, + "step": 7837 + }, + { + "epoch": 5.637834921776659, + "grad_norm": 0.016925370742192827, + "learning_rate": 1.2082405496353888e-06, + "loss": 0.0, + "step": 7838 + }, + { + "epoch": 5.63855421686747, + "grad_norm": 0.6579429614906833, + "learning_rate": 1.2079068106941982e-06, + "loss": 0.0055, + "step": 7839 + }, + { + "epoch": 5.639273511958281, + "grad_norm": 2.582673268580817, + "learning_rate": 1.2075730867817147e-06, + "loss": 0.0389, + "step": 7840 + }, + { + "epoch": 5.639992807049092, + "grad_norm": 4.808876512742077, + "learning_rate": 1.2072393779151097e-06, + "loss": 0.1045, + "step": 7841 + }, + { + "epoch": 5.6407121021399025, + "grad_norm": 1.0647245753175043, + "learning_rate": 1.206905684111552e-06, + "loss": 0.0105, + "step": 7842 + }, + { + "epoch": 5.641431397230714, + "grad_norm": 0.014227356178043658, + "learning_rate": 1.2065720053882117e-06, + "loss": 0.0001, + "step": 7843 + }, + { + "epoch": 5.6421506923215246, + "grad_norm": 0.764160833728507, + "learning_rate": 1.2062383417622566e-06, + "loss": 0.0045, + "step": 7844 + }, + { + "epoch": 5.642869987412336, + "grad_norm": 3.543807748816618, + "learning_rate": 1.2059046932508547e-06, + "loss": 0.0595, + "step": 7845 + }, + { + "epoch": 5.643589282503147, + "grad_norm": 1.9257392895770402, + "learning_rate": 1.2055710598711728e-06, + "loss": 0.0322, + "step": 7846 + }, + { + "epoch": 5.644308577593958, + "grad_norm": 2.4648373157610526, + "learning_rate": 1.2052374416403765e-06, + "loss": 0.0192, + "step": 7847 + }, + { + "epoch": 5.645027872684769, + "grad_norm": 4.289695775476988, + "learning_rate": 1.2049038385756305e-06, + "loss": 0.0275, + "step": 7848 + }, + { + "epoch": 5.64574716777558, + "grad_norm": 4.224846429564521, + "learning_rate": 1.204570250694101e-06, + "loss": 0.1134, + "step": 7849 + }, + { + "epoch": 5.646466462866391, + "grad_norm": 1.718965570796652, + "learning_rate": 1.2042366780129506e-06, + "loss": 0.0234, + "step": 7850 + }, + { + "epoch": 5.647185757957202, + "grad_norm": 0.2935105818633598, + "learning_rate": 1.2039031205493423e-06, + "loss": 0.0006, + "step": 7851 + }, + { + "epoch": 5.647905053048013, + "grad_norm": 4.017891463466948, + "learning_rate": 1.203569578320438e-06, + "loss": 0.082, + "step": 7852 + }, + { + "epoch": 5.648624348138824, + "grad_norm": 0.5110664004977995, + "learning_rate": 1.2032360513433996e-06, + "loss": 0.0036, + "step": 7853 + }, + { + "epoch": 5.649343643229635, + "grad_norm": 4.744195719999465, + "learning_rate": 1.2029025396353867e-06, + "loss": 0.0636, + "step": 7854 + }, + { + "epoch": 5.6500629383204455, + "grad_norm": 3.6413086573506894, + "learning_rate": 1.2025690432135603e-06, + "loss": 0.0493, + "step": 7855 + }, + { + "epoch": 5.650782233411257, + "grad_norm": 2.5042906443890924, + "learning_rate": 1.2022355620950785e-06, + "loss": 0.0493, + "step": 7856 + }, + { + "epoch": 5.651501528502068, + "grad_norm": 0.22553888070210285, + "learning_rate": 1.2019020962971e-06, + "loss": 0.0004, + "step": 7857 + }, + { + "epoch": 5.652220823592879, + "grad_norm": 4.248798800296586, + "learning_rate": 1.2015686458367817e-06, + "loss": 0.0649, + "step": 7858 + }, + { + "epoch": 5.65294011868369, + "grad_norm": 2.7683317996682426, + "learning_rate": 1.20123521073128e-06, + "loss": 0.0377, + "step": 7859 + }, + { + "epoch": 5.653659413774501, + "grad_norm": 1.873060495363098, + "learning_rate": 1.200901790997752e-06, + "loss": 0.0223, + "step": 7860 + }, + { + "epoch": 5.654378708865312, + "grad_norm": 2.777544779220471, + "learning_rate": 1.2005683866533515e-06, + "loss": 0.046, + "step": 7861 + }, + { + "epoch": 5.655098003956123, + "grad_norm": 1.4398891063221493, + "learning_rate": 1.2002349977152333e-06, + "loss": 0.0199, + "step": 7862 + }, + { + "epoch": 5.655817299046934, + "grad_norm": 0.05507673064960859, + "learning_rate": 1.1999016242005504e-06, + "loss": 0.0002, + "step": 7863 + }, + { + "epoch": 5.656536594137745, + "grad_norm": 0.6310287417846425, + "learning_rate": 1.1995682661264555e-06, + "loss": 0.0051, + "step": 7864 + }, + { + "epoch": 5.657255889228556, + "grad_norm": 0.0034576939545976093, + "learning_rate": 1.199234923510101e-06, + "loss": 0.0, + "step": 7865 + }, + { + "epoch": 5.657975184319367, + "grad_norm": 3.5591816974234556, + "learning_rate": 1.1989015963686384e-06, + "loss": 0.0448, + "step": 7866 + }, + { + "epoch": 5.658694479410178, + "grad_norm": 2.959568043171761, + "learning_rate": 1.1985682847192167e-06, + "loss": 0.0741, + "step": 7867 + }, + { + "epoch": 5.659413774500989, + "grad_norm": 0.02636665741957143, + "learning_rate": 1.1982349885789857e-06, + "loss": 0.0001, + "step": 7868 + }, + { + "epoch": 5.6601330695918, + "grad_norm": 6.81086371694984, + "learning_rate": 1.1979017079650938e-06, + "loss": 0.1535, + "step": 7869 + }, + { + "epoch": 5.660852364682611, + "grad_norm": 4.197209948700095, + "learning_rate": 1.19756844289469e-06, + "loss": 0.0724, + "step": 7870 + }, + { + "epoch": 5.661571659773422, + "grad_norm": 5.88603247435558, + "learning_rate": 1.1972351933849208e-06, + "loss": 0.1279, + "step": 7871 + }, + { + "epoch": 5.662290954864233, + "grad_norm": 1.607901670558626, + "learning_rate": 1.1969019594529322e-06, + "loss": 0.0075, + "step": 7872 + }, + { + "epoch": 5.663010249955044, + "grad_norm": 2.2727663271567073, + "learning_rate": 1.1965687411158698e-06, + "loss": 0.0233, + "step": 7873 + }, + { + "epoch": 5.663729545045855, + "grad_norm": 4.833705041099494, + "learning_rate": 1.1962355383908786e-06, + "loss": 0.0703, + "step": 7874 + }, + { + "epoch": 5.664448840136666, + "grad_norm": 2.01443817104122, + "learning_rate": 1.1959023512951016e-06, + "loss": 0.0305, + "step": 7875 + }, + { + "epoch": 5.665168135227477, + "grad_norm": 2.1287883513518815, + "learning_rate": 1.195569179845683e-06, + "loss": 0.0234, + "step": 7876 + }, + { + "epoch": 5.665887430318288, + "grad_norm": 2.578398540284161, + "learning_rate": 1.1952360240597643e-06, + "loss": 0.0328, + "step": 7877 + }, + { + "epoch": 5.666606725409099, + "grad_norm": 1.521987209700204, + "learning_rate": 1.1949028839544873e-06, + "loss": 0.0183, + "step": 7878 + }, + { + "epoch": 5.66732602049991, + "grad_norm": 1.5305112878349247, + "learning_rate": 1.1945697595469926e-06, + "loss": 0.0092, + "step": 7879 + }, + { + "epoch": 5.668045315590721, + "grad_norm": 4.798291506104338, + "learning_rate": 1.1942366508544194e-06, + "loss": 0.0774, + "step": 7880 + }, + { + "epoch": 5.668764610681532, + "grad_norm": 2.7012626434204083, + "learning_rate": 1.1939035578939075e-06, + "loss": 0.032, + "step": 7881 + }, + { + "epoch": 5.669483905772343, + "grad_norm": 7.232651105581225, + "learning_rate": 1.1935704806825954e-06, + "loss": 0.1077, + "step": 7882 + }, + { + "epoch": 5.670203200863154, + "grad_norm": 1.7517866066590138, + "learning_rate": 1.1932374192376194e-06, + "loss": 0.0161, + "step": 7883 + }, + { + "epoch": 5.670922495953965, + "grad_norm": 1.3336707088986923, + "learning_rate": 1.192904373576117e-06, + "loss": 0.0135, + "step": 7884 + }, + { + "epoch": 5.6716417910447765, + "grad_norm": 1.0313351958977925, + "learning_rate": 1.192571343715223e-06, + "loss": 0.0016, + "step": 7885 + }, + { + "epoch": 5.672361086135587, + "grad_norm": 3.0635930088158774, + "learning_rate": 1.1922383296720736e-06, + "loss": 0.0471, + "step": 7886 + }, + { + "epoch": 5.673080381226399, + "grad_norm": 3.0163943792115546, + "learning_rate": 1.1919053314638024e-06, + "loss": 0.061, + "step": 7887 + }, + { + "epoch": 5.673799676317209, + "grad_norm": 1.930356084278348, + "learning_rate": 1.1915723491075428e-06, + "loss": 0.0168, + "step": 7888 + }, + { + "epoch": 5.67451897140802, + "grad_norm": 0.04272022445788369, + "learning_rate": 1.1912393826204271e-06, + "loss": 0.0001, + "step": 7889 + }, + { + "epoch": 5.675238266498831, + "grad_norm": 4.340950512172067, + "learning_rate": 1.1909064320195871e-06, + "loss": 0.0889, + "step": 7890 + }, + { + "epoch": 5.675957561589642, + "grad_norm": 1.8863441688908682, + "learning_rate": 1.190573497322154e-06, + "loss": 0.0029, + "step": 7891 + }, + { + "epoch": 5.676676856680453, + "grad_norm": 0.031143970883467625, + "learning_rate": 1.1902405785452576e-06, + "loss": 0.0, + "step": 7892 + }, + { + "epoch": 5.677396151771264, + "grad_norm": 0.09114734962813811, + "learning_rate": 1.1899076757060274e-06, + "loss": 0.0004, + "step": 7893 + }, + { + "epoch": 5.678115446862075, + "grad_norm": 0.744938696605259, + "learning_rate": 1.1895747888215913e-06, + "loss": 0.0049, + "step": 7894 + }, + { + "epoch": 5.678834741952886, + "grad_norm": 2.5271695767831894, + "learning_rate": 1.1892419179090774e-06, + "loss": 0.0341, + "step": 7895 + }, + { + "epoch": 5.679554037043697, + "grad_norm": 4.450570006401814, + "learning_rate": 1.188909062985612e-06, + "loss": 0.053, + "step": 7896 + }, + { + "epoch": 5.680273332134508, + "grad_norm": 1.60540444861197, + "learning_rate": 1.188576224068322e-06, + "loss": 0.0155, + "step": 7897 + }, + { + "epoch": 5.6809926272253195, + "grad_norm": 4.714747965722082, + "learning_rate": 1.1882434011743316e-06, + "loss": 0.0877, + "step": 7898 + }, + { + "epoch": 5.68171192231613, + "grad_norm": 2.309874082635509, + "learning_rate": 1.1879105943207656e-06, + "loss": 0.0304, + "step": 7899 + }, + { + "epoch": 5.682431217406942, + "grad_norm": 0.3786691464529665, + "learning_rate": 1.1875778035247474e-06, + "loss": 0.0007, + "step": 7900 + }, + { + "epoch": 5.683150512497752, + "grad_norm": 3.4823440258807663, + "learning_rate": 1.187245028803399e-06, + "loss": 0.0585, + "step": 7901 + }, + { + "epoch": 5.683869807588563, + "grad_norm": 3.3396610035688186, + "learning_rate": 1.1869122701738438e-06, + "loss": 0.026, + "step": 7902 + }, + { + "epoch": 5.684589102679374, + "grad_norm": 1.2875272630255925, + "learning_rate": 1.1865795276532017e-06, + "loss": 0.0098, + "step": 7903 + }, + { + "epoch": 5.685308397770186, + "grad_norm": 2.46630498225818, + "learning_rate": 1.1862468012585928e-06, + "loss": 0.0536, + "step": 7904 + }, + { + "epoch": 5.686027692860996, + "grad_norm": 3.5843404055356904, + "learning_rate": 1.185914091007137e-06, + "loss": 0.0565, + "step": 7905 + }, + { + "epoch": 5.686746987951807, + "grad_norm": 2.6779744933153586, + "learning_rate": 1.185581396915952e-06, + "loss": 0.069, + "step": 7906 + }, + { + "epoch": 5.687466283042618, + "grad_norm": 0.0010535713874001201, + "learning_rate": 1.1852487190021567e-06, + "loss": 0.0, + "step": 7907 + }, + { + "epoch": 5.688185578133429, + "grad_norm": 1.8488484857984813, + "learning_rate": 1.184916057282867e-06, + "loss": 0.0168, + "step": 7908 + }, + { + "epoch": 5.68890487322424, + "grad_norm": 5.454141412388462, + "learning_rate": 1.1845834117751996e-06, + "loss": 0.0668, + "step": 7909 + }, + { + "epoch": 5.689624168315051, + "grad_norm": 1.5503145785360486, + "learning_rate": 1.1842507824962694e-06, + "loss": 0.0108, + "step": 7910 + }, + { + "epoch": 5.6903434634058625, + "grad_norm": 0.6558820934087332, + "learning_rate": 1.1839181694631904e-06, + "loss": 0.0011, + "step": 7911 + }, + { + "epoch": 5.691062758496673, + "grad_norm": 2.343740990062217, + "learning_rate": 1.1835855726930766e-06, + "loss": 0.0219, + "step": 7912 + }, + { + "epoch": 5.691782053587485, + "grad_norm": 1.4001044742323838, + "learning_rate": 1.1832529922030407e-06, + "loss": 0.0144, + "step": 7913 + }, + { + "epoch": 5.692501348678295, + "grad_norm": 2.938051062391753, + "learning_rate": 1.1829204280101942e-06, + "loss": 0.0454, + "step": 7914 + }, + { + "epoch": 5.693220643769107, + "grad_norm": 2.7650136933916, + "learning_rate": 1.1825878801316482e-06, + "loss": 0.0316, + "step": 7915 + }, + { + "epoch": 5.693939938859917, + "grad_norm": 2.9571809074668023, + "learning_rate": 1.182255348584513e-06, + "loss": 0.0337, + "step": 7916 + }, + { + "epoch": 5.694659233950729, + "grad_norm": 1.1654578898798915, + "learning_rate": 1.181922833385898e-06, + "loss": 0.0027, + "step": 7917 + }, + { + "epoch": 5.695378529041539, + "grad_norm": 2.0690445314708947, + "learning_rate": 1.1815903345529117e-06, + "loss": 0.0265, + "step": 7918 + }, + { + "epoch": 5.69609782413235, + "grad_norm": 0.37169023919833954, + "learning_rate": 1.1812578521026615e-06, + "loss": 0.0007, + "step": 7919 + }, + { + "epoch": 5.696817119223161, + "grad_norm": 1.87705359492762, + "learning_rate": 1.1809253860522543e-06, + "loss": 0.0326, + "step": 7920 + }, + { + "epoch": 5.697536414313972, + "grad_norm": 3.952309997594218, + "learning_rate": 1.1805929364187964e-06, + "loss": 0.0388, + "step": 7921 + }, + { + "epoch": 5.698255709404783, + "grad_norm": 0.2054905385564442, + "learning_rate": 1.1802605032193916e-06, + "loss": 0.0004, + "step": 7922 + }, + { + "epoch": 5.698975004495594, + "grad_norm": 2.795279051396003, + "learning_rate": 1.1799280864711461e-06, + "loss": 0.0344, + "step": 7923 + }, + { + "epoch": 5.6996942995864055, + "grad_norm": 4.69639715504311, + "learning_rate": 1.1795956861911621e-06, + "loss": 0.0899, + "step": 7924 + }, + { + "epoch": 5.700413594677216, + "grad_norm": 1.9896036936129309, + "learning_rate": 1.1792633023965433e-06, + "loss": 0.0113, + "step": 7925 + }, + { + "epoch": 5.701132889768028, + "grad_norm": 0.09245432932001309, + "learning_rate": 1.17893093510439e-06, + "loss": 0.0005, + "step": 7926 + }, + { + "epoch": 5.701852184858838, + "grad_norm": 3.8495756721242764, + "learning_rate": 1.1785985843318031e-06, + "loss": 0.0634, + "step": 7927 + }, + { + "epoch": 5.70257147994965, + "grad_norm": 4.148901303113292, + "learning_rate": 1.1782662500958841e-06, + "loss": 0.0275, + "step": 7928 + }, + { + "epoch": 5.70329077504046, + "grad_norm": 5.369712629345001, + "learning_rate": 1.177933932413731e-06, + "loss": 0.1032, + "step": 7929 + }, + { + "epoch": 5.704010070131272, + "grad_norm": 4.628361825010907, + "learning_rate": 1.1776016313024427e-06, + "loss": 0.062, + "step": 7930 + }, + { + "epoch": 5.704729365222082, + "grad_norm": 3.27568900486773, + "learning_rate": 1.1772693467791163e-06, + "loss": 0.0327, + "step": 7931 + }, + { + "epoch": 5.705448660312893, + "grad_norm": 4.553661905349076, + "learning_rate": 1.1769370788608482e-06, + "loss": 0.0172, + "step": 7932 + }, + { + "epoch": 5.706167955403704, + "grad_norm": 0.0056352083851909825, + "learning_rate": 1.1766048275647348e-06, + "loss": 0.0, + "step": 7933 + }, + { + "epoch": 5.706887250494516, + "grad_norm": 7.206487053087452, + "learning_rate": 1.1762725929078708e-06, + "loss": 0.1505, + "step": 7934 + }, + { + "epoch": 5.707606545585326, + "grad_norm": 4.535441492844, + "learning_rate": 1.1759403749073501e-06, + "loss": 0.0719, + "step": 7935 + }, + { + "epoch": 5.708325840676137, + "grad_norm": 2.114711585063322, + "learning_rate": 1.175608173580266e-06, + "loss": 0.0061, + "step": 7936 + }, + { + "epoch": 5.7090451357669485, + "grad_norm": 10.9562665781455, + "learning_rate": 1.1752759889437102e-06, + "loss": 0.1736, + "step": 7937 + }, + { + "epoch": 5.709764430857759, + "grad_norm": 4.019854331487344, + "learning_rate": 1.1749438210147759e-06, + "loss": 0.0337, + "step": 7938 + }, + { + "epoch": 5.710483725948571, + "grad_norm": 0.945671877105733, + "learning_rate": 1.1746116698105521e-06, + "loss": 0.0019, + "step": 7939 + }, + { + "epoch": 5.711203021039381, + "grad_norm": 0.24376262452195094, + "learning_rate": 1.1742795353481292e-06, + "loss": 0.0004, + "step": 7940 + }, + { + "epoch": 5.711922316130193, + "grad_norm": 1.5775040931175264, + "learning_rate": 1.1739474176445961e-06, + "loss": 0.017, + "step": 7941 + }, + { + "epoch": 5.712641611221003, + "grad_norm": 4.325732366087104, + "learning_rate": 1.1736153167170404e-06, + "loss": 0.0963, + "step": 7942 + }, + { + "epoch": 5.713360906311815, + "grad_norm": 0.5131660521623692, + "learning_rate": 1.1732832325825492e-06, + "loss": 0.0037, + "step": 7943 + }, + { + "epoch": 5.714080201402625, + "grad_norm": 1.7511523061644945, + "learning_rate": 1.17295116525821e-06, + "loss": 0.0246, + "step": 7944 + }, + { + "epoch": 5.714799496493437, + "grad_norm": 0.22860315492215652, + "learning_rate": 1.1726191147611073e-06, + "loss": 0.0006, + "step": 7945 + }, + { + "epoch": 5.715518791584247, + "grad_norm": 6.654895157306521, + "learning_rate": 1.1722870811083258e-06, + "loss": 0.131, + "step": 7946 + }, + { + "epoch": 5.716238086675059, + "grad_norm": 3.3320196822831702, + "learning_rate": 1.1719550643169494e-06, + "loss": 0.047, + "step": 7947 + }, + { + "epoch": 5.716957381765869, + "grad_norm": 4.530000886946475, + "learning_rate": 1.1716230644040603e-06, + "loss": 0.0472, + "step": 7948 + }, + { + "epoch": 5.71767667685668, + "grad_norm": 2.903162268738178, + "learning_rate": 1.1712910813867413e-06, + "loss": 0.0389, + "step": 7949 + }, + { + "epoch": 5.7183959719474915, + "grad_norm": 2.9832669364706828, + "learning_rate": 1.1709591152820733e-06, + "loss": 0.0424, + "step": 7950 + }, + { + "epoch": 5.719115267038302, + "grad_norm": 2.227701480563803, + "learning_rate": 1.1706271661071363e-06, + "loss": 0.0031, + "step": 7951 + }, + { + "epoch": 5.719834562129114, + "grad_norm": 1.7520930086741107, + "learning_rate": 1.17029523387901e-06, + "loss": 0.0133, + "step": 7952 + }, + { + "epoch": 5.720553857219924, + "grad_norm": 0.22096470969421592, + "learning_rate": 1.1699633186147722e-06, + "loss": 0.0006, + "step": 7953 + }, + { + "epoch": 5.721273152310736, + "grad_norm": 3.446405042460787, + "learning_rate": 1.1696314203315014e-06, + "loss": 0.0541, + "step": 7954 + }, + { + "epoch": 5.721992447401546, + "grad_norm": 5.982588210448654, + "learning_rate": 1.1692995390462741e-06, + "loss": 0.0897, + "step": 7955 + }, + { + "epoch": 5.722711742492358, + "grad_norm": 1.2067546031815235, + "learning_rate": 1.168967674776166e-06, + "loss": 0.0067, + "step": 7956 + }, + { + "epoch": 5.723431037583168, + "grad_norm": 5.48952783523368, + "learning_rate": 1.1686358275382521e-06, + "loss": 0.1214, + "step": 7957 + }, + { + "epoch": 5.72415033267398, + "grad_norm": 2.8227194603252728, + "learning_rate": 1.1683039973496057e-06, + "loss": 0.0594, + "step": 7958 + }, + { + "epoch": 5.72486962776479, + "grad_norm": 4.179659852110548, + "learning_rate": 1.1679721842273017e-06, + "loss": 0.0849, + "step": 7959 + }, + { + "epoch": 5.725588922855602, + "grad_norm": 0.08711089606877226, + "learning_rate": 1.167640388188412e-06, + "loss": 0.0002, + "step": 7960 + }, + { + "epoch": 5.726308217946412, + "grad_norm": 4.947029882536398, + "learning_rate": 1.1673086092500075e-06, + "loss": 0.119, + "step": 7961 + }, + { + "epoch": 5.727027513037224, + "grad_norm": 7.445964705856657, + "learning_rate": 1.1669768474291592e-06, + "loss": 0.1137, + "step": 7962 + }, + { + "epoch": 5.7277468081280345, + "grad_norm": 4.059894834109739, + "learning_rate": 1.1666451027429365e-06, + "loss": 0.0869, + "step": 7963 + }, + { + "epoch": 5.728466103218846, + "grad_norm": 2.7816056704952747, + "learning_rate": 1.1663133752084079e-06, + "loss": 0.0342, + "step": 7964 + }, + { + "epoch": 5.729185398309657, + "grad_norm": 3.5095002335565413, + "learning_rate": 1.1659816648426427e-06, + "loss": 0.0711, + "step": 7965 + }, + { + "epoch": 5.729904693400467, + "grad_norm": 1.1637335020857522, + "learning_rate": 1.165649971662707e-06, + "loss": 0.0114, + "step": 7966 + }, + { + "epoch": 5.730623988491279, + "grad_norm": 2.298335591618006, + "learning_rate": 1.165318295685667e-06, + "loss": 0.0229, + "step": 7967 + }, + { + "epoch": 5.731343283582089, + "grad_norm": 0.7973126420317768, + "learning_rate": 1.1649866369285885e-06, + "loss": 0.0052, + "step": 7968 + }, + { + "epoch": 5.732062578672901, + "grad_norm": 1.5207713318206644, + "learning_rate": 1.1646549954085354e-06, + "loss": 0.0254, + "step": 7969 + }, + { + "epoch": 5.732781873763711, + "grad_norm": 0.16741434156739676, + "learning_rate": 1.1643233711425715e-06, + "loss": 0.0005, + "step": 7970 + }, + { + "epoch": 5.733501168854523, + "grad_norm": 2.6530128329089915, + "learning_rate": 1.1639917641477596e-06, + "loss": 0.0441, + "step": 7971 + }, + { + "epoch": 5.734220463945333, + "grad_norm": 0.06464983986571868, + "learning_rate": 1.1636601744411612e-06, + "loss": 0.0002, + "step": 7972 + }, + { + "epoch": 5.734939759036145, + "grad_norm": 3.30438083552059, + "learning_rate": 1.1633286020398372e-06, + "loss": 0.059, + "step": 7973 + }, + { + "epoch": 5.735659054126955, + "grad_norm": 3.9934242622186154, + "learning_rate": 1.1629970469608473e-06, + "loss": 0.1054, + "step": 7974 + }, + { + "epoch": 5.736378349217767, + "grad_norm": 0.3072843928316565, + "learning_rate": 1.1626655092212514e-06, + "loss": 0.0021, + "step": 7975 + }, + { + "epoch": 5.7370976443085775, + "grad_norm": 2.0412971961555826, + "learning_rate": 1.162333988838107e-06, + "loss": 0.022, + "step": 7976 + }, + { + "epoch": 5.737816939399389, + "grad_norm": 5.448626052921805, + "learning_rate": 1.1620024858284718e-06, + "loss": 0.0729, + "step": 7977 + }, + { + "epoch": 5.7385362344902, + "grad_norm": 2.8159639359583637, + "learning_rate": 1.161671000209402e-06, + "loss": 0.0439, + "step": 7978 + }, + { + "epoch": 5.73925552958101, + "grad_norm": 3.7311362532656505, + "learning_rate": 1.1613395319979528e-06, + "loss": 0.0583, + "step": 7979 + }, + { + "epoch": 5.739974824671822, + "grad_norm": 0.2899932022017597, + "learning_rate": 1.1610080812111793e-06, + "loss": 0.0019, + "step": 7980 + }, + { + "epoch": 5.740694119762633, + "grad_norm": 2.6359644251906906, + "learning_rate": 1.1606766478661355e-06, + "loss": 0.0055, + "step": 7981 + }, + { + "epoch": 5.741413414853444, + "grad_norm": 3.912638055408291, + "learning_rate": 1.1603452319798735e-06, + "loss": 0.0408, + "step": 7982 + }, + { + "epoch": 5.742132709944254, + "grad_norm": 0.05092465951638287, + "learning_rate": 1.1600138335694459e-06, + "loss": 0.0001, + "step": 7983 + }, + { + "epoch": 5.742852005035066, + "grad_norm": 2.2021607497013385, + "learning_rate": 1.1596824526519034e-06, + "loss": 0.0212, + "step": 7984 + }, + { + "epoch": 5.743571300125876, + "grad_norm": 5.364399590025706, + "learning_rate": 1.1593510892442953e-06, + "loss": 0.1148, + "step": 7985 + }, + { + "epoch": 5.744290595216688, + "grad_norm": 3.9052190031501883, + "learning_rate": 1.1590197433636723e-06, + "loss": 0.037, + "step": 7986 + }, + { + "epoch": 5.745009890307498, + "grad_norm": 1.797131922531542, + "learning_rate": 1.1586884150270822e-06, + "loss": 0.0156, + "step": 7987 + }, + { + "epoch": 5.74572918539831, + "grad_norm": 3.3403361192961336, + "learning_rate": 1.158357104251572e-06, + "loss": 0.0797, + "step": 7988 + }, + { + "epoch": 5.7464484804891205, + "grad_norm": 0.0730018969769519, + "learning_rate": 1.1580258110541888e-06, + "loss": 0.0002, + "step": 7989 + }, + { + "epoch": 5.747167775579932, + "grad_norm": 3.454066945961893, + "learning_rate": 1.1576945354519775e-06, + "loss": 0.0781, + "step": 7990 + }, + { + "epoch": 5.747887070670743, + "grad_norm": 2.3308077246908603, + "learning_rate": 1.1573632774619836e-06, + "loss": 0.0521, + "step": 7991 + }, + { + "epoch": 5.748606365761554, + "grad_norm": 1.060912626798597, + "learning_rate": 1.1570320371012505e-06, + "loss": 0.0111, + "step": 7992 + }, + { + "epoch": 5.749325660852365, + "grad_norm": 2.0273063834775917, + "learning_rate": 1.1567008143868212e-06, + "loss": 0.0371, + "step": 7993 + }, + { + "epoch": 5.750044955943176, + "grad_norm": 2.5068662563454214, + "learning_rate": 1.1563696093357378e-06, + "loss": 0.0127, + "step": 7994 + }, + { + "epoch": 5.750764251033987, + "grad_norm": 5.623300543673391, + "learning_rate": 1.1560384219650403e-06, + "loss": 0.0892, + "step": 7995 + }, + { + "epoch": 5.751483546124797, + "grad_norm": 3.6898150459174337, + "learning_rate": 1.1557072522917705e-06, + "loss": 0.0719, + "step": 7996 + }, + { + "epoch": 5.752202841215609, + "grad_norm": 2.121086541641249, + "learning_rate": 1.1553761003329673e-06, + "loss": 0.0325, + "step": 7997 + }, + { + "epoch": 5.752922136306419, + "grad_norm": 2.716810738976316, + "learning_rate": 1.1550449661056684e-06, + "loss": 0.0633, + "step": 7998 + }, + { + "epoch": 5.753641431397231, + "grad_norm": 4.503438686756663, + "learning_rate": 1.1547138496269116e-06, + "loss": 0.0558, + "step": 7999 + }, + { + "epoch": 5.754360726488041, + "grad_norm": 6.932915936095513, + "learning_rate": 1.1543827509137328e-06, + "loss": 0.0958, + "step": 8000 + }, + { + "epoch": 5.755080021578853, + "grad_norm": 3.288181724513349, + "learning_rate": 1.1540516699831685e-06, + "loss": 0.0277, + "step": 8001 + }, + { + "epoch": 5.7557993166696635, + "grad_norm": 4.676089260800219, + "learning_rate": 1.1537206068522535e-06, + "loss": 0.0937, + "step": 8002 + }, + { + "epoch": 5.756518611760475, + "grad_norm": 3.720276485269355, + "learning_rate": 1.1533895615380207e-06, + "loss": 0.0799, + "step": 8003 + }, + { + "epoch": 5.757237906851286, + "grad_norm": 2.349464013102645, + "learning_rate": 1.1530585340575037e-06, + "loss": 0.045, + "step": 8004 + }, + { + "epoch": 5.757957201942097, + "grad_norm": 3.943513711427351, + "learning_rate": 1.1527275244277342e-06, + "loss": 0.1041, + "step": 8005 + }, + { + "epoch": 5.758676497032908, + "grad_norm": 3.0482979779344217, + "learning_rate": 1.1523965326657426e-06, + "loss": 0.0296, + "step": 8006 + }, + { + "epoch": 5.759395792123719, + "grad_norm": 0.44664517198077197, + "learning_rate": 1.1520655587885603e-06, + "loss": 0.0031, + "step": 8007 + }, + { + "epoch": 5.76011508721453, + "grad_norm": 4.4653985149028586, + "learning_rate": 1.1517346028132154e-06, + "loss": 0.0692, + "step": 8008 + }, + { + "epoch": 5.76083438230534, + "grad_norm": 4.452319604207101, + "learning_rate": 1.1514036647567368e-06, + "loss": 0.1397, + "step": 8009 + }, + { + "epoch": 5.761553677396152, + "grad_norm": 4.128605875235768, + "learning_rate": 1.1510727446361515e-06, + "loss": 0.0978, + "step": 8010 + }, + { + "epoch": 5.762272972486963, + "grad_norm": 3.7330804375785736, + "learning_rate": 1.1507418424684857e-06, + "loss": 0.025, + "step": 8011 + }, + { + "epoch": 5.762992267577774, + "grad_norm": 2.7843638805188973, + "learning_rate": 1.1504109582707653e-06, + "loss": 0.0354, + "step": 8012 + }, + { + "epoch": 5.763711562668584, + "grad_norm": 1.9025772297285057, + "learning_rate": 1.150080092060015e-06, + "loss": 0.0319, + "step": 8013 + }, + { + "epoch": 5.764430857759396, + "grad_norm": 1.4527450265564463, + "learning_rate": 1.1497492438532584e-06, + "loss": 0.0059, + "step": 8014 + }, + { + "epoch": 5.7651501528502065, + "grad_norm": 2.06504681903627, + "learning_rate": 1.1494184136675178e-06, + "loss": 0.0173, + "step": 8015 + }, + { + "epoch": 5.765869447941018, + "grad_norm": 0.31462944122229497, + "learning_rate": 1.1490876015198148e-06, + "loss": 0.0003, + "step": 8016 + }, + { + "epoch": 5.766588743031829, + "grad_norm": 5.318312983711191, + "learning_rate": 1.148756807427171e-06, + "loss": 0.0723, + "step": 8017 + }, + { + "epoch": 5.76730803812264, + "grad_norm": 2.784221381052251, + "learning_rate": 1.1484260314066065e-06, + "loss": 0.0499, + "step": 8018 + }, + { + "epoch": 5.768027333213451, + "grad_norm": 2.3549364638243278, + "learning_rate": 1.1480952734751395e-06, + "loss": 0.0156, + "step": 8019 + }, + { + "epoch": 5.768746628304262, + "grad_norm": 0.04632186057212055, + "learning_rate": 1.1477645336497889e-06, + "loss": 0.0001, + "step": 8020 + }, + { + "epoch": 5.769465923395073, + "grad_norm": 2.6127503240821905, + "learning_rate": 1.1474338119475704e-06, + "loss": 0.028, + "step": 8021 + }, + { + "epoch": 5.770185218485884, + "grad_norm": 3.0449368630714195, + "learning_rate": 1.1471031083855017e-06, + "loss": 0.0387, + "step": 8022 + }, + { + "epoch": 5.770904513576695, + "grad_norm": 0.7102793761095662, + "learning_rate": 1.1467724229805979e-06, + "loss": 0.0049, + "step": 8023 + }, + { + "epoch": 5.771623808667506, + "grad_norm": 3.9394771432893028, + "learning_rate": 1.1464417557498727e-06, + "loss": 0.0978, + "step": 8024 + }, + { + "epoch": 5.772343103758317, + "grad_norm": 1.0322757836594902, + "learning_rate": 1.1461111067103398e-06, + "loss": 0.0028, + "step": 8025 + }, + { + "epoch": 5.773062398849127, + "grad_norm": 4.365756939425793, + "learning_rate": 1.1457804758790117e-06, + "loss": 0.065, + "step": 8026 + }, + { + "epoch": 5.773781693939939, + "grad_norm": 0.18387923052933516, + "learning_rate": 1.1454498632728998e-06, + "loss": 0.0015, + "step": 8027 + }, + { + "epoch": 5.7745009890307495, + "grad_norm": 0.5279486357204373, + "learning_rate": 1.145119268909015e-06, + "loss": 0.0045, + "step": 8028 + }, + { + "epoch": 5.775220284121561, + "grad_norm": 0.5023013278616948, + "learning_rate": 1.144788692804367e-06, + "loss": 0.0007, + "step": 8029 + }, + { + "epoch": 5.7759395792123716, + "grad_norm": 4.694660974870802, + "learning_rate": 1.144458134975964e-06, + "loss": 0.0917, + "step": 8030 + }, + { + "epoch": 5.776658874303183, + "grad_norm": 3.6463575072936356, + "learning_rate": 1.1441275954408144e-06, + "loss": 0.0907, + "step": 8031 + }, + { + "epoch": 5.777378169393994, + "grad_norm": 2.019833597302573, + "learning_rate": 1.1437970742159239e-06, + "loss": 0.0148, + "step": 8032 + }, + { + "epoch": 5.778097464484805, + "grad_norm": 1.922771030747328, + "learning_rate": 1.1434665713183e-06, + "loss": 0.019, + "step": 8033 + }, + { + "epoch": 5.778816759575616, + "grad_norm": 2.3380965023455826, + "learning_rate": 1.143136086764947e-06, + "loss": 0.0372, + "step": 8034 + }, + { + "epoch": 5.779536054666427, + "grad_norm": 2.6168731280689292, + "learning_rate": 1.1428056205728684e-06, + "loss": 0.0383, + "step": 8035 + }, + { + "epoch": 5.780255349757238, + "grad_norm": 0.14550788783508134, + "learning_rate": 1.1424751727590677e-06, + "loss": 0.0003, + "step": 8036 + }, + { + "epoch": 5.780974644848049, + "grad_norm": 9.943023914366503, + "learning_rate": 1.1421447433405467e-06, + "loss": 0.1396, + "step": 8037 + }, + { + "epoch": 5.78169393993886, + "grad_norm": 2.200348534773601, + "learning_rate": 1.1418143323343073e-06, + "loss": 0.0277, + "step": 8038 + }, + { + "epoch": 5.782413235029671, + "grad_norm": 8.281358958403853, + "learning_rate": 1.1414839397573492e-06, + "loss": 0.1649, + "step": 8039 + }, + { + "epoch": 5.783132530120482, + "grad_norm": 0.31584662079174297, + "learning_rate": 1.1411535656266716e-06, + "loss": 0.0005, + "step": 8040 + }, + { + "epoch": 5.783851825211293, + "grad_norm": 4.116148622463772, + "learning_rate": 1.1408232099592732e-06, + "loss": 0.044, + "step": 8041 + }, + { + "epoch": 5.784571120302104, + "grad_norm": 1.8668098558041246, + "learning_rate": 1.1404928727721507e-06, + "loss": 0.0035, + "step": 8042 + }, + { + "epoch": 5.7852904153929146, + "grad_norm": 5.328946508268484, + "learning_rate": 1.1401625540823015e-06, + "loss": 0.1196, + "step": 8043 + }, + { + "epoch": 5.786009710483726, + "grad_norm": 6.029316589847938, + "learning_rate": 1.1398322539067205e-06, + "loss": 0.1512, + "step": 8044 + }, + { + "epoch": 5.786729005574537, + "grad_norm": 1.390302530666408, + "learning_rate": 1.1395019722624023e-06, + "loss": 0.0179, + "step": 8045 + }, + { + "epoch": 5.787448300665348, + "grad_norm": 2.9104882784150377, + "learning_rate": 1.1391717091663403e-06, + "loss": 0.0621, + "step": 8046 + }, + { + "epoch": 5.788167595756159, + "grad_norm": 2.809041875396382, + "learning_rate": 1.1388414646355276e-06, + "loss": 0.0337, + "step": 8047 + }, + { + "epoch": 5.78888689084697, + "grad_norm": 1.2556993519884754, + "learning_rate": 1.138511238686955e-06, + "loss": 0.023, + "step": 8048 + }, + { + "epoch": 5.789606185937781, + "grad_norm": 2.236732052543416, + "learning_rate": 1.138181031337614e-06, + "loss": 0.032, + "step": 8049 + }, + { + "epoch": 5.790325481028592, + "grad_norm": 2.666305700813717, + "learning_rate": 1.1378508426044943e-06, + "loss": 0.0554, + "step": 8050 + }, + { + "epoch": 5.791044776119403, + "grad_norm": 0.6399448475445596, + "learning_rate": 1.1375206725045842e-06, + "loss": 0.0054, + "step": 8051 + }, + { + "epoch": 5.791764071210214, + "grad_norm": 0.3691672106532632, + "learning_rate": 1.137190521054872e-06, + "loss": 0.0005, + "step": 8052 + }, + { + "epoch": 5.792483366301025, + "grad_norm": 4.82471929670556, + "learning_rate": 1.1368603882723437e-06, + "loss": 0.0773, + "step": 8053 + }, + { + "epoch": 5.793202661391836, + "grad_norm": 1.7202414056474664, + "learning_rate": 1.1365302741739867e-06, + "loss": 0.0227, + "step": 8054 + }, + { + "epoch": 5.793921956482647, + "grad_norm": 3.7060027860341203, + "learning_rate": 1.136200178776785e-06, + "loss": 0.094, + "step": 8055 + }, + { + "epoch": 5.7946412515734576, + "grad_norm": 0.9655938758277931, + "learning_rate": 1.135870102097723e-06, + "loss": 0.0168, + "step": 8056 + }, + { + "epoch": 5.795360546664269, + "grad_norm": 3.153692337079725, + "learning_rate": 1.135540044153783e-06, + "loss": 0.0486, + "step": 8057 + }, + { + "epoch": 5.7960798417550805, + "grad_norm": 2.650314726968641, + "learning_rate": 1.1352100049619472e-06, + "loss": 0.0538, + "step": 8058 + }, + { + "epoch": 5.796799136845891, + "grad_norm": 2.1519505249147177, + "learning_rate": 1.1348799845391974e-06, + "loss": 0.0352, + "step": 8059 + }, + { + "epoch": 5.797518431936702, + "grad_norm": 1.9694612559376747, + "learning_rate": 1.1345499829025135e-06, + "loss": 0.0286, + "step": 8060 + }, + { + "epoch": 5.798237727027513, + "grad_norm": 3.993019531097813, + "learning_rate": 1.1342200000688746e-06, + "loss": 0.0717, + "step": 8061 + }, + { + "epoch": 5.798957022118324, + "grad_norm": 2.648095724863305, + "learning_rate": 1.1338900360552589e-06, + "loss": 0.0204, + "step": 8062 + }, + { + "epoch": 5.799676317209135, + "grad_norm": 13.655309559334098, + "learning_rate": 1.1335600908786428e-06, + "loss": 0.0612, + "step": 8063 + }, + { + "epoch": 5.800395612299946, + "grad_norm": 0.09962672584958455, + "learning_rate": 1.1332301645560038e-06, + "loss": 0.0004, + "step": 8064 + }, + { + "epoch": 5.801114907390757, + "grad_norm": 2.092732606596833, + "learning_rate": 1.1329002571043167e-06, + "loss": 0.0438, + "step": 8065 + }, + { + "epoch": 5.801834202481568, + "grad_norm": 1.3065667616840158, + "learning_rate": 1.1325703685405557e-06, + "loss": 0.009, + "step": 8066 + }, + { + "epoch": 5.802553497572379, + "grad_norm": 3.699831940251251, + "learning_rate": 1.1322404988816943e-06, + "loss": 0.0625, + "step": 8067 + }, + { + "epoch": 5.80327279266319, + "grad_norm": 2.5407882982763943, + "learning_rate": 1.131910648144705e-06, + "loss": 0.0562, + "step": 8068 + }, + { + "epoch": 5.803992087754001, + "grad_norm": 1.950950588285628, + "learning_rate": 1.1315808163465584e-06, + "loss": 0.03, + "step": 8069 + }, + { + "epoch": 5.804711382844812, + "grad_norm": 2.7720040667349717, + "learning_rate": 1.131251003504226e-06, + "loss": 0.0196, + "step": 8070 + }, + { + "epoch": 5.8054306779356235, + "grad_norm": 0.09894486431206434, + "learning_rate": 1.1309212096346765e-06, + "loss": 0.0002, + "step": 8071 + }, + { + "epoch": 5.806149973026434, + "grad_norm": 0.14713976531797432, + "learning_rate": 1.130591434754879e-06, + "loss": 0.0004, + "step": 8072 + }, + { + "epoch": 5.806869268117245, + "grad_norm": 7.103782408209249, + "learning_rate": 1.1302616788818008e-06, + "loss": 0.1283, + "step": 8073 + }, + { + "epoch": 5.807588563208056, + "grad_norm": 1.5297898890924764, + "learning_rate": 1.1299319420324073e-06, + "loss": 0.0252, + "step": 8074 + }, + { + "epoch": 5.808307858298867, + "grad_norm": 3.1674733051580786, + "learning_rate": 1.1296022242236658e-06, + "loss": 0.0379, + "step": 8075 + }, + { + "epoch": 5.809027153389678, + "grad_norm": 1.217174943364638, + "learning_rate": 1.1292725254725399e-06, + "loss": 0.0106, + "step": 8076 + }, + { + "epoch": 5.809746448480489, + "grad_norm": 0.35408694434379656, + "learning_rate": 1.1289428457959935e-06, + "loss": 0.0005, + "step": 8077 + }, + { + "epoch": 5.8104657435713, + "grad_norm": 2.7627620162047286, + "learning_rate": 1.128613185210989e-06, + "loss": 0.0356, + "step": 8078 + }, + { + "epoch": 5.811185038662111, + "grad_norm": 0.14519863913808162, + "learning_rate": 1.128283543734488e-06, + "loss": 0.0004, + "step": 8079 + }, + { + "epoch": 5.811904333752922, + "grad_norm": 3.368618981824394, + "learning_rate": 1.1279539213834513e-06, + "loss": 0.072, + "step": 8080 + }, + { + "epoch": 5.812623628843733, + "grad_norm": 3.2431893047797367, + "learning_rate": 1.1276243181748385e-06, + "loss": 0.0581, + "step": 8081 + }, + { + "epoch": 5.813342923934544, + "grad_norm": 2.5557412358116225, + "learning_rate": 1.1272947341256082e-06, + "loss": 0.0503, + "step": 8082 + }, + { + "epoch": 5.814062219025355, + "grad_norm": 5.076409618851997, + "learning_rate": 1.1269651692527181e-06, + "loss": 0.0806, + "step": 8083 + }, + { + "epoch": 5.8147815141161665, + "grad_norm": 0.2002167289058098, + "learning_rate": 1.1266356235731246e-06, + "loss": 0.0006, + "step": 8084 + }, + { + "epoch": 5.815500809206977, + "grad_norm": 2.598288058424609, + "learning_rate": 1.1263060971037839e-06, + "loss": 0.0531, + "step": 8085 + }, + { + "epoch": 5.816220104297788, + "grad_norm": 1.9754499469951359, + "learning_rate": 1.1259765898616506e-06, + "loss": 0.0258, + "step": 8086 + }, + { + "epoch": 5.816939399388599, + "grad_norm": 0.08848776162996513, + "learning_rate": 1.1256471018636781e-06, + "loss": 0.0002, + "step": 8087 + }, + { + "epoch": 5.817658694479411, + "grad_norm": 6.4603420408232175, + "learning_rate": 1.1253176331268196e-06, + "loss": 0.0886, + "step": 8088 + }, + { + "epoch": 5.818377989570221, + "grad_norm": 0.59898225556208, + "learning_rate": 1.1249881836680261e-06, + "loss": 0.0045, + "step": 8089 + }, + { + "epoch": 5.819097284661032, + "grad_norm": 1.5093173632917087, + "learning_rate": 1.124658753504249e-06, + "loss": 0.0279, + "step": 8090 + }, + { + "epoch": 5.819816579751843, + "grad_norm": 3.316368879660235, + "learning_rate": 1.1243293426524387e-06, + "loss": 0.0636, + "step": 8091 + }, + { + "epoch": 5.820535874842654, + "grad_norm": 3.527080351119607, + "learning_rate": 1.1239999511295428e-06, + "loss": 0.0557, + "step": 8092 + }, + { + "epoch": 5.821255169933465, + "grad_norm": 2.3871702741146716, + "learning_rate": 1.1236705789525097e-06, + "loss": 0.0425, + "step": 8093 + }, + { + "epoch": 5.821974465024276, + "grad_norm": 1.2961833982385687, + "learning_rate": 1.1233412261382855e-06, + "loss": 0.0165, + "step": 8094 + }, + { + "epoch": 5.822693760115087, + "grad_norm": 1.9805760073677197, + "learning_rate": 1.1230118927038158e-06, + "loss": 0.0325, + "step": 8095 + }, + { + "epoch": 5.823413055205898, + "grad_norm": 4.3544827283303515, + "learning_rate": 1.1226825786660468e-06, + "loss": 0.0302, + "step": 8096 + }, + { + "epoch": 5.8241323502967095, + "grad_norm": 0.01641297287257619, + "learning_rate": 1.1223532840419213e-06, + "loss": 0.0001, + "step": 8097 + }, + { + "epoch": 5.82485164538752, + "grad_norm": 0.47088456581955324, + "learning_rate": 1.1220240088483826e-06, + "loss": 0.0009, + "step": 8098 + }, + { + "epoch": 5.825570940478332, + "grad_norm": 1.3970161570766604, + "learning_rate": 1.121694753102372e-06, + "loss": 0.0187, + "step": 8099 + }, + { + "epoch": 5.826290235569142, + "grad_norm": 4.643780619044687, + "learning_rate": 1.12136551682083e-06, + "loss": 0.088, + "step": 8100 + }, + { + "epoch": 5.827009530659954, + "grad_norm": 0.5270930640713518, + "learning_rate": 1.1210363000206974e-06, + "loss": 0.0035, + "step": 8101 + }, + { + "epoch": 5.827728825750764, + "grad_norm": 4.721462935227409, + "learning_rate": 1.1207071027189123e-06, + "loss": 0.0665, + "step": 8102 + }, + { + "epoch": 5.828448120841575, + "grad_norm": 8.334949917953903, + "learning_rate": 1.1203779249324127e-06, + "loss": 0.1848, + "step": 8103 + }, + { + "epoch": 5.829167415932386, + "grad_norm": 2.4055446304519643, + "learning_rate": 1.120048766678135e-06, + "loss": 0.0327, + "step": 8104 + }, + { + "epoch": 5.829886711023197, + "grad_norm": 0.03952224795029684, + "learning_rate": 1.1197196279730151e-06, + "loss": 0.0002, + "step": 8105 + }, + { + "epoch": 5.830606006114008, + "grad_norm": 1.4167755311346548, + "learning_rate": 1.1193905088339881e-06, + "loss": 0.0107, + "step": 8106 + }, + { + "epoch": 5.831325301204819, + "grad_norm": 2.942057720519932, + "learning_rate": 1.1190614092779876e-06, + "loss": 0.0389, + "step": 8107 + }, + { + "epoch": 5.83204459629563, + "grad_norm": 1.7322007095006828, + "learning_rate": 1.1187323293219465e-06, + "loss": 0.017, + "step": 8108 + }, + { + "epoch": 5.832763891386441, + "grad_norm": 0.07148823930476832, + "learning_rate": 1.1184032689827962e-06, + "loss": 0.0003, + "step": 8109 + }, + { + "epoch": 5.8334831864772525, + "grad_norm": 7.932332088436659, + "learning_rate": 1.1180742282774666e-06, + "loss": 0.2111, + "step": 8110 + }, + { + "epoch": 5.834202481568063, + "grad_norm": 1.266429682228784, + "learning_rate": 1.1177452072228896e-06, + "loss": 0.0196, + "step": 8111 + }, + { + "epoch": 5.834921776658875, + "grad_norm": 2.453501072124315, + "learning_rate": 1.1174162058359926e-06, + "loss": 0.0141, + "step": 8112 + }, + { + "epoch": 5.835641071749685, + "grad_norm": 2.1115950529925973, + "learning_rate": 1.1170872241337034e-06, + "loss": 0.0233, + "step": 8113 + }, + { + "epoch": 5.836360366840497, + "grad_norm": 4.183404806246734, + "learning_rate": 1.1167582621329484e-06, + "loss": 0.0807, + "step": 8114 + }, + { + "epoch": 5.837079661931307, + "grad_norm": 0.5734457320850825, + "learning_rate": 1.1164293198506542e-06, + "loss": 0.003, + "step": 8115 + }, + { + "epoch": 5.837798957022119, + "grad_norm": 1.045905674417317, + "learning_rate": 1.1161003973037437e-06, + "loss": 0.0043, + "step": 8116 + }, + { + "epoch": 5.838518252112929, + "grad_norm": 0.2765736231413716, + "learning_rate": 1.1157714945091425e-06, + "loss": 0.0016, + "step": 8117 + }, + { + "epoch": 5.839237547203741, + "grad_norm": 2.2068410491251806, + "learning_rate": 1.1154426114837723e-06, + "loss": 0.0288, + "step": 8118 + }, + { + "epoch": 5.839956842294551, + "grad_norm": 2.332234106772409, + "learning_rate": 1.1151137482445547e-06, + "loss": 0.0251, + "step": 8119 + }, + { + "epoch": 5.840676137385362, + "grad_norm": 1.4771683402666593, + "learning_rate": 1.1147849048084103e-06, + "loss": 0.0156, + "step": 8120 + }, + { + "epoch": 5.841395432476173, + "grad_norm": 2.0399942680855427, + "learning_rate": 1.114456081192259e-06, + "loss": 0.0225, + "step": 8121 + }, + { + "epoch": 5.842114727566984, + "grad_norm": 0.36421914004573214, + "learning_rate": 1.114127277413019e-06, + "loss": 0.001, + "step": 8122 + }, + { + "epoch": 5.8428340226577955, + "grad_norm": 3.9945466822051063, + "learning_rate": 1.113798493487608e-06, + "loss": 0.0759, + "step": 8123 + }, + { + "epoch": 5.843553317748606, + "grad_norm": 6.185490132591463, + "learning_rate": 1.1134697294329425e-06, + "loss": 0.0559, + "step": 8124 + }, + { + "epoch": 5.844272612839418, + "grad_norm": 2.7372379624697643, + "learning_rate": 1.1131409852659382e-06, + "loss": 0.0348, + "step": 8125 + }, + { + "epoch": 5.844991907930228, + "grad_norm": 6.558424017030756, + "learning_rate": 1.1128122610035083e-06, + "loss": 0.1626, + "step": 8126 + }, + { + "epoch": 5.84571120302104, + "grad_norm": 1.9029697496249245, + "learning_rate": 1.112483556662568e-06, + "loss": 0.0096, + "step": 8127 + }, + { + "epoch": 5.84643049811185, + "grad_norm": 3.017079876756938, + "learning_rate": 1.1121548722600293e-06, + "loss": 0.0687, + "step": 8128 + }, + { + "epoch": 5.847149793202662, + "grad_norm": 1.599689600678877, + "learning_rate": 1.111826207812803e-06, + "loss": 0.0161, + "step": 8129 + }, + { + "epoch": 5.847869088293472, + "grad_norm": 2.439068607051563, + "learning_rate": 1.1114975633378e-06, + "loss": 0.0231, + "step": 8130 + }, + { + "epoch": 5.848588383384284, + "grad_norm": 4.253821725565907, + "learning_rate": 1.1111689388519282e-06, + "loss": 0.0831, + "step": 8131 + }, + { + "epoch": 5.849307678475094, + "grad_norm": 2.475228107924166, + "learning_rate": 1.1108403343720978e-06, + "loss": 0.0617, + "step": 8132 + }, + { + "epoch": 5.850026973565905, + "grad_norm": 2.01151685742089, + "learning_rate": 1.1105117499152157e-06, + "loss": 0.0441, + "step": 8133 + }, + { + "epoch": 5.850746268656716, + "grad_norm": 0.20320278688170243, + "learning_rate": 1.1101831854981874e-06, + "loss": 0.0004, + "step": 8134 + }, + { + "epoch": 5.851465563747528, + "grad_norm": 5.954862596606296, + "learning_rate": 1.1098546411379186e-06, + "loss": 0.0675, + "step": 8135 + }, + { + "epoch": 5.8521848588383385, + "grad_norm": 5.369379530214561, + "learning_rate": 1.1095261168513136e-06, + "loss": 0.0655, + "step": 8136 + }, + { + "epoch": 5.852904153929149, + "grad_norm": 4.23120167672999, + "learning_rate": 1.1091976126552748e-06, + "loss": 0.0796, + "step": 8137 + }, + { + "epoch": 5.853623449019961, + "grad_norm": 2.2580017863492174, + "learning_rate": 1.1088691285667054e-06, + "loss": 0.0184, + "step": 8138 + }, + { + "epoch": 5.854342744110771, + "grad_norm": 4.441533407889899, + "learning_rate": 1.1085406646025057e-06, + "loss": 0.0493, + "step": 8139 + }, + { + "epoch": 5.855062039201583, + "grad_norm": 1.4231613025685148, + "learning_rate": 1.1082122207795762e-06, + "loss": 0.0162, + "step": 8140 + }, + { + "epoch": 5.855781334292393, + "grad_norm": 5.461699348894913, + "learning_rate": 1.1078837971148157e-06, + "loss": 0.0668, + "step": 8141 + }, + { + "epoch": 5.856500629383205, + "grad_norm": 1.0767402008573685, + "learning_rate": 1.1075553936251216e-06, + "loss": 0.0054, + "step": 8142 + }, + { + "epoch": 5.857219924474015, + "grad_norm": 1.5502466566417419, + "learning_rate": 1.1072270103273921e-06, + "loss": 0.0143, + "step": 8143 + }, + { + "epoch": 5.857939219564827, + "grad_norm": 2.548058516669674, + "learning_rate": 1.1068986472385222e-06, + "loss": 0.0306, + "step": 8144 + }, + { + "epoch": 5.858658514655637, + "grad_norm": 1.8126332999480148, + "learning_rate": 1.1065703043754069e-06, + "loss": 0.018, + "step": 8145 + }, + { + "epoch": 5.859377809746449, + "grad_norm": 2.9547722503013274, + "learning_rate": 1.1062419817549402e-06, + "loss": 0.0174, + "step": 8146 + }, + { + "epoch": 5.860097104837259, + "grad_norm": 0.4477894754574051, + "learning_rate": 1.105913679394014e-06, + "loss": 0.0031, + "step": 8147 + }, + { + "epoch": 5.860816399928071, + "grad_norm": 4.143271055524187, + "learning_rate": 1.1055853973095214e-06, + "loss": 0.0477, + "step": 8148 + }, + { + "epoch": 5.8615356950188815, + "grad_norm": 1.3825487024372694, + "learning_rate": 1.1052571355183527e-06, + "loss": 0.0184, + "step": 8149 + }, + { + "epoch": 5.862254990109692, + "grad_norm": 1.8808104184803627, + "learning_rate": 1.1049288940373973e-06, + "loss": 0.019, + "step": 8150 + }, + { + "epoch": 5.862974285200504, + "grad_norm": 2.4495294179313674, + "learning_rate": 1.104600672883544e-06, + "loss": 0.0412, + "step": 8151 + }, + { + "epoch": 5.863693580291314, + "grad_norm": 4.526217401326612, + "learning_rate": 1.1042724720736792e-06, + "loss": 0.0888, + "step": 8152 + }, + { + "epoch": 5.864412875382126, + "grad_norm": 3.8546676751738955, + "learning_rate": 1.1039442916246912e-06, + "loss": 0.0934, + "step": 8153 + }, + { + "epoch": 5.865132170472936, + "grad_norm": 3.015077513092094, + "learning_rate": 1.1036161315534645e-06, + "loss": 0.0397, + "step": 8154 + }, + { + "epoch": 5.865851465563748, + "grad_norm": 2.1065340224124665, + "learning_rate": 1.103287991876884e-06, + "loss": 0.0344, + "step": 8155 + }, + { + "epoch": 5.866570760654558, + "grad_norm": 2.509628782149461, + "learning_rate": 1.1029598726118322e-06, + "loss": 0.0438, + "step": 8156 + }, + { + "epoch": 5.86729005574537, + "grad_norm": 2.2476158460271938, + "learning_rate": 1.1026317737751923e-06, + "loss": 0.0328, + "step": 8157 + }, + { + "epoch": 5.86800935083618, + "grad_norm": 2.065767154968053, + "learning_rate": 1.1023036953838448e-06, + "loss": 0.0294, + "step": 8158 + }, + { + "epoch": 5.868728645926992, + "grad_norm": 0.006327415915982132, + "learning_rate": 1.1019756374546704e-06, + "loss": 0.0, + "step": 8159 + }, + { + "epoch": 5.869447941017802, + "grad_norm": 1.9114421014629235, + "learning_rate": 1.1016476000045486e-06, + "loss": 0.0071, + "step": 8160 + }, + { + "epoch": 5.870167236108614, + "grad_norm": 3.3152951339412393, + "learning_rate": 1.1013195830503567e-06, + "loss": 0.0444, + "step": 8161 + }, + { + "epoch": 5.8708865311994245, + "grad_norm": 5.199392381105336, + "learning_rate": 1.1009915866089723e-06, + "loss": 0.097, + "step": 8162 + }, + { + "epoch": 5.871605826290235, + "grad_norm": 2.5866881916469806, + "learning_rate": 1.1006636106972704e-06, + "loss": 0.0217, + "step": 8163 + }, + { + "epoch": 5.872325121381047, + "grad_norm": 2.044122296114998, + "learning_rate": 1.100335655332128e-06, + "loss": 0.0346, + "step": 8164 + }, + { + "epoch": 5.873044416471858, + "grad_norm": 3.1138510878209256, + "learning_rate": 1.100007720530417e-06, + "loss": 0.0459, + "step": 8165 + }, + { + "epoch": 5.873763711562669, + "grad_norm": 0.8155596450545165, + "learning_rate": 1.0996798063090112e-06, + "loss": 0.0054, + "step": 8166 + }, + { + "epoch": 5.874483006653479, + "grad_norm": 2.393414355129619, + "learning_rate": 1.0993519126847822e-06, + "loss": 0.0362, + "step": 8167 + }, + { + "epoch": 5.875202301744291, + "grad_norm": 3.256445344471677, + "learning_rate": 1.0990240396745997e-06, + "loss": 0.0451, + "step": 8168 + }, + { + "epoch": 5.875921596835101, + "grad_norm": 0.744588446545627, + "learning_rate": 1.098696187295335e-06, + "loss": 0.0042, + "step": 8169 + }, + { + "epoch": 5.876640891925913, + "grad_norm": 3.5766191307547417, + "learning_rate": 1.098368355563856e-06, + "loss": 0.0401, + "step": 8170 + }, + { + "epoch": 5.877360187016723, + "grad_norm": 0.561270262120307, + "learning_rate": 1.09804054449703e-06, + "loss": 0.0042, + "step": 8171 + }, + { + "epoch": 5.878079482107535, + "grad_norm": 2.132637553990446, + "learning_rate": 1.0977127541117238e-06, + "loss": 0.0151, + "step": 8172 + }, + { + "epoch": 5.878798777198345, + "grad_norm": 3.3560936135088566, + "learning_rate": 1.097384984424802e-06, + "loss": 0.0059, + "step": 8173 + }, + { + "epoch": 5.879518072289157, + "grad_norm": 3.392690626263601, + "learning_rate": 1.09705723545313e-06, + "loss": 0.0336, + "step": 8174 + }, + { + "epoch": 5.8802373673799675, + "grad_norm": 0.05026288545250037, + "learning_rate": 1.0967295072135707e-06, + "loss": 0.0001, + "step": 8175 + }, + { + "epoch": 5.880956662470779, + "grad_norm": 3.0729230621791994, + "learning_rate": 1.096401799722986e-06, + "loss": 0.0797, + "step": 8176 + }, + { + "epoch": 5.88167595756159, + "grad_norm": 0.20277946659673088, + "learning_rate": 1.0960741129982373e-06, + "loss": 0.0007, + "step": 8177 + }, + { + "epoch": 5.882395252652401, + "grad_norm": 5.410326046909365, + "learning_rate": 1.0957464470561845e-06, + "loss": 0.0641, + "step": 8178 + }, + { + "epoch": 5.883114547743212, + "grad_norm": 2.7386515804249796, + "learning_rate": 1.0954188019136864e-06, + "loss": 0.041, + "step": 8179 + }, + { + "epoch": 5.883833842834022, + "grad_norm": 3.8944205834371983, + "learning_rate": 1.0950911775876013e-06, + "loss": 0.0552, + "step": 8180 + }, + { + "epoch": 5.884553137924834, + "grad_norm": 0.5425452805108778, + "learning_rate": 1.0947635740947863e-06, + "loss": 0.0008, + "step": 8181 + }, + { + "epoch": 5.885272433015644, + "grad_norm": 0.0983631083428855, + "learning_rate": 1.0944359914520963e-06, + "loss": 0.0004, + "step": 8182 + }, + { + "epoch": 5.885991728106456, + "grad_norm": 2.9571098409117926, + "learning_rate": 1.0941084296763867e-06, + "loss": 0.0432, + "step": 8183 + }, + { + "epoch": 5.886711023197266, + "grad_norm": 0.047248729367984334, + "learning_rate": 1.0937808887845104e-06, + "loss": 0.0002, + "step": 8184 + }, + { + "epoch": 5.887430318288078, + "grad_norm": 0.31161670031238825, + "learning_rate": 1.093453368793321e-06, + "loss": 0.0017, + "step": 8185 + }, + { + "epoch": 5.888149613378888, + "grad_norm": 3.7878603865370413, + "learning_rate": 1.0931258697196694e-06, + "loss": 0.0869, + "step": 8186 + }, + { + "epoch": 5.8888689084697, + "grad_norm": 3.5894145944317613, + "learning_rate": 1.0927983915804064e-06, + "loss": 0.0308, + "step": 8187 + }, + { + "epoch": 5.8895882035605105, + "grad_norm": 0.15126422748401017, + "learning_rate": 1.0924709343923805e-06, + "loss": 0.0008, + "step": 8188 + }, + { + "epoch": 5.890307498651322, + "grad_norm": 0.04986182976784408, + "learning_rate": 1.09214349817244e-06, + "loss": 0.0002, + "step": 8189 + }, + { + "epoch": 5.891026793742133, + "grad_norm": 0.2762435435261827, + "learning_rate": 1.0918160829374331e-06, + "loss": 0.0011, + "step": 8190 + }, + { + "epoch": 5.891746088832944, + "grad_norm": 4.700056971833961, + "learning_rate": 1.0914886887042053e-06, + "loss": 0.0735, + "step": 8191 + }, + { + "epoch": 5.892465383923755, + "grad_norm": 2.149306052476679, + "learning_rate": 1.0911613154896013e-06, + "loss": 0.0052, + "step": 8192 + }, + { + "epoch": 5.893184679014566, + "grad_norm": 2.337270667855665, + "learning_rate": 1.0908339633104656e-06, + "loss": 0.0197, + "step": 8193 + }, + { + "epoch": 5.893903974105377, + "grad_norm": 6.946430787289715, + "learning_rate": 1.0905066321836404e-06, + "loss": 0.0992, + "step": 8194 + }, + { + "epoch": 5.894623269196188, + "grad_norm": 2.674130893616468, + "learning_rate": 1.0901793221259681e-06, + "loss": 0.0312, + "step": 8195 + }, + { + "epoch": 5.895342564286999, + "grad_norm": 0.04323555852630184, + "learning_rate": 1.089852033154289e-06, + "loss": 0.0001, + "step": 8196 + }, + { + "epoch": 5.896061859377809, + "grad_norm": 1.8703998485689581, + "learning_rate": 1.089524765285443e-06, + "loss": 0.0194, + "step": 8197 + }, + { + "epoch": 5.896781154468621, + "grad_norm": 2.0488169241583756, + "learning_rate": 1.0891975185362682e-06, + "loss": 0.0233, + "step": 8198 + }, + { + "epoch": 5.897500449559431, + "grad_norm": 3.964501949967531, + "learning_rate": 1.0888702929236023e-06, + "loss": 0.0325, + "step": 8199 + }, + { + "epoch": 5.898219744650243, + "grad_norm": 2.036017153371765, + "learning_rate": 1.0885430884642813e-06, + "loss": 0.0194, + "step": 8200 + }, + { + "epoch": 5.8989390397410535, + "grad_norm": 3.6352026416046708, + "learning_rate": 1.0882159051751408e-06, + "loss": 0.1093, + "step": 8201 + }, + { + "epoch": 5.899658334831865, + "grad_norm": 2.583622719758922, + "learning_rate": 1.087888743073015e-06, + "loss": 0.0126, + "step": 8202 + }, + { + "epoch": 5.900377629922676, + "grad_norm": 0.007485425779039772, + "learning_rate": 1.0875616021747368e-06, + "loss": 0.0, + "step": 8203 + }, + { + "epoch": 5.901096925013487, + "grad_norm": 2.5825991029659887, + "learning_rate": 1.0872344824971381e-06, + "loss": 0.0281, + "step": 8204 + }, + { + "epoch": 5.901816220104298, + "grad_norm": 2.239847975402559, + "learning_rate": 1.0869073840570494e-06, + "loss": 0.016, + "step": 8205 + }, + { + "epoch": 5.902535515195109, + "grad_norm": 2.5725433549291177, + "learning_rate": 1.0865803068713015e-06, + "loss": 0.0066, + "step": 8206 + }, + { + "epoch": 5.90325481028592, + "grad_norm": 1.224379226810905, + "learning_rate": 1.0862532509567224e-06, + "loss": 0.0117, + "step": 8207 + }, + { + "epoch": 5.903974105376731, + "grad_norm": 1.1894091165611778, + "learning_rate": 1.0859262163301398e-06, + "loss": 0.0034, + "step": 8208 + }, + { + "epoch": 5.904693400467542, + "grad_norm": 3.7342229741159585, + "learning_rate": 1.0855992030083808e-06, + "loss": 0.0601, + "step": 8209 + }, + { + "epoch": 5.905412695558352, + "grad_norm": 2.261762333610983, + "learning_rate": 1.0852722110082693e-06, + "loss": 0.0373, + "step": 8210 + }, + { + "epoch": 5.906131990649164, + "grad_norm": 3.578754280904679, + "learning_rate": 1.0849452403466313e-06, + "loss": 0.0714, + "step": 8211 + }, + { + "epoch": 5.906851285739975, + "grad_norm": 0.6042634957137347, + "learning_rate": 1.084618291040289e-06, + "loss": 0.0044, + "step": 8212 + }, + { + "epoch": 5.907570580830786, + "grad_norm": 8.31925655299347, + "learning_rate": 1.0842913631060649e-06, + "loss": 0.1942, + "step": 8213 + }, + { + "epoch": 5.9082898759215965, + "grad_norm": 3.3107277332761145, + "learning_rate": 1.08396445656078e-06, + "loss": 0.0586, + "step": 8214 + }, + { + "epoch": 5.909009171012408, + "grad_norm": 0.18316098168317677, + "learning_rate": 1.0836375714212537e-06, + "loss": 0.0003, + "step": 8215 + }, + { + "epoch": 5.909728466103219, + "grad_norm": 2.3120271323689, + "learning_rate": 1.0833107077043058e-06, + "loss": 0.0261, + "step": 8216 + }, + { + "epoch": 5.91044776119403, + "grad_norm": 2.729457640398173, + "learning_rate": 1.0829838654267533e-06, + "loss": 0.0286, + "step": 8217 + }, + { + "epoch": 5.911167056284841, + "grad_norm": 3.2578739277123225, + "learning_rate": 1.082657044605413e-06, + "loss": 0.1557, + "step": 8218 + }, + { + "epoch": 5.911886351375652, + "grad_norm": 4.979746607191992, + "learning_rate": 1.0823302452571002e-06, + "loss": 0.0697, + "step": 8219 + }, + { + "epoch": 5.912605646466463, + "grad_norm": 0.7140124013427698, + "learning_rate": 1.0820034673986297e-06, + "loss": 0.001, + "step": 8220 + }, + { + "epoch": 5.913324941557274, + "grad_norm": 1.204992372750038, + "learning_rate": 1.0816767110468139e-06, + "loss": 0.0095, + "step": 8221 + }, + { + "epoch": 5.914044236648085, + "grad_norm": 0.073346266263922, + "learning_rate": 1.0813499762184665e-06, + "loss": 0.0002, + "step": 8222 + }, + { + "epoch": 5.914763531738896, + "grad_norm": 3.996719903493425, + "learning_rate": 1.0810232629303977e-06, + "loss": 0.0682, + "step": 8223 + }, + { + "epoch": 5.915482826829707, + "grad_norm": 1.7560649657304546, + "learning_rate": 1.0806965711994172e-06, + "loss": 0.0305, + "step": 8224 + }, + { + "epoch": 5.916202121920518, + "grad_norm": 0.5120222396398785, + "learning_rate": 1.0803699010423343e-06, + "loss": 0.0011, + "step": 8225 + }, + { + "epoch": 5.916921417011329, + "grad_norm": 2.6017630559611695, + "learning_rate": 1.080043252475956e-06, + "loss": 0.0248, + "step": 8226 + }, + { + "epoch": 5.9176407121021395, + "grad_norm": 0.7356094945332432, + "learning_rate": 1.0797166255170904e-06, + "loss": 0.0082, + "step": 8227 + }, + { + "epoch": 5.918360007192951, + "grad_norm": 2.988356992425312, + "learning_rate": 1.079390020182542e-06, + "loss": 0.059, + "step": 8228 + }, + { + "epoch": 5.919079302283762, + "grad_norm": 1.4380732855798963, + "learning_rate": 1.0790634364891155e-06, + "loss": 0.0177, + "step": 8229 + }, + { + "epoch": 5.919798597374573, + "grad_norm": 1.7211276918195026, + "learning_rate": 1.078736874453614e-06, + "loss": 0.0174, + "step": 8230 + }, + { + "epoch": 5.920517892465384, + "grad_norm": 5.966106561870387, + "learning_rate": 1.0784103340928394e-06, + "loss": 0.0878, + "step": 8231 + }, + { + "epoch": 5.921237187556195, + "grad_norm": 4.200827555158548, + "learning_rate": 1.0780838154235937e-06, + "loss": 0.0483, + "step": 8232 + }, + { + "epoch": 5.921956482647006, + "grad_norm": 3.1687289379269354, + "learning_rate": 1.0777573184626765e-06, + "loss": 0.0376, + "step": 8233 + }, + { + "epoch": 5.922675777737817, + "grad_norm": 3.124027211092012, + "learning_rate": 1.077430843226886e-06, + "loss": 0.0474, + "step": 8234 + }, + { + "epoch": 5.923395072828628, + "grad_norm": 1.9867740350257643, + "learning_rate": 1.0771043897330208e-06, + "loss": 0.0087, + "step": 8235 + }, + { + "epoch": 5.924114367919439, + "grad_norm": 0.24911252408814874, + "learning_rate": 1.0767779579978768e-06, + "loss": 0.0005, + "step": 8236 + }, + { + "epoch": 5.92483366301025, + "grad_norm": 2.1481599286040556, + "learning_rate": 1.0764515480382499e-06, + "loss": 0.018, + "step": 8237 + }, + { + "epoch": 5.925552958101061, + "grad_norm": 2.2435869814002727, + "learning_rate": 1.0761251598709344e-06, + "loss": 0.0428, + "step": 8238 + }, + { + "epoch": 5.926272253191872, + "grad_norm": 4.500991537008146, + "learning_rate": 1.0757987935127238e-06, + "loss": 0.0803, + "step": 8239 + }, + { + "epoch": 5.9269915482826825, + "grad_norm": 7.028911091120986, + "learning_rate": 1.0754724489804098e-06, + "loss": 0.1243, + "step": 8240 + }, + { + "epoch": 5.927710843373494, + "grad_norm": 2.8736907329611894, + "learning_rate": 1.0751461262907834e-06, + "loss": 0.0703, + "step": 8241 + }, + { + "epoch": 5.928430138464305, + "grad_norm": 0.645748038606214, + "learning_rate": 1.074819825460634e-06, + "loss": 0.0015, + "step": 8242 + }, + { + "epoch": 5.929149433555116, + "grad_norm": 1.9842096494568051, + "learning_rate": 1.0744935465067516e-06, + "loss": 0.0201, + "step": 8243 + }, + { + "epoch": 5.929868728645927, + "grad_norm": 3.6685635765757785, + "learning_rate": 1.0741672894459233e-06, + "loss": 0.0571, + "step": 8244 + }, + { + "epoch": 5.930588023736738, + "grad_norm": 3.2203953155561265, + "learning_rate": 1.0738410542949352e-06, + "loss": 0.0449, + "step": 8245 + }, + { + "epoch": 5.931307318827549, + "grad_norm": 1.578072237009142, + "learning_rate": 1.0735148410705736e-06, + "loss": 0.0129, + "step": 8246 + }, + { + "epoch": 5.93202661391836, + "grad_norm": 1.768940034726935, + "learning_rate": 1.0731886497896212e-06, + "loss": 0.0145, + "step": 8247 + }, + { + "epoch": 5.932745909009171, + "grad_norm": 0.1375979214346885, + "learning_rate": 1.0728624804688626e-06, + "loss": 0.0005, + "step": 8248 + }, + { + "epoch": 5.933465204099982, + "grad_norm": 1.8716530127886566, + "learning_rate": 1.072536333125079e-06, + "loss": 0.0164, + "step": 8249 + }, + { + "epoch": 5.934184499190793, + "grad_norm": 2.8738779614647356, + "learning_rate": 1.0722102077750514e-06, + "loss": 0.0225, + "step": 8250 + }, + { + "epoch": 5.934903794281604, + "grad_norm": 4.075770125644606, + "learning_rate": 1.0718841044355596e-06, + "loss": 0.0489, + "step": 8251 + }, + { + "epoch": 5.935623089372415, + "grad_norm": 1.721372427219796, + "learning_rate": 1.071558023123382e-06, + "loss": 0.0307, + "step": 8252 + }, + { + "epoch": 5.936342384463226, + "grad_norm": 1.4220872707096548, + "learning_rate": 1.0712319638552966e-06, + "loss": 0.0162, + "step": 8253 + }, + { + "epoch": 5.937061679554037, + "grad_norm": 3.936931319815165, + "learning_rate": 1.0709059266480792e-06, + "loss": 0.1252, + "step": 8254 + }, + { + "epoch": 5.937780974644848, + "grad_norm": 1.9805382194250618, + "learning_rate": 1.0705799115185051e-06, + "loss": 0.0252, + "step": 8255 + }, + { + "epoch": 5.938500269735659, + "grad_norm": 0.2783049904778019, + "learning_rate": 1.0702539184833484e-06, + "loss": 0.0008, + "step": 8256 + }, + { + "epoch": 5.93921956482647, + "grad_norm": 0.03339517973851255, + "learning_rate": 1.0699279475593814e-06, + "loss": 0.0001, + "step": 8257 + }, + { + "epoch": 5.939938859917281, + "grad_norm": 1.9590568735702638, + "learning_rate": 1.069601998763377e-06, + "loss": 0.0225, + "step": 8258 + }, + { + "epoch": 5.940658155008092, + "grad_norm": 1.3310892510833916, + "learning_rate": 1.0692760721121058e-06, + "loss": 0.0168, + "step": 8259 + }, + { + "epoch": 5.941377450098903, + "grad_norm": 6.1026200271982916, + "learning_rate": 1.0689501676223363e-06, + "loss": 0.1223, + "step": 8260 + }, + { + "epoch": 5.942096745189714, + "grad_norm": 2.550927251502519, + "learning_rate": 1.0686242853108373e-06, + "loss": 0.0246, + "step": 8261 + }, + { + "epoch": 5.942816040280525, + "grad_norm": 2.120004677363871, + "learning_rate": 1.0682984251943758e-06, + "loss": 0.0202, + "step": 8262 + }, + { + "epoch": 5.943535335371336, + "grad_norm": 3.7198928707684327, + "learning_rate": 1.0679725872897185e-06, + "loss": 0.045, + "step": 8263 + }, + { + "epoch": 5.944254630462147, + "grad_norm": 3.5957023034824505, + "learning_rate": 1.06764677161363e-06, + "loss": 0.041, + "step": 8264 + }, + { + "epoch": 5.944973925552958, + "grad_norm": 5.3477814333983105, + "learning_rate": 1.067320978182874e-06, + "loss": 0.0693, + "step": 8265 + }, + { + "epoch": 5.945693220643769, + "grad_norm": 5.806415363124138, + "learning_rate": 1.0669952070142134e-06, + "loss": 0.1061, + "step": 8266 + }, + { + "epoch": 5.94641251573458, + "grad_norm": 1.60008120676573, + "learning_rate": 1.0666694581244095e-06, + "loss": 0.0362, + "step": 8267 + }, + { + "epoch": 5.947131810825391, + "grad_norm": 4.799260591769724, + "learning_rate": 1.0663437315302223e-06, + "loss": 0.1067, + "step": 8268 + }, + { + "epoch": 5.947851105916202, + "grad_norm": 2.6087512590898556, + "learning_rate": 1.0660180272484116e-06, + "loss": 0.0335, + "step": 8269 + }, + { + "epoch": 5.9485704010070135, + "grad_norm": 3.9165202790200375, + "learning_rate": 1.0656923452957355e-06, + "loss": 0.0231, + "step": 8270 + }, + { + "epoch": 5.949289696097824, + "grad_norm": 2.129247272067369, + "learning_rate": 1.0653666856889505e-06, + "loss": 0.0405, + "step": 8271 + }, + { + "epoch": 5.950008991188636, + "grad_norm": 0.8799597412569004, + "learning_rate": 1.0650410484448122e-06, + "loss": 0.0092, + "step": 8272 + }, + { + "epoch": 5.950728286279446, + "grad_norm": 6.589599417940666, + "learning_rate": 1.0647154335800756e-06, + "loss": 0.1146, + "step": 8273 + }, + { + "epoch": 5.951447581370257, + "grad_norm": 0.3287111630734672, + "learning_rate": 1.0643898411114941e-06, + "loss": 0.0006, + "step": 8274 + }, + { + "epoch": 5.952166876461068, + "grad_norm": 0.49812483022088755, + "learning_rate": 1.06406427105582e-06, + "loss": 0.0032, + "step": 8275 + }, + { + "epoch": 5.952886171551879, + "grad_norm": 1.0069107048993367, + "learning_rate": 1.0637387234298047e-06, + "loss": 0.0138, + "step": 8276 + }, + { + "epoch": 5.95360546664269, + "grad_norm": 3.4268261094410906, + "learning_rate": 1.0634131982501975e-06, + "loss": 0.0296, + "step": 8277 + }, + { + "epoch": 5.954324761733501, + "grad_norm": 2.3542275810183524, + "learning_rate": 1.0630876955337472e-06, + "loss": 0.028, + "step": 8278 + }, + { + "epoch": 5.955044056824312, + "grad_norm": 2.9212372890650142, + "learning_rate": 1.0627622152972027e-06, + "loss": 0.0306, + "step": 8279 + }, + { + "epoch": 5.955763351915123, + "grad_norm": 1.7003273717484655, + "learning_rate": 1.0624367575573098e-06, + "loss": 0.0085, + "step": 8280 + }, + { + "epoch": 5.956482647005934, + "grad_norm": 3.5269994611904703, + "learning_rate": 1.0621113223308135e-06, + "loss": 0.0725, + "step": 8281 + }, + { + "epoch": 5.957201942096745, + "grad_norm": 2.9246683890806495, + "learning_rate": 1.061785909634459e-06, + "loss": 0.0412, + "step": 8282 + }, + { + "epoch": 5.9579212371875565, + "grad_norm": 5.585421449917058, + "learning_rate": 1.0614605194849877e-06, + "loss": 0.0564, + "step": 8283 + }, + { + "epoch": 5.958640532278367, + "grad_norm": 1.2927275566127703, + "learning_rate": 1.061135151899143e-06, + "loss": 0.0114, + "step": 8284 + }, + { + "epoch": 5.959359827369179, + "grad_norm": 0.39354599628058234, + "learning_rate": 1.0608098068936651e-06, + "loss": 0.0017, + "step": 8285 + }, + { + "epoch": 5.960079122459989, + "grad_norm": 2.1178618206238924, + "learning_rate": 1.0604844844852939e-06, + "loss": 0.0277, + "step": 8286 + }, + { + "epoch": 5.9607984175508, + "grad_norm": 3.3650012837518557, + "learning_rate": 1.0601591846907674e-06, + "loss": 0.054, + "step": 8287 + }, + { + "epoch": 5.961517712641611, + "grad_norm": 1.9413095553412019, + "learning_rate": 1.0598339075268228e-06, + "loss": 0.0193, + "step": 8288 + }, + { + "epoch": 5.962237007732423, + "grad_norm": 5.718405785145276, + "learning_rate": 1.0595086530101963e-06, + "loss": 0.1515, + "step": 8289 + }, + { + "epoch": 5.962956302823233, + "grad_norm": 4.744081471594903, + "learning_rate": 1.059183421157623e-06, + "loss": 0.0598, + "step": 8290 + }, + { + "epoch": 5.963675597914044, + "grad_norm": 5.987060825985791, + "learning_rate": 1.0588582119858365e-06, + "loss": 0.0332, + "step": 8291 + }, + { + "epoch": 5.964394893004855, + "grad_norm": 1.5108363087865933, + "learning_rate": 1.0585330255115695e-06, + "loss": 0.0299, + "step": 8292 + }, + { + "epoch": 5.965114188095666, + "grad_norm": 0.8007609571666593, + "learning_rate": 1.0582078617515532e-06, + "loss": 0.0116, + "step": 8293 + }, + { + "epoch": 5.965833483186477, + "grad_norm": 4.4065173169999365, + "learning_rate": 1.0578827207225173e-06, + "loss": 0.0844, + "step": 8294 + }, + { + "epoch": 5.966552778277288, + "grad_norm": 1.7800247055075173, + "learning_rate": 1.0575576024411923e-06, + "loss": 0.0197, + "step": 8295 + }, + { + "epoch": 5.9672720733680995, + "grad_norm": 1.1251230233859937, + "learning_rate": 1.0572325069243054e-06, + "loss": 0.0133, + "step": 8296 + }, + { + "epoch": 5.96799136845891, + "grad_norm": 2.3150667905076174, + "learning_rate": 1.0569074341885833e-06, + "loss": 0.023, + "step": 8297 + }, + { + "epoch": 5.968710663549722, + "grad_norm": 0.43270059493358953, + "learning_rate": 1.0565823842507512e-06, + "loss": 0.0016, + "step": 8298 + }, + { + "epoch": 5.969429958640532, + "grad_norm": 2.9037191899378696, + "learning_rate": 1.0562573571275335e-06, + "loss": 0.0319, + "step": 8299 + }, + { + "epoch": 5.970149253731344, + "grad_norm": 1.2459027369516829, + "learning_rate": 1.055932352835654e-06, + "loss": 0.0091, + "step": 8300 + }, + { + "epoch": 5.970868548822154, + "grad_norm": 2.752336466381236, + "learning_rate": 1.055607371391835e-06, + "loss": 0.0476, + "step": 8301 + }, + { + "epoch": 5.971587843912966, + "grad_norm": 4.023795470685008, + "learning_rate": 1.0552824128127965e-06, + "loss": 0.0961, + "step": 8302 + }, + { + "epoch": 5.972307139003776, + "grad_norm": 2.5494180031967035, + "learning_rate": 1.0549574771152584e-06, + "loss": 0.0371, + "step": 8303 + }, + { + "epoch": 5.973026434094587, + "grad_norm": 1.2430320660965122, + "learning_rate": 1.0546325643159393e-06, + "loss": 0.0129, + "step": 8304 + }, + { + "epoch": 5.973745729185398, + "grad_norm": 0.06729775455353829, + "learning_rate": 1.0543076744315568e-06, + "loss": 0.0001, + "step": 8305 + }, + { + "epoch": 5.974465024276209, + "grad_norm": 1.3629243771362083, + "learning_rate": 1.0539828074788269e-06, + "loss": 0.0082, + "step": 8306 + }, + { + "epoch": 5.97518431936702, + "grad_norm": 5.963380624323929, + "learning_rate": 1.0536579634744645e-06, + "loss": 0.0866, + "step": 8307 + }, + { + "epoch": 5.975903614457831, + "grad_norm": 3.1590691817960295, + "learning_rate": 1.0533331424351834e-06, + "loss": 0.0636, + "step": 8308 + }, + { + "epoch": 5.9766229095486425, + "grad_norm": 0.6062446104456216, + "learning_rate": 1.0530083443776961e-06, + "loss": 0.0006, + "step": 8309 + }, + { + "epoch": 5.977342204639453, + "grad_norm": 4.456582055881374, + "learning_rate": 1.052683569318714e-06, + "loss": 0.0569, + "step": 8310 + }, + { + "epoch": 5.978061499730265, + "grad_norm": 2.5865198490712658, + "learning_rate": 1.0523588172749478e-06, + "loss": 0.0417, + "step": 8311 + }, + { + "epoch": 5.978780794821075, + "grad_norm": 0.15620827594611705, + "learning_rate": 1.0520340882631063e-06, + "loss": 0.0008, + "step": 8312 + }, + { + "epoch": 5.979500089911887, + "grad_norm": 1.5451954597831263, + "learning_rate": 1.051709382299897e-06, + "loss": 0.0273, + "step": 8313 + }, + { + "epoch": 5.980219385002697, + "grad_norm": 2.793685907775458, + "learning_rate": 1.0513846994020273e-06, + "loss": 0.0314, + "step": 8314 + }, + { + "epoch": 5.980938680093509, + "grad_norm": 5.010576402758322, + "learning_rate": 1.0510600395862016e-06, + "loss": 0.0522, + "step": 8315 + }, + { + "epoch": 5.981657975184319, + "grad_norm": 6.5645554896346745, + "learning_rate": 1.0507354028691257e-06, + "loss": 0.1199, + "step": 8316 + }, + { + "epoch": 5.98237727027513, + "grad_norm": 0.9044122622712482, + "learning_rate": 1.0504107892675017e-06, + "loss": 0.0048, + "step": 8317 + }, + { + "epoch": 5.983096565365941, + "grad_norm": 5.215984059244124, + "learning_rate": 1.0500861987980323e-06, + "loss": 0.015, + "step": 8318 + }, + { + "epoch": 5.983815860456753, + "grad_norm": 2.452970450481428, + "learning_rate": 1.0497616314774176e-06, + "loss": 0.0353, + "step": 8319 + }, + { + "epoch": 5.984535155547563, + "grad_norm": 1.569567769466921, + "learning_rate": 1.0494370873223566e-06, + "loss": 0.0021, + "step": 8320 + }, + { + "epoch": 5.985254450638374, + "grad_norm": 3.8724481904560077, + "learning_rate": 1.0491125663495488e-06, + "loss": 0.0455, + "step": 8321 + }, + { + "epoch": 5.9859737457291855, + "grad_norm": 6.11296205823761, + "learning_rate": 1.0487880685756912e-06, + "loss": 0.0527, + "step": 8322 + }, + { + "epoch": 5.986693040819996, + "grad_norm": 4.775299794251412, + "learning_rate": 1.0484635940174794e-06, + "loss": 0.1212, + "step": 8323 + }, + { + "epoch": 5.987412335910808, + "grad_norm": 3.693916270897016, + "learning_rate": 1.0481391426916084e-06, + "loss": 0.0504, + "step": 8324 + }, + { + "epoch": 5.988131631001618, + "grad_norm": 1.950006282187902, + "learning_rate": 1.0478147146147716e-06, + "loss": 0.0176, + "step": 8325 + }, + { + "epoch": 5.98885092609243, + "grad_norm": 4.312331002064015, + "learning_rate": 1.0474903098036615e-06, + "loss": 0.0581, + "step": 8326 + }, + { + "epoch": 5.98957022118324, + "grad_norm": 2.7839837378878567, + "learning_rate": 1.0471659282749695e-06, + "loss": 0.0256, + "step": 8327 + }, + { + "epoch": 5.990289516274052, + "grad_norm": 0.5605883765240076, + "learning_rate": 1.0468415700453856e-06, + "loss": 0.0019, + "step": 8328 + }, + { + "epoch": 5.991008811364862, + "grad_norm": 3.6476014142667, + "learning_rate": 1.0465172351315982e-06, + "loss": 0.0635, + "step": 8329 + }, + { + "epoch": 5.991728106455674, + "grad_norm": 2.3291738895358094, + "learning_rate": 1.0461929235502952e-06, + "loss": 0.0359, + "step": 8330 + }, + { + "epoch": 5.992447401546484, + "grad_norm": 2.1952357949621693, + "learning_rate": 1.0458686353181624e-06, + "loss": 0.0317, + "step": 8331 + }, + { + "epoch": 5.993166696637296, + "grad_norm": 2.587139429066047, + "learning_rate": 1.045544370451886e-06, + "loss": 0.0201, + "step": 8332 + }, + { + "epoch": 5.993885991728106, + "grad_norm": 0.5240653679586268, + "learning_rate": 1.0452201289681495e-06, + "loss": 0.0014, + "step": 8333 + }, + { + "epoch": 5.994605286818917, + "grad_norm": 2.8698656339899773, + "learning_rate": 1.0448959108836357e-06, + "loss": 0.0495, + "step": 8334 + }, + { + "epoch": 5.9953245819097285, + "grad_norm": 12.61295203320308, + "learning_rate": 1.0445717162150261e-06, + "loss": 0.0712, + "step": 8335 + }, + { + "epoch": 5.996043877000539, + "grad_norm": 1.1988591105530846, + "learning_rate": 1.0442475449790007e-06, + "loss": 0.0028, + "step": 8336 + }, + { + "epoch": 5.996763172091351, + "grad_norm": 0.20103404112486362, + "learning_rate": 1.0439233971922396e-06, + "loss": 0.0004, + "step": 8337 + }, + { + "epoch": 5.997482467182161, + "grad_norm": 4.537653428117391, + "learning_rate": 1.0435992728714207e-06, + "loss": 0.0769, + "step": 8338 + }, + { + "epoch": 5.998201762272973, + "grad_norm": 3.525432328817224, + "learning_rate": 1.04327517203322e-06, + "loss": 0.0691, + "step": 8339 + }, + { + "epoch": 5.998921057363783, + "grad_norm": 3.336508881655443, + "learning_rate": 1.0429510946943136e-06, + "loss": 0.0284, + "step": 8340 + }, + { + "epoch": 5.999640352454595, + "grad_norm": 3.31141561470824, + "learning_rate": 1.0426270408713754e-06, + "loss": 0.0296, + "step": 8341 + }, + { + "epoch": 6.000359647545405, + "grad_norm": 2.1104003479731026, + "learning_rate": 1.042303010581079e-06, + "loss": 0.0359, + "step": 8342 + }, + { + "epoch": 6.001078942636217, + "grad_norm": 0.5408937720970376, + "learning_rate": 1.0419790038400965e-06, + "loss": 0.0023, + "step": 8343 + }, + { + "epoch": 6.001798237727027, + "grad_norm": 1.6735815008283859, + "learning_rate": 1.0416550206650981e-06, + "loss": 0.0211, + "step": 8344 + }, + { + "epoch": 6.002517532817839, + "grad_norm": 1.2986484366670608, + "learning_rate": 1.0413310610727534e-06, + "loss": 0.0127, + "step": 8345 + }, + { + "epoch": 6.003236827908649, + "grad_norm": 2.6122784454431036, + "learning_rate": 1.0410071250797306e-06, + "loss": 0.0167, + "step": 8346 + }, + { + "epoch": 6.003956122999461, + "grad_norm": 4.214241251691837, + "learning_rate": 1.040683212702697e-06, + "loss": 0.0607, + "step": 8347 + }, + { + "epoch": 6.0046754180902715, + "grad_norm": 2.1340170035139705, + "learning_rate": 1.0403593239583188e-06, + "loss": 0.0143, + "step": 8348 + }, + { + "epoch": 6.005394713181083, + "grad_norm": 0.006438428227090916, + "learning_rate": 1.04003545886326e-06, + "loss": 0.0, + "step": 8349 + }, + { + "epoch": 6.006114008271894, + "grad_norm": 2.9780415106633886, + "learning_rate": 1.0397116174341845e-06, + "loss": 0.0708, + "step": 8350 + }, + { + "epoch": 6.006833303362704, + "grad_norm": 1.91674935497692, + "learning_rate": 1.039387799687754e-06, + "loss": 0.0192, + "step": 8351 + }, + { + "epoch": 6.007552598453516, + "grad_norm": 0.08374556223387493, + "learning_rate": 1.0390640056406293e-06, + "loss": 0.0003, + "step": 8352 + }, + { + "epoch": 6.008271893544326, + "grad_norm": 1.330735356607771, + "learning_rate": 1.0387402353094714e-06, + "loss": 0.0188, + "step": 8353 + }, + { + "epoch": 6.008991188635138, + "grad_norm": 3.4145994120108836, + "learning_rate": 1.0384164887109382e-06, + "loss": 0.0296, + "step": 8354 + }, + { + "epoch": 6.009710483725948, + "grad_norm": 1.7579032981119727, + "learning_rate": 1.0380927658616868e-06, + "loss": 0.013, + "step": 8355 + }, + { + "epoch": 6.01042977881676, + "grad_norm": 0.12605551979185137, + "learning_rate": 1.0377690667783734e-06, + "loss": 0.0002, + "step": 8356 + }, + { + "epoch": 6.01114907390757, + "grad_norm": 1.9570010675189464, + "learning_rate": 1.0374453914776523e-06, + "loss": 0.0204, + "step": 8357 + }, + { + "epoch": 6.011868368998382, + "grad_norm": 0.588033071952894, + "learning_rate": 1.0371217399761783e-06, + "loss": 0.0022, + "step": 8358 + }, + { + "epoch": 6.012587664089192, + "grad_norm": 3.3193154405485332, + "learning_rate": 1.0367981122906036e-06, + "loss": 0.03, + "step": 8359 + }, + { + "epoch": 6.013306959180004, + "grad_norm": 0.7996377792563844, + "learning_rate": 1.036474508437579e-06, + "loss": 0.0062, + "step": 8360 + }, + { + "epoch": 6.0140262542708145, + "grad_norm": 0.06372124545238748, + "learning_rate": 1.0361509284337548e-06, + "loss": 0.0002, + "step": 8361 + }, + { + "epoch": 6.014745549361626, + "grad_norm": 0.03642519281008519, + "learning_rate": 1.035827372295779e-06, + "loss": 0.0002, + "step": 8362 + }, + { + "epoch": 6.015464844452437, + "grad_norm": 1.0115387847585329, + "learning_rate": 1.0355038400403002e-06, + "loss": 0.0095, + "step": 8363 + }, + { + "epoch": 6.016184139543248, + "grad_norm": 0.058367620806626906, + "learning_rate": 1.035180331683964e-06, + "loss": 0.0002, + "step": 8364 + }, + { + "epoch": 6.016903434634059, + "grad_norm": 3.0273259242926187, + "learning_rate": 1.034856847243416e-06, + "loss": 0.0261, + "step": 8365 + }, + { + "epoch": 6.017622729724869, + "grad_norm": 1.914756615960446, + "learning_rate": 1.0345333867352998e-06, + "loss": 0.019, + "step": 8366 + }, + { + "epoch": 6.018342024815681, + "grad_norm": 1.398245900323466, + "learning_rate": 1.0342099501762573e-06, + "loss": 0.0128, + "step": 8367 + }, + { + "epoch": 6.019061319906491, + "grad_norm": 1.7154025778714008, + "learning_rate": 1.033886537582931e-06, + "loss": 0.0205, + "step": 8368 + }, + { + "epoch": 6.019780614997303, + "grad_norm": 3.0154372301262433, + "learning_rate": 1.0335631489719607e-06, + "loss": 0.0076, + "step": 8369 + }, + { + "epoch": 6.020499910088113, + "grad_norm": 1.3973763916606563, + "learning_rate": 1.033239784359985e-06, + "loss": 0.0103, + "step": 8370 + }, + { + "epoch": 6.021219205178925, + "grad_norm": 1.7138863714347263, + "learning_rate": 1.0329164437636418e-06, + "loss": 0.0108, + "step": 8371 + }, + { + "epoch": 6.021938500269735, + "grad_norm": 5.14570664432999, + "learning_rate": 1.0325931271995676e-06, + "loss": 0.0786, + "step": 8372 + }, + { + "epoch": 6.022657795360547, + "grad_norm": 0.31201034150490903, + "learning_rate": 1.0322698346843968e-06, + "loss": 0.0023, + "step": 8373 + }, + { + "epoch": 6.0233770904513575, + "grad_norm": 1.3802159953025759, + "learning_rate": 1.0319465662347649e-06, + "loss": 0.0132, + "step": 8374 + }, + { + "epoch": 6.024096385542169, + "grad_norm": 0.7370485750001731, + "learning_rate": 1.0316233218673035e-06, + "loss": 0.0023, + "step": 8375 + }, + { + "epoch": 6.02481568063298, + "grad_norm": 3.208218151601074, + "learning_rate": 1.0313001015986446e-06, + "loss": 0.0565, + "step": 8376 + }, + { + "epoch": 6.025534975723791, + "grad_norm": 4.141763136555581, + "learning_rate": 1.0309769054454186e-06, + "loss": 0.0274, + "step": 8377 + }, + { + "epoch": 6.026254270814602, + "grad_norm": 0.6638678614054352, + "learning_rate": 1.030653733424253e-06, + "loss": 0.0043, + "step": 8378 + }, + { + "epoch": 6.026973565905413, + "grad_norm": 4.8766613743704195, + "learning_rate": 1.0303305855517776e-06, + "loss": 0.0864, + "step": 8379 + }, + { + "epoch": 6.027692860996224, + "grad_norm": 1.3062470399187955, + "learning_rate": 1.0300074618446178e-06, + "loss": 0.0078, + "step": 8380 + }, + { + "epoch": 6.028412156087034, + "grad_norm": 0.1601294708326566, + "learning_rate": 1.0296843623193995e-06, + "loss": 0.0008, + "step": 8381 + }, + { + "epoch": 6.029131451177846, + "grad_norm": 0.433678440159115, + "learning_rate": 1.029361286992746e-06, + "loss": 0.0012, + "step": 8382 + }, + { + "epoch": 6.029850746268656, + "grad_norm": 2.697498504950228, + "learning_rate": 1.0290382358812805e-06, + "loss": 0.0397, + "step": 8383 + }, + { + "epoch": 6.030570041359468, + "grad_norm": 0.22976415871716435, + "learning_rate": 1.0287152090016244e-06, + "loss": 0.0007, + "step": 8384 + }, + { + "epoch": 6.031289336450278, + "grad_norm": 0.7292103556828585, + "learning_rate": 1.0283922063703983e-06, + "loss": 0.0041, + "step": 8385 + }, + { + "epoch": 6.03200863154109, + "grad_norm": 0.1796511338510292, + "learning_rate": 1.0280692280042211e-06, + "loss": 0.0003, + "step": 8386 + }, + { + "epoch": 6.0327279266319005, + "grad_norm": 3.1686022821393083, + "learning_rate": 1.0277462739197109e-06, + "loss": 0.0259, + "step": 8387 + }, + { + "epoch": 6.033447221722712, + "grad_norm": 3.0108204421888733, + "learning_rate": 1.0274233441334827e-06, + "loss": 0.0323, + "step": 8388 + }, + { + "epoch": 6.034166516813523, + "grad_norm": 2.4609272389099828, + "learning_rate": 1.0271004386621542e-06, + "loss": 0.0356, + "step": 8389 + }, + { + "epoch": 6.034885811904334, + "grad_norm": 6.38929612435006, + "learning_rate": 1.0267775575223382e-06, + "loss": 0.0909, + "step": 8390 + }, + { + "epoch": 6.035605106995145, + "grad_norm": 2.6204822400383914, + "learning_rate": 1.0264547007306473e-06, + "loss": 0.0449, + "step": 8391 + }, + { + "epoch": 6.036324402085956, + "grad_norm": 2.2669169082428082, + "learning_rate": 1.0261318683036933e-06, + "loss": 0.0243, + "step": 8392 + }, + { + "epoch": 6.037043697176767, + "grad_norm": 2.979873309196892, + "learning_rate": 1.0258090602580867e-06, + "loss": 0.0562, + "step": 8393 + }, + { + "epoch": 6.037762992267578, + "grad_norm": 0.9907572092151078, + "learning_rate": 1.0254862766104356e-06, + "loss": 0.009, + "step": 8394 + }, + { + "epoch": 6.038482287358389, + "grad_norm": 2.62839967212115, + "learning_rate": 1.0251635173773492e-06, + "loss": 0.0396, + "step": 8395 + }, + { + "epoch": 6.039201582449199, + "grad_norm": 3.437610333994754, + "learning_rate": 1.0248407825754334e-06, + "loss": 0.058, + "step": 8396 + }, + { + "epoch": 6.039920877540011, + "grad_norm": 0.428600761305421, + "learning_rate": 1.0245180722212935e-06, + "loss": 0.0023, + "step": 8397 + }, + { + "epoch": 6.040640172630821, + "grad_norm": 1.451719993583136, + "learning_rate": 1.0241953863315336e-06, + "loss": 0.0277, + "step": 8398 + }, + { + "epoch": 6.041359467721633, + "grad_norm": 0.06930149930974282, + "learning_rate": 1.0238727249227558e-06, + "loss": 0.0001, + "step": 8399 + }, + { + "epoch": 6.0420787628124435, + "grad_norm": 2.0977205411572295, + "learning_rate": 1.0235500880115627e-06, + "loss": 0.0185, + "step": 8400 + }, + { + "epoch": 6.042798057903255, + "grad_norm": 2.0181988217593916, + "learning_rate": 1.0232274756145534e-06, + "loss": 0.0164, + "step": 8401 + }, + { + "epoch": 6.043517352994066, + "grad_norm": 1.0034454718304202, + "learning_rate": 1.022904887748328e-06, + "loss": 0.0048, + "step": 8402 + }, + { + "epoch": 6.044236648084877, + "grad_norm": 4.362789624274764, + "learning_rate": 1.0225823244294834e-06, + "loss": 0.0888, + "step": 8403 + }, + { + "epoch": 6.044955943175688, + "grad_norm": 2.988330687742684, + "learning_rate": 1.022259785674616e-06, + "loss": 0.0464, + "step": 8404 + }, + { + "epoch": 6.045675238266499, + "grad_norm": 1.7430545882644701, + "learning_rate": 1.0219372715003215e-06, + "loss": 0.0307, + "step": 8405 + }, + { + "epoch": 6.04639453335731, + "grad_norm": 2.8756332109906055, + "learning_rate": 1.0216147819231935e-06, + "loss": 0.03, + "step": 8406 + }, + { + "epoch": 6.047113828448121, + "grad_norm": 7.0672225302343525, + "learning_rate": 1.021292316959825e-06, + "loss": 0.1446, + "step": 8407 + }, + { + "epoch": 6.047833123538932, + "grad_norm": 0.3696625638950881, + "learning_rate": 1.0209698766268072e-06, + "loss": 0.0026, + "step": 8408 + }, + { + "epoch": 6.048552418629743, + "grad_norm": 3.0272786530177376, + "learning_rate": 1.0206474609407294e-06, + "loss": 0.0645, + "step": 8409 + }, + { + "epoch": 6.049271713720554, + "grad_norm": 2.0442063363273313, + "learning_rate": 1.0203250699181816e-06, + "loss": 0.0135, + "step": 8410 + }, + { + "epoch": 6.049991008811364, + "grad_norm": 2.973601475138253, + "learning_rate": 1.0200027035757513e-06, + "loss": 0.0432, + "step": 8411 + }, + { + "epoch": 6.050710303902176, + "grad_norm": 2.474538568409765, + "learning_rate": 1.0196803619300244e-06, + "loss": 0.0097, + "step": 8412 + }, + { + "epoch": 6.0514295989929865, + "grad_norm": 2.5474124731936683, + "learning_rate": 1.0193580449975865e-06, + "loss": 0.0238, + "step": 8413 + }, + { + "epoch": 6.052148894083798, + "grad_norm": 2.787729235115647, + "learning_rate": 1.0190357527950198e-06, + "loss": 0.0232, + "step": 8414 + }, + { + "epoch": 6.052868189174609, + "grad_norm": 1.1900640892057746, + "learning_rate": 1.0187134853389087e-06, + "loss": 0.018, + "step": 8415 + }, + { + "epoch": 6.05358748426542, + "grad_norm": 0.7923593711493319, + "learning_rate": 1.0183912426458337e-06, + "loss": 0.0094, + "step": 8416 + }, + { + "epoch": 6.054306779356231, + "grad_norm": 2.305839625309817, + "learning_rate": 1.0180690247323742e-06, + "loss": 0.027, + "step": 8417 + }, + { + "epoch": 6.055026074447042, + "grad_norm": 1.9084447406632408, + "learning_rate": 1.01774683161511e-06, + "loss": 0.0243, + "step": 8418 + }, + { + "epoch": 6.055745369537853, + "grad_norm": 1.0776640881732706, + "learning_rate": 1.0174246633106178e-06, + "loss": 0.0096, + "step": 8419 + }, + { + "epoch": 6.056464664628664, + "grad_norm": 1.6603484988478745, + "learning_rate": 1.0171025198354732e-06, + "loss": 0.0063, + "step": 8420 + }, + { + "epoch": 6.057183959719475, + "grad_norm": 2.9028259844232176, + "learning_rate": 1.016780401206252e-06, + "loss": 0.0115, + "step": 8421 + }, + { + "epoch": 6.057903254810286, + "grad_norm": 2.7025724555299115, + "learning_rate": 1.0164583074395277e-06, + "loss": 0.0293, + "step": 8422 + }, + { + "epoch": 6.058622549901097, + "grad_norm": 0.7874191847635824, + "learning_rate": 1.016136238551872e-06, + "loss": 0.0061, + "step": 8423 + }, + { + "epoch": 6.059341844991908, + "grad_norm": 0.20334812557603985, + "learning_rate": 1.0158141945598565e-06, + "loss": 0.0005, + "step": 8424 + }, + { + "epoch": 6.060061140082719, + "grad_norm": 2.305569802649439, + "learning_rate": 1.0154921754800502e-06, + "loss": 0.0078, + "step": 8425 + }, + { + "epoch": 6.06078043517353, + "grad_norm": 2.593081048751438, + "learning_rate": 1.0151701813290223e-06, + "loss": 0.0344, + "step": 8426 + }, + { + "epoch": 6.061499730264341, + "grad_norm": 3.9503486150767664, + "learning_rate": 1.0148482121233394e-06, + "loss": 0.0524, + "step": 8427 + }, + { + "epoch": 6.062219025355152, + "grad_norm": 3.1466490931057387, + "learning_rate": 1.014526267879568e-06, + "loss": 0.0231, + "step": 8428 + }, + { + "epoch": 6.062938320445963, + "grad_norm": 1.059863306690755, + "learning_rate": 1.0142043486142722e-06, + "loss": 0.0064, + "step": 8429 + }, + { + "epoch": 6.063657615536774, + "grad_norm": 2.0599329483492, + "learning_rate": 1.013882454344015e-06, + "loss": 0.0117, + "step": 8430 + }, + { + "epoch": 6.064376910627585, + "grad_norm": 4.174621348424846, + "learning_rate": 1.0135605850853595e-06, + "loss": 0.0414, + "step": 8431 + }, + { + "epoch": 6.065096205718396, + "grad_norm": 0.02989400790221348, + "learning_rate": 1.0132387408548657e-06, + "loss": 0.0002, + "step": 8432 + }, + { + "epoch": 6.065815500809207, + "grad_norm": 1.4873012375179493, + "learning_rate": 1.012916921669093e-06, + "loss": 0.0126, + "step": 8433 + }, + { + "epoch": 6.066534795900018, + "grad_norm": 0.5667758064304308, + "learning_rate": 1.0125951275445999e-06, + "loss": 0.0051, + "step": 8434 + }, + { + "epoch": 6.067254090990829, + "grad_norm": 2.6233361972782605, + "learning_rate": 1.0122733584979429e-06, + "loss": 0.0482, + "step": 8435 + }, + { + "epoch": 6.06797338608164, + "grad_norm": 1.2361848075374684, + "learning_rate": 1.011951614545678e-06, + "loss": 0.009, + "step": 8436 + }, + { + "epoch": 6.068692681172451, + "grad_norm": 0.8981702393291024, + "learning_rate": 1.0116298957043592e-06, + "loss": 0.0036, + "step": 8437 + }, + { + "epoch": 6.069411976263262, + "grad_norm": 0.019846749736468987, + "learning_rate": 1.0113082019905396e-06, + "loss": 0.0001, + "step": 8438 + }, + { + "epoch": 6.070131271354073, + "grad_norm": 4.439646489414389, + "learning_rate": 1.010986533420771e-06, + "loss": 0.0504, + "step": 8439 + }, + { + "epoch": 6.070850566444884, + "grad_norm": 4.610895225759734, + "learning_rate": 1.0106648900116035e-06, + "loss": 0.0553, + "step": 8440 + }, + { + "epoch": 6.0715698615356954, + "grad_norm": 3.1163014236558415, + "learning_rate": 1.010343271779586e-06, + "loss": 0.0236, + "step": 8441 + }, + { + "epoch": 6.072289156626506, + "grad_norm": 3.3620458423760176, + "learning_rate": 1.0100216787412675e-06, + "loss": 0.0685, + "step": 8442 + }, + { + "epoch": 6.073008451717317, + "grad_norm": 5.911378765302203, + "learning_rate": 1.0097001109131934e-06, + "loss": 0.0685, + "step": 8443 + }, + { + "epoch": 6.073727746808128, + "grad_norm": 0.07020470450047181, + "learning_rate": 1.0093785683119093e-06, + "loss": 0.0002, + "step": 8444 + }, + { + "epoch": 6.074447041898939, + "grad_norm": 0.16261135055324827, + "learning_rate": 1.0090570509539593e-06, + "loss": 0.0005, + "step": 8445 + }, + { + "epoch": 6.07516633698975, + "grad_norm": 1.279069804609872, + "learning_rate": 1.008735558855885e-06, + "loss": 0.0136, + "step": 8446 + }, + { + "epoch": 6.075885632080561, + "grad_norm": 2.1748938247059373, + "learning_rate": 1.0084140920342295e-06, + "loss": 0.0048, + "step": 8447 + }, + { + "epoch": 6.076604927171372, + "grad_norm": 2.3812630603129064, + "learning_rate": 1.008092650505532e-06, + "loss": 0.0247, + "step": 8448 + }, + { + "epoch": 6.077324222262183, + "grad_norm": 0.4821147286938914, + "learning_rate": 1.007771234286331e-06, + "loss": 0.0014, + "step": 8449 + }, + { + "epoch": 6.078043517352994, + "grad_norm": 1.2544893967097768, + "learning_rate": 1.0074498433931638e-06, + "loss": 0.017, + "step": 8450 + }, + { + "epoch": 6.078762812443805, + "grad_norm": 1.8983130369231784, + "learning_rate": 1.0071284778425665e-06, + "loss": 0.0078, + "step": 8451 + }, + { + "epoch": 6.079482107534616, + "grad_norm": 0.942780411141087, + "learning_rate": 1.0068071376510746e-06, + "loss": 0.0016, + "step": 8452 + }, + { + "epoch": 6.080201402625427, + "grad_norm": 0.020054138068458974, + "learning_rate": 1.0064858228352215e-06, + "loss": 0.0001, + "step": 8453 + }, + { + "epoch": 6.080920697716238, + "grad_norm": 0.39225621524582893, + "learning_rate": 1.006164533411539e-06, + "loss": 0.0003, + "step": 8454 + }, + { + "epoch": 6.081639992807049, + "grad_norm": 0.8573809728619263, + "learning_rate": 1.005843269396558e-06, + "loss": 0.0065, + "step": 8455 + }, + { + "epoch": 6.0823592878978605, + "grad_norm": 0.040861905433969135, + "learning_rate": 1.0055220308068082e-06, + "loss": 0.0, + "step": 8456 + }, + { + "epoch": 6.083078582988671, + "grad_norm": 5.213431125588534, + "learning_rate": 1.0052008176588182e-06, + "loss": 0.0793, + "step": 8457 + }, + { + "epoch": 6.083797878079482, + "grad_norm": 0.46042687127137555, + "learning_rate": 1.0048796299691145e-06, + "loss": 0.0019, + "step": 8458 + }, + { + "epoch": 6.084517173170293, + "grad_norm": 4.551255681462903, + "learning_rate": 1.0045584677542232e-06, + "loss": 0.0678, + "step": 8459 + }, + { + "epoch": 6.085236468261104, + "grad_norm": 0.06419780088016096, + "learning_rate": 1.0042373310306684e-06, + "loss": 0.0005, + "step": 8460 + }, + { + "epoch": 6.085955763351915, + "grad_norm": 3.243984053657666, + "learning_rate": 1.0039162198149733e-06, + "loss": 0.0283, + "step": 8461 + }, + { + "epoch": 6.086675058442726, + "grad_norm": 1.384696745279857, + "learning_rate": 1.0035951341236591e-06, + "loss": 0.0179, + "step": 8462 + }, + { + "epoch": 6.087394353533537, + "grad_norm": 2.4895908506089093, + "learning_rate": 1.003274073973247e-06, + "loss": 0.0392, + "step": 8463 + }, + { + "epoch": 6.088113648624348, + "grad_norm": 3.1157307989043455, + "learning_rate": 1.002953039380256e-06, + "loss": 0.0387, + "step": 8464 + }, + { + "epoch": 6.088832943715159, + "grad_norm": 2.7464534115711636, + "learning_rate": 1.0026320303612033e-06, + "loss": 0.0181, + "step": 8465 + }, + { + "epoch": 6.08955223880597, + "grad_norm": 4.868606040368892, + "learning_rate": 1.002311046932606e-06, + "loss": 0.0051, + "step": 8466 + }, + { + "epoch": 6.090271533896781, + "grad_norm": 3.9322612163070687, + "learning_rate": 1.001990089110978e-06, + "loss": 0.0655, + "step": 8467 + }, + { + "epoch": 6.090990828987592, + "grad_norm": 2.528099736766746, + "learning_rate": 1.0016691569128353e-06, + "loss": 0.0253, + "step": 8468 + }, + { + "epoch": 6.0917101240784035, + "grad_norm": 1.2635516441965084, + "learning_rate": 1.001348250354689e-06, + "loss": 0.0021, + "step": 8469 + }, + { + "epoch": 6.092429419169214, + "grad_norm": 0.39845343916548387, + "learning_rate": 1.0010273694530507e-06, + "loss": 0.003, + "step": 8470 + }, + { + "epoch": 6.093148714260026, + "grad_norm": 1.0079716181926737, + "learning_rate": 1.00070651422443e-06, + "loss": 0.0049, + "step": 8471 + }, + { + "epoch": 6.093868009350836, + "grad_norm": 6.289485708249232, + "learning_rate": 1.0003856846853354e-06, + "loss": 0.0585, + "step": 8472 + }, + { + "epoch": 6.094587304441647, + "grad_norm": 0.1946674201246327, + "learning_rate": 1.0000648808522746e-06, + "loss": 0.0004, + "step": 8473 + }, + { + "epoch": 6.095306599532458, + "grad_norm": 3.4330590273054082, + "learning_rate": 9.997441027417534e-07, + "loss": 0.0734, + "step": 8474 + }, + { + "epoch": 6.096025894623269, + "grad_norm": 6.911493118441967, + "learning_rate": 9.994233503702761e-07, + "loss": 0.09, + "step": 8475 + }, + { + "epoch": 6.09674518971408, + "grad_norm": 3.678270198071335, + "learning_rate": 9.991026237543461e-07, + "loss": 0.0541, + "step": 8476 + }, + { + "epoch": 6.097464484804891, + "grad_norm": 2.1076511935487705, + "learning_rate": 9.987819229104653e-07, + "loss": 0.024, + "step": 8477 + }, + { + "epoch": 6.098183779895702, + "grad_norm": 0.025833293949921034, + "learning_rate": 9.984612478551348e-07, + "loss": 0.0001, + "step": 8478 + }, + { + "epoch": 6.098903074986513, + "grad_norm": 4.619692683577929, + "learning_rate": 9.981405986048532e-07, + "loss": 0.0542, + "step": 8479 + }, + { + "epoch": 6.099622370077324, + "grad_norm": 1.7909571602468113, + "learning_rate": 9.978199751761193e-07, + "loss": 0.0259, + "step": 8480 + }, + { + "epoch": 6.100341665168135, + "grad_norm": 1.0398290417629246, + "learning_rate": 9.974993775854289e-07, + "loss": 0.008, + "step": 8481 + }, + { + "epoch": 6.1010609602589465, + "grad_norm": 0.7264081045084986, + "learning_rate": 9.971788058492776e-07, + "loss": 0.0011, + "step": 8482 + }, + { + "epoch": 6.101780255349757, + "grad_norm": 0.009044899944708057, + "learning_rate": 9.968582599841587e-07, + "loss": 0.0, + "step": 8483 + }, + { + "epoch": 6.102499550440569, + "grad_norm": 3.014279164176206, + "learning_rate": 9.965377400065666e-07, + "loss": 0.0347, + "step": 8484 + }, + { + "epoch": 6.103218845531379, + "grad_norm": 4.428066117077157, + "learning_rate": 9.962172459329916e-07, + "loss": 0.088, + "step": 8485 + }, + { + "epoch": 6.103938140622191, + "grad_norm": 1.5354222779735067, + "learning_rate": 9.958967777799233e-07, + "loss": 0.0095, + "step": 8486 + }, + { + "epoch": 6.104657435713001, + "grad_norm": 0.1617444341225184, + "learning_rate": 9.955763355638508e-07, + "loss": 0.0004, + "step": 8487 + }, + { + "epoch": 6.105376730803812, + "grad_norm": 4.49699679775746, + "learning_rate": 9.952559193012607e-07, + "loss": 0.0327, + "step": 8488 + }, + { + "epoch": 6.106096025894623, + "grad_norm": 0.12238344579290962, + "learning_rate": 9.949355290086402e-07, + "loss": 0.0003, + "step": 8489 + }, + { + "epoch": 6.106815320985434, + "grad_norm": 0.10596699183290596, + "learning_rate": 9.946151647024736e-07, + "loss": 0.0005, + "step": 8490 + }, + { + "epoch": 6.107534616076245, + "grad_norm": 1.7340085384774355, + "learning_rate": 9.942948263992437e-07, + "loss": 0.008, + "step": 8491 + }, + { + "epoch": 6.108253911167056, + "grad_norm": 0.040591024350734216, + "learning_rate": 9.939745141154326e-07, + "loss": 0.0001, + "step": 8492 + }, + { + "epoch": 6.108973206257867, + "grad_norm": 0.9838167169368015, + "learning_rate": 9.93654227867521e-07, + "loss": 0.0038, + "step": 8493 + }, + { + "epoch": 6.109692501348678, + "grad_norm": 4.104391729034057, + "learning_rate": 9.933339676719885e-07, + "loss": 0.0607, + "step": 8494 + }, + { + "epoch": 6.1104117964394895, + "grad_norm": 0.051543901640092385, + "learning_rate": 9.930137335453126e-07, + "loss": 0.0002, + "step": 8495 + }, + { + "epoch": 6.1111310915303, + "grad_norm": 2.357017600646483, + "learning_rate": 9.9269352550397e-07, + "loss": 0.0311, + "step": 8496 + }, + { + "epoch": 6.111850386621112, + "grad_norm": 0.8468828227989329, + "learning_rate": 9.923733435644365e-07, + "loss": 0.0025, + "step": 8497 + }, + { + "epoch": 6.112569681711922, + "grad_norm": 0.30718332434706763, + "learning_rate": 9.920531877431847e-07, + "loss": 0.0013, + "step": 8498 + }, + { + "epoch": 6.113288976802734, + "grad_norm": 0.05101392848364552, + "learning_rate": 9.917330580566888e-07, + "loss": 0.0001, + "step": 8499 + }, + { + "epoch": 6.114008271893544, + "grad_norm": 5.787087390260647, + "learning_rate": 9.914129545214194e-07, + "loss": 0.0243, + "step": 8500 + }, + { + "epoch": 6.114727566984356, + "grad_norm": 2.003262670592137, + "learning_rate": 9.91092877153846e-07, + "loss": 0.033, + "step": 8501 + }, + { + "epoch": 6.115446862075166, + "grad_norm": 0.7126613360709099, + "learning_rate": 9.907728259704377e-07, + "loss": 0.0059, + "step": 8502 + }, + { + "epoch": 6.116166157165978, + "grad_norm": 1.5480393967246622, + "learning_rate": 9.90452800987661e-07, + "loss": 0.0148, + "step": 8503 + }, + { + "epoch": 6.116885452256788, + "grad_norm": 3.5162084463376893, + "learning_rate": 9.90132802221982e-07, + "loss": 0.0629, + "step": 8504 + }, + { + "epoch": 6.117604747347599, + "grad_norm": 0.42046874342520196, + "learning_rate": 9.898128296898657e-07, + "loss": 0.0026, + "step": 8505 + }, + { + "epoch": 6.11832404243841, + "grad_norm": 1.4435547170155343, + "learning_rate": 9.89492883407775e-07, + "loss": 0.0195, + "step": 8506 + }, + { + "epoch": 6.119043337529221, + "grad_norm": 3.3242291386963423, + "learning_rate": 9.891729633921719e-07, + "loss": 0.026, + "step": 8507 + }, + { + "epoch": 6.1197626326200325, + "grad_norm": 3.0049783073322627, + "learning_rate": 9.888530696595162e-07, + "loss": 0.0063, + "step": 8508 + }, + { + "epoch": 6.120481927710843, + "grad_norm": 0.053378868052763845, + "learning_rate": 9.88533202226267e-07, + "loss": 0.0003, + "step": 8509 + }, + { + "epoch": 6.121201222801655, + "grad_norm": 1.8226577971557256, + "learning_rate": 9.882133611088827e-07, + "loss": 0.0232, + "step": 8510 + }, + { + "epoch": 6.121920517892465, + "grad_norm": 0.15385094741905508, + "learning_rate": 9.878935463238194e-07, + "loss": 0.0004, + "step": 8511 + }, + { + "epoch": 6.122639812983277, + "grad_norm": 2.253958535485105, + "learning_rate": 9.875737578875324e-07, + "loss": 0.0306, + "step": 8512 + }, + { + "epoch": 6.123359108074087, + "grad_norm": 1.5970306382103086, + "learning_rate": 9.87253995816475e-07, + "loss": 0.0154, + "step": 8513 + }, + { + "epoch": 6.124078403164899, + "grad_norm": 1.3443153119918532, + "learning_rate": 9.869342601270992e-07, + "loss": 0.0057, + "step": 8514 + }, + { + "epoch": 6.124797698255709, + "grad_norm": 0.031830542236788995, + "learning_rate": 9.866145508358566e-07, + "loss": 0.0001, + "step": 8515 + }, + { + "epoch": 6.125516993346521, + "grad_norm": 0.03732297138782374, + "learning_rate": 9.862948679591967e-07, + "loss": 0.0001, + "step": 8516 + }, + { + "epoch": 6.126236288437331, + "grad_norm": 3.01698615974581, + "learning_rate": 9.859752115135677e-07, + "loss": 0.0197, + "step": 8517 + }, + { + "epoch": 6.126955583528143, + "grad_norm": 3.1780375158930734, + "learning_rate": 9.856555815154162e-07, + "loss": 0.0434, + "step": 8518 + }, + { + "epoch": 6.127674878618953, + "grad_norm": 1.6133261317203802, + "learning_rate": 9.853359779811876e-07, + "loss": 0.0181, + "step": 8519 + }, + { + "epoch": 6.128394173709764, + "grad_norm": 2.393466024359795, + "learning_rate": 9.85016400927327e-07, + "loss": 0.022, + "step": 8520 + }, + { + "epoch": 6.1291134688005755, + "grad_norm": 0.45512390575253686, + "learning_rate": 9.846968503702765e-07, + "loss": 0.006, + "step": 8521 + }, + { + "epoch": 6.129832763891386, + "grad_norm": 2.1274534437299173, + "learning_rate": 9.843773263264776e-07, + "loss": 0.0156, + "step": 8522 + }, + { + "epoch": 6.130552058982198, + "grad_norm": 2.9157924532047343, + "learning_rate": 9.840578288123704e-07, + "loss": 0.0466, + "step": 8523 + }, + { + "epoch": 6.131271354073008, + "grad_norm": 0.018555166187732484, + "learning_rate": 9.837383578443937e-07, + "loss": 0.0001, + "step": 8524 + }, + { + "epoch": 6.13199064916382, + "grad_norm": 2.2100877699966603, + "learning_rate": 9.83418913438984e-07, + "loss": 0.009, + "step": 8525 + }, + { + "epoch": 6.13270994425463, + "grad_norm": 1.1523276209699829, + "learning_rate": 9.830994956125788e-07, + "loss": 0.0109, + "step": 8526 + }, + { + "epoch": 6.133429239345442, + "grad_norm": 2.201654155913906, + "learning_rate": 9.827801043816118e-07, + "loss": 0.0127, + "step": 8527 + }, + { + "epoch": 6.134148534436252, + "grad_norm": 0.20951605851953184, + "learning_rate": 9.824607397625163e-07, + "loss": 0.0008, + "step": 8528 + }, + { + "epoch": 6.134867829527064, + "grad_norm": 0.3139299909880094, + "learning_rate": 9.821414017717243e-07, + "loss": 0.0022, + "step": 8529 + }, + { + "epoch": 6.135587124617874, + "grad_norm": 2.0673530157553546, + "learning_rate": 9.818220904256662e-07, + "loss": 0.0082, + "step": 8530 + }, + { + "epoch": 6.136306419708686, + "grad_norm": 3.4344319414068307, + "learning_rate": 9.815028057407711e-07, + "loss": 0.0591, + "step": 8531 + }, + { + "epoch": 6.137025714799496, + "grad_norm": 3.9369478005501377, + "learning_rate": 9.811835477334672e-07, + "loss": 0.0312, + "step": 8532 + }, + { + "epoch": 6.137745009890308, + "grad_norm": 2.240001845335932, + "learning_rate": 9.808643164201804e-07, + "loss": 0.02, + "step": 8533 + }, + { + "epoch": 6.1384643049811185, + "grad_norm": 0.7020130964632549, + "learning_rate": 9.80545111817336e-07, + "loss": 0.0044, + "step": 8534 + }, + { + "epoch": 6.139183600071929, + "grad_norm": 3.3440907981961265, + "learning_rate": 9.802259339413567e-07, + "loss": 0.0208, + "step": 8535 + }, + { + "epoch": 6.139902895162741, + "grad_norm": 4.768452685995969, + "learning_rate": 9.799067828086663e-07, + "loss": 0.0765, + "step": 8536 + }, + { + "epoch": 6.140622190253551, + "grad_norm": 5.077296209060865, + "learning_rate": 9.795876584356846e-07, + "loss": 0.0293, + "step": 8537 + }, + { + "epoch": 6.141341485344363, + "grad_norm": 1.4029540218757899, + "learning_rate": 9.792685608388317e-07, + "loss": 0.0075, + "step": 8538 + }, + { + "epoch": 6.142060780435173, + "grad_norm": 0.21132006521878496, + "learning_rate": 9.789494900345253e-07, + "loss": 0.0008, + "step": 8539 + }, + { + "epoch": 6.142780075525985, + "grad_norm": 3.41472293390178, + "learning_rate": 9.786304460391816e-07, + "loss": 0.0609, + "step": 8540 + }, + { + "epoch": 6.143499370616795, + "grad_norm": 2.113249691966292, + "learning_rate": 9.783114288692176e-07, + "loss": 0.0115, + "step": 8541 + }, + { + "epoch": 6.144218665707607, + "grad_norm": 3.2403226053872314, + "learning_rate": 9.77992438541046e-07, + "loss": 0.0298, + "step": 8542 + }, + { + "epoch": 6.144937960798417, + "grad_norm": 0.36339408821365193, + "learning_rate": 9.776734750710803e-07, + "loss": 0.0013, + "step": 8543 + }, + { + "epoch": 6.145657255889229, + "grad_norm": 2.6605807767494176, + "learning_rate": 9.773545384757308e-07, + "loss": 0.0157, + "step": 8544 + }, + { + "epoch": 6.146376550980039, + "grad_norm": 1.8692237315184714, + "learning_rate": 9.770356287714081e-07, + "loss": 0.0197, + "step": 8545 + }, + { + "epoch": 6.147095846070851, + "grad_norm": 3.9311926761663325, + "learning_rate": 9.767167459745194e-07, + "loss": 0.0318, + "step": 8546 + }, + { + "epoch": 6.1478151411616615, + "grad_norm": 0.814654681555939, + "learning_rate": 9.763978901014735e-07, + "loss": 0.0009, + "step": 8547 + }, + { + "epoch": 6.148534436252473, + "grad_norm": 0.0031918244426353687, + "learning_rate": 9.760790611686752e-07, + "loss": 0.0, + "step": 8548 + }, + { + "epoch": 6.149253731343284, + "grad_norm": 0.0031549120610918865, + "learning_rate": 9.757602591925289e-07, + "loss": 0.0, + "step": 8549 + }, + { + "epoch": 6.149973026434094, + "grad_norm": 1.2570293056160355, + "learning_rate": 9.754414841894378e-07, + "loss": 0.006, + "step": 8550 + }, + { + "epoch": 6.150692321524906, + "grad_norm": 0.0764333052661694, + "learning_rate": 9.751227361758026e-07, + "loss": 0.0001, + "step": 8551 + }, + { + "epoch": 6.151411616615716, + "grad_norm": 1.9227684323743406, + "learning_rate": 9.748040151680243e-07, + "loss": 0.0169, + "step": 8552 + }, + { + "epoch": 6.152130911706528, + "grad_norm": 3.629714741671002, + "learning_rate": 9.744853211825015e-07, + "loss": 0.0379, + "step": 8553 + }, + { + "epoch": 6.152850206797338, + "grad_norm": 0.8470387774669427, + "learning_rate": 9.741666542356313e-07, + "loss": 0.0053, + "step": 8554 + }, + { + "epoch": 6.15356950188815, + "grad_norm": 7.9938761303807535, + "learning_rate": 9.738480143438098e-07, + "loss": 0.1347, + "step": 8555 + }, + { + "epoch": 6.15428879697896, + "grad_norm": 1.1520238470151334, + "learning_rate": 9.735294015234315e-07, + "loss": 0.0095, + "step": 8556 + }, + { + "epoch": 6.155008092069772, + "grad_norm": 5.13412707962924, + "learning_rate": 9.7321081579089e-07, + "loss": 0.0492, + "step": 8557 + }, + { + "epoch": 6.155727387160582, + "grad_norm": 0.509356248994015, + "learning_rate": 9.728922571625765e-07, + "loss": 0.0017, + "step": 8558 + }, + { + "epoch": 6.156446682251394, + "grad_norm": 0.007640307224511344, + "learning_rate": 9.72573725654882e-07, + "loss": 0.0, + "step": 8559 + }, + { + "epoch": 6.1571659773422045, + "grad_norm": 2.6756424740806746, + "learning_rate": 9.722552212841949e-07, + "loss": 0.0148, + "step": 8560 + }, + { + "epoch": 6.157885272433016, + "grad_norm": 0.5900079624462072, + "learning_rate": 9.719367440669027e-07, + "loss": 0.0038, + "step": 8561 + }, + { + "epoch": 6.158604567523827, + "grad_norm": 1.0010981322568615, + "learning_rate": 9.716182940193924e-07, + "loss": 0.007, + "step": 8562 + }, + { + "epoch": 6.159323862614638, + "grad_norm": 5.9973697348538835, + "learning_rate": 9.712998711580485e-07, + "loss": 0.088, + "step": 8563 + }, + { + "epoch": 6.160043157705449, + "grad_norm": 5.0919060338717275, + "learning_rate": 9.709814754992545e-07, + "loss": 0.0663, + "step": 8564 + }, + { + "epoch": 6.160762452796259, + "grad_norm": 2.409177472071483, + "learning_rate": 9.70663107059392e-07, + "loss": 0.0119, + "step": 8565 + }, + { + "epoch": 6.161481747887071, + "grad_norm": 3.301915677584, + "learning_rate": 9.703447658548425e-07, + "loss": 0.0506, + "step": 8566 + }, + { + "epoch": 6.162201042977881, + "grad_norm": 3.701544030174339, + "learning_rate": 9.700264519019833e-07, + "loss": 0.0654, + "step": 8567 + }, + { + "epoch": 6.162920338068693, + "grad_norm": 0.40730387971481163, + "learning_rate": 9.697081652171942e-07, + "loss": 0.0018, + "step": 8568 + }, + { + "epoch": 6.163639633159503, + "grad_norm": 7.578171707690718, + "learning_rate": 9.693899058168507e-07, + "loss": 0.1238, + "step": 8569 + }, + { + "epoch": 6.164358928250315, + "grad_norm": 2.4180944893857332, + "learning_rate": 9.690716737173285e-07, + "loss": 0.0207, + "step": 8570 + }, + { + "epoch": 6.165078223341125, + "grad_norm": 3.3201851490104763, + "learning_rate": 9.687534689350004e-07, + "loss": 0.0466, + "step": 8571 + }, + { + "epoch": 6.165797518431937, + "grad_norm": 6.736263803668472, + "learning_rate": 9.684352914862388e-07, + "loss": 0.0696, + "step": 8572 + }, + { + "epoch": 6.1665168135227475, + "grad_norm": 0.022954221922700187, + "learning_rate": 9.681171413874146e-07, + "loss": 0.0001, + "step": 8573 + }, + { + "epoch": 6.167236108613559, + "grad_norm": 2.109099271856594, + "learning_rate": 9.677990186548972e-07, + "loss": 0.0151, + "step": 8574 + }, + { + "epoch": 6.16795540370437, + "grad_norm": 3.4917433704725216, + "learning_rate": 9.674809233050547e-07, + "loss": 0.0137, + "step": 8575 + }, + { + "epoch": 6.168674698795181, + "grad_norm": 3.6002272104701603, + "learning_rate": 9.671628553542536e-07, + "loss": 0.044, + "step": 8576 + }, + { + "epoch": 6.169393993885992, + "grad_norm": 1.5563726402974796, + "learning_rate": 9.66844814818858e-07, + "loss": 0.0154, + "step": 8577 + }, + { + "epoch": 6.170113288976803, + "grad_norm": 1.9854407794212072, + "learning_rate": 9.665268017152332e-07, + "loss": 0.0088, + "step": 8578 + }, + { + "epoch": 6.170832584067614, + "grad_norm": 1.7711785274926892, + "learning_rate": 9.662088160597416e-07, + "loss": 0.0069, + "step": 8579 + }, + { + "epoch": 6.171551879158425, + "grad_norm": 1.6792187465261341, + "learning_rate": 9.65890857868743e-07, + "loss": 0.0115, + "step": 8580 + }, + { + "epoch": 6.172271174249236, + "grad_norm": 2.748599768515513, + "learning_rate": 9.655729271585972e-07, + "loss": 0.0283, + "step": 8581 + }, + { + "epoch": 6.172990469340046, + "grad_norm": 0.037668857617870526, + "learning_rate": 9.652550239456622e-07, + "loss": 0.0001, + "step": 8582 + }, + { + "epoch": 6.173709764430858, + "grad_norm": 0.2508474357629885, + "learning_rate": 9.649371482462953e-07, + "loss": 0.0013, + "step": 8583 + }, + { + "epoch": 6.174429059521668, + "grad_norm": 7.977318198921218, + "learning_rate": 9.646193000768515e-07, + "loss": 0.1022, + "step": 8584 + }, + { + "epoch": 6.17514835461248, + "grad_norm": 1.8956357168679108, + "learning_rate": 9.643014794536844e-07, + "loss": 0.0215, + "step": 8585 + }, + { + "epoch": 6.1758676497032905, + "grad_norm": 0.26509182044689217, + "learning_rate": 9.639836863931466e-07, + "loss": 0.0007, + "step": 8586 + }, + { + "epoch": 6.176586944794102, + "grad_norm": 2.228750109597591, + "learning_rate": 9.63665920911589e-07, + "loss": 0.0092, + "step": 8587 + }, + { + "epoch": 6.177306239884913, + "grad_norm": 3.9551492005583326, + "learning_rate": 9.633481830253612e-07, + "loss": 0.0606, + "step": 8588 + }, + { + "epoch": 6.178025534975724, + "grad_norm": 3.698079806385527, + "learning_rate": 9.630304727508116e-07, + "loss": 0.0393, + "step": 8589 + }, + { + "epoch": 6.178744830066535, + "grad_norm": 3.835530458454227, + "learning_rate": 9.627127901042868e-07, + "loss": 0.0264, + "step": 8590 + }, + { + "epoch": 6.179464125157346, + "grad_norm": 2.3762003603140625, + "learning_rate": 9.623951351021322e-07, + "loss": 0.0305, + "step": 8591 + }, + { + "epoch": 6.180183420248157, + "grad_norm": 0.817599062623928, + "learning_rate": 9.620775077606916e-07, + "loss": 0.0062, + "step": 8592 + }, + { + "epoch": 6.180902715338968, + "grad_norm": 1.618197521422518, + "learning_rate": 9.617599080963073e-07, + "loss": 0.023, + "step": 8593 + }, + { + "epoch": 6.181622010429779, + "grad_norm": 2.2618816160050703, + "learning_rate": 9.61442336125321e-07, + "loss": 0.0102, + "step": 8594 + }, + { + "epoch": 6.18234130552059, + "grad_norm": 3.4286395574326005, + "learning_rate": 9.611247918640715e-07, + "loss": 0.0279, + "step": 8595 + }, + { + "epoch": 6.183060600611401, + "grad_norm": 3.486200490744778, + "learning_rate": 9.608072753288976e-07, + "loss": 0.0588, + "step": 8596 + }, + { + "epoch": 6.183779895702211, + "grad_norm": 4.679796175779051, + "learning_rate": 9.60489786536136e-07, + "loss": 0.0544, + "step": 8597 + }, + { + "epoch": 6.184499190793023, + "grad_norm": 9.367199866191028, + "learning_rate": 9.601723255021211e-07, + "loss": 0.2251, + "step": 8598 + }, + { + "epoch": 6.1852184858838335, + "grad_norm": 0.32409126264178384, + "learning_rate": 9.598548922431888e-07, + "loss": 0.0023, + "step": 8599 + }, + { + "epoch": 6.185937780974645, + "grad_norm": 0.40725695855349753, + "learning_rate": 9.595374867756701e-07, + "loss": 0.0015, + "step": 8600 + }, + { + "epoch": 6.186657076065456, + "grad_norm": 5.609016064888883, + "learning_rate": 9.592201091158964e-07, + "loss": 0.0639, + "step": 8601 + }, + { + "epoch": 6.187376371156267, + "grad_norm": 5.188074343370156, + "learning_rate": 9.58902759280198e-07, + "loss": 0.0525, + "step": 8602 + }, + { + "epoch": 6.188095666247078, + "grad_norm": 4.686387779334003, + "learning_rate": 9.585854372849015e-07, + "loss": 0.0738, + "step": 8603 + }, + { + "epoch": 6.188814961337889, + "grad_norm": 1.644534159754913, + "learning_rate": 9.58268143146335e-07, + "loss": 0.0195, + "step": 8604 + }, + { + "epoch": 6.1895342564287, + "grad_norm": 0.027747916583594195, + "learning_rate": 9.579508768808237e-07, + "loss": 0.0001, + "step": 8605 + }, + { + "epoch": 6.190253551519511, + "grad_norm": 5.945371484117064, + "learning_rate": 9.576336385046914e-07, + "loss": 0.1041, + "step": 8606 + }, + { + "epoch": 6.190972846610322, + "grad_norm": 1.6510444168366114, + "learning_rate": 9.573164280342604e-07, + "loss": 0.0168, + "step": 8607 + }, + { + "epoch": 6.191692141701133, + "grad_norm": 3.0059180804408228, + "learning_rate": 9.569992454858518e-07, + "loss": 0.0485, + "step": 8608 + }, + { + "epoch": 6.192411436791944, + "grad_norm": 6.136701836566348, + "learning_rate": 9.566820908757852e-07, + "loss": 0.0441, + "step": 8609 + }, + { + "epoch": 6.193130731882755, + "grad_norm": 1.3918659980115966, + "learning_rate": 9.563649642203787e-07, + "loss": 0.0094, + "step": 8610 + }, + { + "epoch": 6.193850026973566, + "grad_norm": 4.04021005718987, + "learning_rate": 9.560478655359494e-07, + "loss": 0.0686, + "step": 8611 + }, + { + "epoch": 6.1945693220643765, + "grad_norm": 3.4765792412013514, + "learning_rate": 9.557307948388124e-07, + "loss": 0.062, + "step": 8612 + }, + { + "epoch": 6.195288617155188, + "grad_norm": 0.47442354772227674, + "learning_rate": 9.554137521452813e-07, + "loss": 0.0027, + "step": 8613 + }, + { + "epoch": 6.196007912245999, + "grad_norm": 0.1468507278071725, + "learning_rate": 9.550967374716679e-07, + "loss": 0.0008, + "step": 8614 + }, + { + "epoch": 6.19672720733681, + "grad_norm": 3.993196198768846, + "learning_rate": 9.547797508342849e-07, + "loss": 0.0247, + "step": 8615 + }, + { + "epoch": 6.197446502427621, + "grad_norm": 3.3175521497775495, + "learning_rate": 9.544627922494405e-07, + "loss": 0.046, + "step": 8616 + }, + { + "epoch": 6.198165797518432, + "grad_norm": 3.0412456157727896, + "learning_rate": 9.54145861733443e-07, + "loss": 0.0478, + "step": 8617 + }, + { + "epoch": 6.198885092609243, + "grad_norm": 3.526395459358004, + "learning_rate": 9.538289593025991e-07, + "loss": 0.0569, + "step": 8618 + }, + { + "epoch": 6.199604387700054, + "grad_norm": 0.08113253632930385, + "learning_rate": 9.535120849732137e-07, + "loss": 0.0003, + "step": 8619 + }, + { + "epoch": 6.200323682790865, + "grad_norm": 7.387011210811843, + "learning_rate": 9.531952387615914e-07, + "loss": 0.1625, + "step": 8620 + }, + { + "epoch": 6.201042977881676, + "grad_norm": 2.173440235331433, + "learning_rate": 9.528784206840338e-07, + "loss": 0.0171, + "step": 8621 + }, + { + "epoch": 6.201762272972487, + "grad_norm": 1.8149659365057547, + "learning_rate": 9.52561630756842e-07, + "loss": 0.0176, + "step": 8622 + }, + { + "epoch": 6.202481568063298, + "grad_norm": 0.003867296765008353, + "learning_rate": 9.522448689963152e-07, + "loss": 0.0, + "step": 8623 + }, + { + "epoch": 6.203200863154109, + "grad_norm": 0.2543036608886057, + "learning_rate": 9.519281354187512e-07, + "loss": 0.001, + "step": 8624 + }, + { + "epoch": 6.20392015824492, + "grad_norm": 3.8910797151070944, + "learning_rate": 9.51611430040447e-07, + "loss": 0.0644, + "step": 8625 + }, + { + "epoch": 6.204639453335731, + "grad_norm": 1.3677200868662456, + "learning_rate": 9.512947528776975e-07, + "loss": 0.0116, + "step": 8626 + }, + { + "epoch": 6.205358748426542, + "grad_norm": 0.42307789875818574, + "learning_rate": 9.509781039467961e-07, + "loss": 0.0017, + "step": 8627 + }, + { + "epoch": 6.206078043517353, + "grad_norm": 1.3826641677674756, + "learning_rate": 9.506614832640351e-07, + "loss": 0.0152, + "step": 8628 + }, + { + "epoch": 6.206797338608164, + "grad_norm": 3.303671847725123, + "learning_rate": 9.503448908457046e-07, + "loss": 0.0418, + "step": 8629 + }, + { + "epoch": 6.207516633698975, + "grad_norm": 1.4432334168938725, + "learning_rate": 9.500283267080949e-07, + "loss": 0.0138, + "step": 8630 + }, + { + "epoch": 6.208235928789786, + "grad_norm": 1.9041323808527733, + "learning_rate": 9.497117908674932e-07, + "loss": 0.016, + "step": 8631 + }, + { + "epoch": 6.208955223880597, + "grad_norm": 2.467797304979863, + "learning_rate": 9.493952833401858e-07, + "loss": 0.0168, + "step": 8632 + }, + { + "epoch": 6.209674518971408, + "grad_norm": 2.3450066129806584, + "learning_rate": 9.490788041424576e-07, + "loss": 0.0204, + "step": 8633 + }, + { + "epoch": 6.210393814062219, + "grad_norm": 2.822121571411155, + "learning_rate": 9.487623532905923e-07, + "loss": 0.0246, + "step": 8634 + }, + { + "epoch": 6.21111310915303, + "grad_norm": 0.51732023728849, + "learning_rate": 9.484459308008707e-07, + "loss": 0.0013, + "step": 8635 + }, + { + "epoch": 6.211832404243841, + "grad_norm": 2.6461739371140878, + "learning_rate": 9.481295366895747e-07, + "loss": 0.0343, + "step": 8636 + }, + { + "epoch": 6.212551699334652, + "grad_norm": 1.8244704184068985, + "learning_rate": 9.478131709729831e-07, + "loss": 0.0113, + "step": 8637 + }, + { + "epoch": 6.213270994425463, + "grad_norm": 3.797255537091449, + "learning_rate": 9.474968336673733e-07, + "loss": 0.0351, + "step": 8638 + }, + { + "epoch": 6.213990289516274, + "grad_norm": 2.4337269302475595, + "learning_rate": 9.47180524789021e-07, + "loss": 0.0234, + "step": 8639 + }, + { + "epoch": 6.2147095846070854, + "grad_norm": 4.622773225364064, + "learning_rate": 9.468642443542007e-07, + "loss": 0.0459, + "step": 8640 + }, + { + "epoch": 6.215428879697896, + "grad_norm": 0.8955940598337416, + "learning_rate": 9.465479923791865e-07, + "loss": 0.0065, + "step": 8641 + }, + { + "epoch": 6.216148174788707, + "grad_norm": 2.2667563806827276, + "learning_rate": 9.462317688802497e-07, + "loss": 0.0211, + "step": 8642 + }, + { + "epoch": 6.216867469879518, + "grad_norm": 0.004415402522682699, + "learning_rate": 9.459155738736605e-07, + "loss": 0.0, + "step": 8643 + }, + { + "epoch": 6.217586764970329, + "grad_norm": 1.5955280101399354, + "learning_rate": 9.455994073756877e-07, + "loss": 0.0053, + "step": 8644 + }, + { + "epoch": 6.21830606006114, + "grad_norm": 0.13080581320092208, + "learning_rate": 9.452832694025982e-07, + "loss": 0.0006, + "step": 8645 + }, + { + "epoch": 6.219025355151951, + "grad_norm": 4.880843140338826, + "learning_rate": 9.449671599706586e-07, + "loss": 0.0685, + "step": 8646 + }, + { + "epoch": 6.219744650242762, + "grad_norm": 0.1643030485600769, + "learning_rate": 9.44651079096133e-07, + "loss": 0.0005, + "step": 8647 + }, + { + "epoch": 6.220463945333573, + "grad_norm": 2.5933684803750032, + "learning_rate": 9.443350267952843e-07, + "loss": 0.0325, + "step": 8648 + }, + { + "epoch": 6.221183240424384, + "grad_norm": 5.312223463690614, + "learning_rate": 9.440190030843739e-07, + "loss": 0.0893, + "step": 8649 + }, + { + "epoch": 6.221902535515195, + "grad_norm": 0.6279482362537864, + "learning_rate": 9.43703007979661e-07, + "loss": 0.0023, + "step": 8650 + }, + { + "epoch": 6.222621830606006, + "grad_norm": 4.544367845633379, + "learning_rate": 9.433870414974061e-07, + "loss": 0.0845, + "step": 8651 + }, + { + "epoch": 6.223341125696817, + "grad_norm": 4.130219562459496, + "learning_rate": 9.430711036538646e-07, + "loss": 0.0549, + "step": 8652 + }, + { + "epoch": 6.2240604207876284, + "grad_norm": 5.125280562083164, + "learning_rate": 9.427551944652923e-07, + "loss": 0.0614, + "step": 8653 + }, + { + "epoch": 6.224779715878439, + "grad_norm": 0.6887524499007363, + "learning_rate": 9.424393139479435e-07, + "loss": 0.0044, + "step": 8654 + }, + { + "epoch": 6.2254990109692505, + "grad_norm": 3.6417008043541, + "learning_rate": 9.421234621180709e-07, + "loss": 0.0503, + "step": 8655 + }, + { + "epoch": 6.226218306060061, + "grad_norm": 3.6032935961488493, + "learning_rate": 9.418076389919249e-07, + "loss": 0.0195, + "step": 8656 + }, + { + "epoch": 6.226937601150873, + "grad_norm": 3.731477663694508, + "learning_rate": 9.414918445857563e-07, + "loss": 0.0296, + "step": 8657 + }, + { + "epoch": 6.227656896241683, + "grad_norm": 3.3473785702996626, + "learning_rate": 9.411760789158128e-07, + "loss": 0.0334, + "step": 8658 + }, + { + "epoch": 6.228376191332494, + "grad_norm": 4.532055492784322, + "learning_rate": 9.408603419983409e-07, + "loss": 0.0456, + "step": 8659 + }, + { + "epoch": 6.229095486423305, + "grad_norm": 1.9890141870030162, + "learning_rate": 9.40544633849586e-07, + "loss": 0.0308, + "step": 8660 + }, + { + "epoch": 6.229814781514116, + "grad_norm": 2.0127354655288046, + "learning_rate": 9.402289544857914e-07, + "loss": 0.012, + "step": 8661 + }, + { + "epoch": 6.230534076604927, + "grad_norm": 1.0098205429481173, + "learning_rate": 9.399133039232002e-07, + "loss": 0.0094, + "step": 8662 + }, + { + "epoch": 6.231253371695738, + "grad_norm": 1.3106118672424405, + "learning_rate": 9.395976821780526e-07, + "loss": 0.0036, + "step": 8663 + }, + { + "epoch": 6.231972666786549, + "grad_norm": 5.769191258352344, + "learning_rate": 9.392820892665881e-07, + "loss": 0.0941, + "step": 8664 + }, + { + "epoch": 6.23269196187736, + "grad_norm": 4.3221816719543575, + "learning_rate": 9.389665252050443e-07, + "loss": 0.0369, + "step": 8665 + }, + { + "epoch": 6.233411256968171, + "grad_norm": 4.026717669641537, + "learning_rate": 9.386509900096573e-07, + "loss": 0.0422, + "step": 8666 + }, + { + "epoch": 6.234130552058982, + "grad_norm": 2.1896699732442437, + "learning_rate": 9.383354836966627e-07, + "loss": 0.0427, + "step": 8667 + }, + { + "epoch": 6.2348498471497935, + "grad_norm": 2.653961405963118, + "learning_rate": 9.380200062822934e-07, + "loss": 0.0365, + "step": 8668 + }, + { + "epoch": 6.235569142240604, + "grad_norm": 0.945724829859345, + "learning_rate": 9.377045577827812e-07, + "loss": 0.0166, + "step": 8669 + }, + { + "epoch": 6.236288437331416, + "grad_norm": 5.07719998888618, + "learning_rate": 9.373891382143568e-07, + "loss": 0.0289, + "step": 8670 + }, + { + "epoch": 6.237007732422226, + "grad_norm": 2.5578906340568452, + "learning_rate": 9.370737475932481e-07, + "loss": 0.0544, + "step": 8671 + }, + { + "epoch": 6.237727027513038, + "grad_norm": 2.3555512347510135, + "learning_rate": 9.367583859356839e-07, + "loss": 0.0067, + "step": 8672 + }, + { + "epoch": 6.238446322603848, + "grad_norm": 0.7235050670716757, + "learning_rate": 9.364430532578896e-07, + "loss": 0.0057, + "step": 8673 + }, + { + "epoch": 6.239165617694659, + "grad_norm": 1.2001034479562829, + "learning_rate": 9.361277495760899e-07, + "loss": 0.0108, + "step": 8674 + }, + { + "epoch": 6.23988491278547, + "grad_norm": 1.668637389241393, + "learning_rate": 9.358124749065067e-07, + "loss": 0.0288, + "step": 8675 + }, + { + "epoch": 6.240604207876281, + "grad_norm": 1.1802216307113136, + "learning_rate": 9.354972292653622e-07, + "loss": 0.0062, + "step": 8676 + }, + { + "epoch": 6.241323502967092, + "grad_norm": 3.4292542737887155, + "learning_rate": 9.351820126688756e-07, + "loss": 0.0531, + "step": 8677 + }, + { + "epoch": 6.242042798057903, + "grad_norm": 3.716804159077039, + "learning_rate": 9.348668251332666e-07, + "loss": 0.0597, + "step": 8678 + }, + { + "epoch": 6.242762093148714, + "grad_norm": 0.8503327840480184, + "learning_rate": 9.345516666747514e-07, + "loss": 0.0017, + "step": 8679 + }, + { + "epoch": 6.243481388239525, + "grad_norm": 3.830534777630263, + "learning_rate": 9.342365373095456e-07, + "loss": 0.0277, + "step": 8680 + }, + { + "epoch": 6.2442006833303365, + "grad_norm": 0.039867186051121876, + "learning_rate": 9.339214370538632e-07, + "loss": 0.0001, + "step": 8681 + }, + { + "epoch": 6.244919978421147, + "grad_norm": 2.3793342354653326, + "learning_rate": 9.336063659239163e-07, + "loss": 0.0493, + "step": 8682 + }, + { + "epoch": 6.245639273511959, + "grad_norm": 0.91028187284196, + "learning_rate": 9.332913239359163e-07, + "loss": 0.0072, + "step": 8683 + }, + { + "epoch": 6.246358568602769, + "grad_norm": 4.359702894381738, + "learning_rate": 9.329763111060727e-07, + "loss": 0.069, + "step": 8684 + }, + { + "epoch": 6.247077863693581, + "grad_norm": 3.9490695335872923, + "learning_rate": 9.32661327450593e-07, + "loss": 0.0383, + "step": 8685 + }, + { + "epoch": 6.247797158784391, + "grad_norm": 2.281645114492608, + "learning_rate": 9.323463729856843e-07, + "loss": 0.0183, + "step": 8686 + }, + { + "epoch": 6.248516453875203, + "grad_norm": 0.20277422399685688, + "learning_rate": 9.320314477275507e-07, + "loss": 0.0003, + "step": 8687 + }, + { + "epoch": 6.249235748966013, + "grad_norm": 4.416624770721724, + "learning_rate": 9.317165516923963e-07, + "loss": 0.026, + "step": 8688 + }, + { + "epoch": 6.249955044056824, + "grad_norm": 2.296892426955382, + "learning_rate": 9.314016848964233e-07, + "loss": 0.0299, + "step": 8689 + }, + { + "epoch": 6.250674339147635, + "grad_norm": 3.3234355239592084, + "learning_rate": 9.310868473558315e-07, + "loss": 0.0354, + "step": 8690 + }, + { + "epoch": 6.251393634238446, + "grad_norm": 1.4145914151438663, + "learning_rate": 9.3077203908682e-07, + "loss": 0.0096, + "step": 8691 + }, + { + "epoch": 6.252112929329257, + "grad_norm": 3.2264756262492624, + "learning_rate": 9.304572601055858e-07, + "loss": 0.0175, + "step": 8692 + }, + { + "epoch": 6.252832224420068, + "grad_norm": 4.634352900295368, + "learning_rate": 9.30142510428326e-07, + "loss": 0.0683, + "step": 8693 + }, + { + "epoch": 6.2535515195108795, + "grad_norm": 2.5394311855199296, + "learning_rate": 9.298277900712344e-07, + "loss": 0.0226, + "step": 8694 + }, + { + "epoch": 6.25427081460169, + "grad_norm": 3.8855576993994947, + "learning_rate": 9.295130990505041e-07, + "loss": 0.0553, + "step": 8695 + }, + { + "epoch": 6.254990109692502, + "grad_norm": 3.889150777042221, + "learning_rate": 9.29198437382326e-07, + "loss": 0.0371, + "step": 8696 + }, + { + "epoch": 6.255709404783312, + "grad_norm": 0.3770238000287799, + "learning_rate": 9.28883805082891e-07, + "loss": 0.0008, + "step": 8697 + }, + { + "epoch": 6.256428699874124, + "grad_norm": 1.858256984963748, + "learning_rate": 9.285692021683854e-07, + "loss": 0.0161, + "step": 8698 + }, + { + "epoch": 6.257147994964934, + "grad_norm": 8.306147147420543, + "learning_rate": 9.282546286549983e-07, + "loss": 0.1059, + "step": 8699 + }, + { + "epoch": 6.257867290055746, + "grad_norm": 2.604310327555615, + "learning_rate": 9.279400845589142e-07, + "loss": 0.0368, + "step": 8700 + }, + { + "epoch": 6.258586585146556, + "grad_norm": 0.6089531010202432, + "learning_rate": 9.27625569896317e-07, + "loss": 0.0034, + "step": 8701 + }, + { + "epoch": 6.259305880237368, + "grad_norm": 1.8808885666591442, + "learning_rate": 9.27311084683389e-07, + "loss": 0.0171, + "step": 8702 + }, + { + "epoch": 6.260025175328178, + "grad_norm": 3.628301524094177, + "learning_rate": 9.269966289363107e-07, + "loss": 0.0488, + "step": 8703 + }, + { + "epoch": 6.26074447041899, + "grad_norm": 0.1132887629675012, + "learning_rate": 9.266822026712622e-07, + "loss": 0.0006, + "step": 8704 + }, + { + "epoch": 6.2614637655098, + "grad_norm": 1.048856510844343, + "learning_rate": 9.263678059044205e-07, + "loss": 0.0011, + "step": 8705 + }, + { + "epoch": 6.262183060600611, + "grad_norm": 1.972456051149795, + "learning_rate": 9.260534386519623e-07, + "loss": 0.0175, + "step": 8706 + }, + { + "epoch": 6.2629023556914225, + "grad_norm": 1.6020931590920062, + "learning_rate": 9.257391009300621e-07, + "loss": 0.014, + "step": 8707 + }, + { + "epoch": 6.263621650782233, + "grad_norm": 0.4621919596308053, + "learning_rate": 9.254247927548928e-07, + "loss": 0.0023, + "step": 8708 + }, + { + "epoch": 6.264340945873045, + "grad_norm": 2.0402621682905604, + "learning_rate": 9.251105141426271e-07, + "loss": 0.0248, + "step": 8709 + }, + { + "epoch": 6.265060240963855, + "grad_norm": 2.48215087315461, + "learning_rate": 9.247962651094349e-07, + "loss": 0.0343, + "step": 8710 + }, + { + "epoch": 6.265779536054667, + "grad_norm": 3.1121830095584544, + "learning_rate": 9.244820456714844e-07, + "loss": 0.0326, + "step": 8711 + }, + { + "epoch": 6.266498831145477, + "grad_norm": 2.6439703865513993, + "learning_rate": 9.241678558449428e-07, + "loss": 0.0341, + "step": 8712 + }, + { + "epoch": 6.267218126236289, + "grad_norm": 5.543390632374985, + "learning_rate": 9.238536956459753e-07, + "loss": 0.1128, + "step": 8713 + }, + { + "epoch": 6.267937421327099, + "grad_norm": 1.9501171934688524, + "learning_rate": 9.235395650907472e-07, + "loss": 0.0171, + "step": 8714 + }, + { + "epoch": 6.268656716417911, + "grad_norm": 1.4094055219388597, + "learning_rate": 9.232254641954204e-07, + "loss": 0.0167, + "step": 8715 + }, + { + "epoch": 6.269376011508721, + "grad_norm": 4.9060089530287545, + "learning_rate": 9.229113929761559e-07, + "loss": 0.0699, + "step": 8716 + }, + { + "epoch": 6.270095306599533, + "grad_norm": 1.052369326031973, + "learning_rate": 9.225973514491134e-07, + "loss": 0.0094, + "step": 8717 + }, + { + "epoch": 6.270814601690343, + "grad_norm": 0.12890389440470423, + "learning_rate": 9.222833396304507e-07, + "loss": 0.0003, + "step": 8718 + }, + { + "epoch": 6.271533896781154, + "grad_norm": 0.2533103897210374, + "learning_rate": 9.219693575363239e-07, + "loss": 0.0002, + "step": 8719 + }, + { + "epoch": 6.2722531918719655, + "grad_norm": 5.381848186996461, + "learning_rate": 9.216554051828888e-07, + "loss": 0.0643, + "step": 8720 + }, + { + "epoch": 6.272972486962776, + "grad_norm": 1.109888953100217, + "learning_rate": 9.213414825862985e-07, + "loss": 0.0071, + "step": 8721 + }, + { + "epoch": 6.273691782053588, + "grad_norm": 3.3406435535386922, + "learning_rate": 9.210275897627044e-07, + "loss": 0.0312, + "step": 8722 + }, + { + "epoch": 6.274411077144398, + "grad_norm": 0.03550273753797966, + "learning_rate": 9.207137267282573e-07, + "loss": 0.0002, + "step": 8723 + }, + { + "epoch": 6.27513037223521, + "grad_norm": 3.877976887027119, + "learning_rate": 9.203998934991056e-07, + "loss": 0.0151, + "step": 8724 + }, + { + "epoch": 6.27584966732602, + "grad_norm": 2.50036782981627, + "learning_rate": 9.200860900913971e-07, + "loss": 0.0253, + "step": 8725 + }, + { + "epoch": 6.276568962416832, + "grad_norm": 4.2029461657831835, + "learning_rate": 9.197723165212772e-07, + "loss": 0.0346, + "step": 8726 + }, + { + "epoch": 6.277288257507642, + "grad_norm": 3.3430025756237005, + "learning_rate": 9.194585728048903e-07, + "loss": 0.0322, + "step": 8727 + }, + { + "epoch": 6.278007552598454, + "grad_norm": 0.14618599311946726, + "learning_rate": 9.191448589583789e-07, + "loss": 0.0005, + "step": 8728 + }, + { + "epoch": 6.278726847689264, + "grad_norm": 1.1487200177018804, + "learning_rate": 9.188311749978836e-07, + "loss": 0.0139, + "step": 8729 + }, + { + "epoch": 6.279446142780076, + "grad_norm": 0.7664608993701781, + "learning_rate": 9.185175209395451e-07, + "loss": 0.0013, + "step": 8730 + }, + { + "epoch": 6.280165437870886, + "grad_norm": 1.461384523219639, + "learning_rate": 9.182038967995011e-07, + "loss": 0.0166, + "step": 8731 + }, + { + "epoch": 6.280884732961698, + "grad_norm": 2.6701705332408947, + "learning_rate": 9.178903025938877e-07, + "loss": 0.068, + "step": 8732 + }, + { + "epoch": 6.2816040280525085, + "grad_norm": 2.3652179857555593, + "learning_rate": 9.175767383388404e-07, + "loss": 0.0276, + "step": 8733 + }, + { + "epoch": 6.28232332314332, + "grad_norm": 2.5547946906635106, + "learning_rate": 9.172632040504916e-07, + "loss": 0.0299, + "step": 8734 + }, + { + "epoch": 6.283042618234131, + "grad_norm": 0.971767935889207, + "learning_rate": 9.169496997449743e-07, + "loss": 0.0103, + "step": 8735 + }, + { + "epoch": 6.283761913324941, + "grad_norm": 2.4535321252237394, + "learning_rate": 9.166362254384186e-07, + "loss": 0.0252, + "step": 8736 + }, + { + "epoch": 6.284481208415753, + "grad_norm": 0.12153868562968924, + "learning_rate": 9.163227811469531e-07, + "loss": 0.0003, + "step": 8737 + }, + { + "epoch": 6.285200503506563, + "grad_norm": 0.19565249670892643, + "learning_rate": 9.16009366886705e-07, + "loss": 0.0005, + "step": 8738 + }, + { + "epoch": 6.285919798597375, + "grad_norm": 0.049246630147249365, + "learning_rate": 9.156959826738004e-07, + "loss": 0.0002, + "step": 8739 + }, + { + "epoch": 6.286639093688185, + "grad_norm": 1.290126063076784, + "learning_rate": 9.153826285243627e-07, + "loss": 0.0123, + "step": 8740 + }, + { + "epoch": 6.287358388778997, + "grad_norm": 3.879669715179887, + "learning_rate": 9.150693044545154e-07, + "loss": 0.0437, + "step": 8741 + }, + { + "epoch": 6.288077683869807, + "grad_norm": 0.10877990476983497, + "learning_rate": 9.147560104803793e-07, + "loss": 0.0003, + "step": 8742 + }, + { + "epoch": 6.288796978960619, + "grad_norm": 1.7426702287521463, + "learning_rate": 9.144427466180736e-07, + "loss": 0.0161, + "step": 8743 + }, + { + "epoch": 6.289516274051429, + "grad_norm": 2.612538177929658, + "learning_rate": 9.141295128837166e-07, + "loss": 0.0485, + "step": 8744 + }, + { + "epoch": 6.290235569142241, + "grad_norm": 3.1158781661974913, + "learning_rate": 9.138163092934241e-07, + "loss": 0.0254, + "step": 8745 + }, + { + "epoch": 6.2909548642330515, + "grad_norm": 3.5932262824903027, + "learning_rate": 9.135031358633123e-07, + "loss": 0.0121, + "step": 8746 + }, + { + "epoch": 6.291674159323863, + "grad_norm": 2.5137501386647165, + "learning_rate": 9.131899926094935e-07, + "loss": 0.0253, + "step": 8747 + }, + { + "epoch": 6.292393454414674, + "grad_norm": 1.7585048664543061, + "learning_rate": 9.128768795480797e-07, + "loss": 0.0237, + "step": 8748 + }, + { + "epoch": 6.293112749505484, + "grad_norm": 0.11477182102499324, + "learning_rate": 9.125637966951811e-07, + "loss": 0.0004, + "step": 8749 + }, + { + "epoch": 6.293832044596296, + "grad_norm": 1.2978801026224418, + "learning_rate": 9.122507440669055e-07, + "loss": 0.0192, + "step": 8750 + }, + { + "epoch": 6.294551339687106, + "grad_norm": 0.26129388542692084, + "learning_rate": 9.11937721679362e-07, + "loss": 0.0008, + "step": 8751 + }, + { + "epoch": 6.295270634777918, + "grad_norm": 0.19505155105379385, + "learning_rate": 9.116247295486547e-07, + "loss": 0.0003, + "step": 8752 + }, + { + "epoch": 6.295989929868728, + "grad_norm": 2.116377307362769, + "learning_rate": 9.11311767690888e-07, + "loss": 0.0316, + "step": 8753 + }, + { + "epoch": 6.29670922495954, + "grad_norm": 0.2717234748102172, + "learning_rate": 9.109988361221641e-07, + "loss": 0.0016, + "step": 8754 + }, + { + "epoch": 6.29742852005035, + "grad_norm": 1.2958226414529195, + "learning_rate": 9.10685934858584e-07, + "loss": 0.0077, + "step": 8755 + }, + { + "epoch": 6.298147815141162, + "grad_norm": 0.09189981156584043, + "learning_rate": 9.103730639162472e-07, + "loss": 0.0002, + "step": 8756 + }, + { + "epoch": 6.298867110231972, + "grad_norm": 0.3201681231797727, + "learning_rate": 9.100602233112513e-07, + "loss": 0.0014, + "step": 8757 + }, + { + "epoch": 6.299586405322784, + "grad_norm": 2.2124117399770493, + "learning_rate": 9.097474130596926e-07, + "loss": 0.0337, + "step": 8758 + }, + { + "epoch": 6.3003057004135945, + "grad_norm": 4.5296142530604655, + "learning_rate": 9.094346331776655e-07, + "loss": 0.0486, + "step": 8759 + }, + { + "epoch": 6.301024995504406, + "grad_norm": 5.071110294982579, + "learning_rate": 9.091218836812633e-07, + "loss": 0.0283, + "step": 8760 + }, + { + "epoch": 6.301744290595217, + "grad_norm": 2.6476865552745963, + "learning_rate": 9.088091645865773e-07, + "loss": 0.0483, + "step": 8761 + }, + { + "epoch": 6.302463585686028, + "grad_norm": 3.197921784289651, + "learning_rate": 9.084964759096978e-07, + "loss": 0.0473, + "step": 8762 + }, + { + "epoch": 6.303182880776839, + "grad_norm": 0.8292633183296441, + "learning_rate": 9.081838176667129e-07, + "loss": 0.0042, + "step": 8763 + }, + { + "epoch": 6.30390217586765, + "grad_norm": 0.1444404852926021, + "learning_rate": 9.078711898737096e-07, + "loss": 0.0004, + "step": 8764 + }, + { + "epoch": 6.304621470958461, + "grad_norm": 1.9287419849183722, + "learning_rate": 9.075585925467729e-07, + "loss": 0.0266, + "step": 8765 + }, + { + "epoch": 6.305340766049271, + "grad_norm": 0.4127492635276675, + "learning_rate": 9.07246025701986e-07, + "loss": 0.0007, + "step": 8766 + }, + { + "epoch": 6.306060061140083, + "grad_norm": 0.5662643939942364, + "learning_rate": 9.069334893554325e-07, + "loss": 0.0021, + "step": 8767 + }, + { + "epoch": 6.306779356230893, + "grad_norm": 1.2597481267276538, + "learning_rate": 9.066209835231917e-07, + "loss": 0.0157, + "step": 8768 + }, + { + "epoch": 6.307498651321705, + "grad_norm": 3.695239463340809, + "learning_rate": 9.063085082213436e-07, + "loss": 0.0199, + "step": 8769 + }, + { + "epoch": 6.308217946412515, + "grad_norm": 0.1268624108913602, + "learning_rate": 9.059960634659645e-07, + "loss": 0.0004, + "step": 8770 + }, + { + "epoch": 6.308937241503327, + "grad_norm": 2.3103427447959555, + "learning_rate": 9.056836492731297e-07, + "loss": 0.0289, + "step": 8771 + }, + { + "epoch": 6.3096565365941375, + "grad_norm": 4.425532371190145, + "learning_rate": 9.053712656589155e-07, + "loss": 0.0369, + "step": 8772 + }, + { + "epoch": 6.310375831684949, + "grad_norm": 2.1569469858447134, + "learning_rate": 9.050589126393935e-07, + "loss": 0.0267, + "step": 8773 + }, + { + "epoch": 6.31109512677576, + "grad_norm": 1.975467014051648, + "learning_rate": 9.047465902306346e-07, + "loss": 0.0262, + "step": 8774 + }, + { + "epoch": 6.311814421866571, + "grad_norm": 3.291578040704041, + "learning_rate": 9.044342984487087e-07, + "loss": 0.0348, + "step": 8775 + }, + { + "epoch": 6.312533716957382, + "grad_norm": 0.07498922851235285, + "learning_rate": 9.041220373096832e-07, + "loss": 0.0002, + "step": 8776 + }, + { + "epoch": 6.313253012048193, + "grad_norm": 2.0155285675200063, + "learning_rate": 9.038098068296254e-07, + "loss": 0.0114, + "step": 8777 + }, + { + "epoch": 6.313972307139004, + "grad_norm": 4.063254184446011, + "learning_rate": 9.034976070245996e-07, + "loss": 0.0288, + "step": 8778 + }, + { + "epoch": 6.314691602229815, + "grad_norm": 1.4151865250333857, + "learning_rate": 9.031854379106687e-07, + "loss": 0.0206, + "step": 8779 + }, + { + "epoch": 6.315410897320626, + "grad_norm": 2.950656271223171, + "learning_rate": 9.028732995038951e-07, + "loss": 0.0228, + "step": 8780 + }, + { + "epoch": 6.316130192411437, + "grad_norm": 5.856603601309274, + "learning_rate": 9.025611918203377e-07, + "loss": 0.0633, + "step": 8781 + }, + { + "epoch": 6.316849487502248, + "grad_norm": 4.065557460275283, + "learning_rate": 9.022491148760567e-07, + "loss": 0.0588, + "step": 8782 + }, + { + "epoch": 6.317568782593058, + "grad_norm": 0.050284582894875786, + "learning_rate": 9.019370686871078e-07, + "loss": 0.0001, + "step": 8783 + }, + { + "epoch": 6.31828807768387, + "grad_norm": 1.8874841974761216, + "learning_rate": 9.016250532695465e-07, + "loss": 0.0144, + "step": 8784 + }, + { + "epoch": 6.3190073727746805, + "grad_norm": 0.008129923242212744, + "learning_rate": 9.013130686394267e-07, + "loss": 0.0, + "step": 8785 + }, + { + "epoch": 6.319726667865492, + "grad_norm": 0.5347973111582662, + "learning_rate": 9.010011148128005e-07, + "loss": 0.0011, + "step": 8786 + }, + { + "epoch": 6.320445962956303, + "grad_norm": 3.166446271021582, + "learning_rate": 9.006891918057177e-07, + "loss": 0.0641, + "step": 8787 + }, + { + "epoch": 6.321165258047114, + "grad_norm": 2.702983419006049, + "learning_rate": 9.003772996342286e-07, + "loss": 0.0295, + "step": 8788 + }, + { + "epoch": 6.321884553137925, + "grad_norm": 1.3508262469141437, + "learning_rate": 9.000654383143804e-07, + "loss": 0.0093, + "step": 8789 + }, + { + "epoch": 6.322603848228736, + "grad_norm": 1.3768150549272304, + "learning_rate": 8.997536078622181e-07, + "loss": 0.0085, + "step": 8790 + }, + { + "epoch": 6.323323143319547, + "grad_norm": 2.1280338501312817, + "learning_rate": 8.994418082937867e-07, + "loss": 0.0182, + "step": 8791 + }, + { + "epoch": 6.324042438410358, + "grad_norm": 0.8645798674552372, + "learning_rate": 8.99130039625128e-07, + "loss": 0.0097, + "step": 8792 + }, + { + "epoch": 6.324761733501169, + "grad_norm": 0.11323485768757152, + "learning_rate": 8.988183018722839e-07, + "loss": 0.0006, + "step": 8793 + }, + { + "epoch": 6.32548102859198, + "grad_norm": 1.7825532300579168, + "learning_rate": 8.985065950512937e-07, + "loss": 0.027, + "step": 8794 + }, + { + "epoch": 6.326200323682791, + "grad_norm": 0.008779503937837047, + "learning_rate": 8.981949191781949e-07, + "loss": 0.0, + "step": 8795 + }, + { + "epoch": 6.326919618773601, + "grad_norm": 6.878053356241148, + "learning_rate": 8.97883274269024e-07, + "loss": 0.0692, + "step": 8796 + }, + { + "epoch": 6.327638913864413, + "grad_norm": 2.456345612991258, + "learning_rate": 8.975716603398152e-07, + "loss": 0.0052, + "step": 8797 + }, + { + "epoch": 6.3283582089552235, + "grad_norm": 2.478860721160586, + "learning_rate": 8.972600774066025e-07, + "loss": 0.0082, + "step": 8798 + }, + { + "epoch": 6.329077504046035, + "grad_norm": 0.038685697918200104, + "learning_rate": 8.969485254854167e-07, + "loss": 0.0001, + "step": 8799 + }, + { + "epoch": 6.329796799136846, + "grad_norm": 1.6493438825410143, + "learning_rate": 8.966370045922881e-07, + "loss": 0.0135, + "step": 8800 + }, + { + "epoch": 6.330516094227657, + "grad_norm": 1.763836087122746, + "learning_rate": 8.963255147432447e-07, + "loss": 0.0299, + "step": 8801 + }, + { + "epoch": 6.331235389318468, + "grad_norm": 0.8146666517376647, + "learning_rate": 8.960140559543126e-07, + "loss": 0.0026, + "step": 8802 + }, + { + "epoch": 6.331954684409279, + "grad_norm": 3.1036159876920926, + "learning_rate": 8.957026282415182e-07, + "loss": 0.0417, + "step": 8803 + }, + { + "epoch": 6.33267397950009, + "grad_norm": 0.020271802628100032, + "learning_rate": 8.953912316208846e-07, + "loss": 0.0001, + "step": 8804 + }, + { + "epoch": 6.333393274590901, + "grad_norm": 1.115716060615625, + "learning_rate": 8.950798661084334e-07, + "loss": 0.0058, + "step": 8805 + }, + { + "epoch": 6.334112569681712, + "grad_norm": 0.02539032997978993, + "learning_rate": 8.94768531720185e-07, + "loss": 0.0001, + "step": 8806 + }, + { + "epoch": 6.334831864772523, + "grad_norm": 0.5898569390590837, + "learning_rate": 8.944572284721579e-07, + "loss": 0.0056, + "step": 8807 + }, + { + "epoch": 6.335551159863334, + "grad_norm": 2.3114917840108267, + "learning_rate": 8.94145956380369e-07, + "loss": 0.0262, + "step": 8808 + }, + { + "epoch": 6.336270454954145, + "grad_norm": 4.032246034811912, + "learning_rate": 8.938347154608345e-07, + "loss": 0.0288, + "step": 8809 + }, + { + "epoch": 6.336989750044956, + "grad_norm": 4.290765073900556, + "learning_rate": 8.935235057295684e-07, + "loss": 0.0219, + "step": 8810 + }, + { + "epoch": 6.337709045135767, + "grad_norm": 3.490300588549557, + "learning_rate": 8.932123272025822e-07, + "loss": 0.0289, + "step": 8811 + }, + { + "epoch": 6.338428340226578, + "grad_norm": 1.4104415814844102, + "learning_rate": 8.92901179895887e-07, + "loss": 0.0058, + "step": 8812 + }, + { + "epoch": 6.339147635317389, + "grad_norm": 3.991976581944841, + "learning_rate": 8.925900638254918e-07, + "loss": 0.0249, + "step": 8813 + }, + { + "epoch": 6.3398669304082, + "grad_norm": 3.114185748372296, + "learning_rate": 8.922789790074041e-07, + "loss": 0.0161, + "step": 8814 + }, + { + "epoch": 6.340586225499011, + "grad_norm": 2.1030677576408996, + "learning_rate": 8.919679254576298e-07, + "loss": 0.0128, + "step": 8815 + }, + { + "epoch": 6.341305520589822, + "grad_norm": 2.376599621009517, + "learning_rate": 8.916569031921732e-07, + "loss": 0.0173, + "step": 8816 + }, + { + "epoch": 6.342024815680633, + "grad_norm": 0.9112607407226676, + "learning_rate": 8.913459122270366e-07, + "loss": 0.0078, + "step": 8817 + }, + { + "epoch": 6.342744110771444, + "grad_norm": 3.095106322354701, + "learning_rate": 8.910349525782212e-07, + "loss": 0.0258, + "step": 8818 + }, + { + "epoch": 6.343463405862255, + "grad_norm": 0.01743533046920908, + "learning_rate": 8.907240242617268e-07, + "loss": 0.0, + "step": 8819 + }, + { + "epoch": 6.344182700953066, + "grad_norm": 3.5979268055251774, + "learning_rate": 8.904131272935509e-07, + "loss": 0.0191, + "step": 8820 + }, + { + "epoch": 6.344901996043877, + "grad_norm": 1.0287290648707992, + "learning_rate": 8.901022616896896e-07, + "loss": 0.01, + "step": 8821 + }, + { + "epoch": 6.345621291134688, + "grad_norm": 2.6215241387000634, + "learning_rate": 8.897914274661377e-07, + "loss": 0.0306, + "step": 8822 + }, + { + "epoch": 6.346340586225499, + "grad_norm": 4.297397455976695, + "learning_rate": 8.894806246388873e-07, + "loss": 0.0444, + "step": 8823 + }, + { + "epoch": 6.34705988131631, + "grad_norm": 0.173835624677954, + "learning_rate": 8.891698532239312e-07, + "loss": 0.0006, + "step": 8824 + }, + { + "epoch": 6.347779176407121, + "grad_norm": 0.4804174832094122, + "learning_rate": 8.888591132372584e-07, + "loss": 0.0012, + "step": 8825 + }, + { + "epoch": 6.348498471497932, + "grad_norm": 1.5346301160358389, + "learning_rate": 8.885484046948571e-07, + "loss": 0.0066, + "step": 8826 + }, + { + "epoch": 6.349217766588743, + "grad_norm": 1.4340519466411996, + "learning_rate": 8.882377276127138e-07, + "loss": 0.0143, + "step": 8827 + }, + { + "epoch": 6.349937061679554, + "grad_norm": 3.7256782392397665, + "learning_rate": 8.879270820068133e-07, + "loss": 0.0269, + "step": 8828 + }, + { + "epoch": 6.350656356770365, + "grad_norm": 4.4359533479588205, + "learning_rate": 8.876164678931383e-07, + "loss": 0.0351, + "step": 8829 + }, + { + "epoch": 6.351375651861176, + "grad_norm": 2.2523592272802824, + "learning_rate": 8.873058852876715e-07, + "loss": 0.0197, + "step": 8830 + }, + { + "epoch": 6.352094946951987, + "grad_norm": 0.007838564905994449, + "learning_rate": 8.869953342063924e-07, + "loss": 0.0, + "step": 8831 + }, + { + "epoch": 6.352814242042798, + "grad_norm": 0.4794383640665072, + "learning_rate": 8.866848146652796e-07, + "loss": 0.0035, + "step": 8832 + }, + { + "epoch": 6.353533537133609, + "grad_norm": 3.4903916804271624, + "learning_rate": 8.863743266803098e-07, + "loss": 0.0584, + "step": 8833 + }, + { + "epoch": 6.35425283222442, + "grad_norm": 1.5636055016124075, + "learning_rate": 8.860638702674576e-07, + "loss": 0.0283, + "step": 8834 + }, + { + "epoch": 6.354972127315231, + "grad_norm": 0.4519207731997015, + "learning_rate": 8.857534454426972e-07, + "loss": 0.002, + "step": 8835 + }, + { + "epoch": 6.355691422406042, + "grad_norm": 2.922634831338771, + "learning_rate": 8.854430522220005e-07, + "loss": 0.0536, + "step": 8836 + }, + { + "epoch": 6.356410717496853, + "grad_norm": 0.5387630513943987, + "learning_rate": 8.851326906213377e-07, + "loss": 0.0012, + "step": 8837 + }, + { + "epoch": 6.357130012587664, + "grad_norm": 3.84264454516704, + "learning_rate": 8.848223606566771e-07, + "loss": 0.0566, + "step": 8838 + }, + { + "epoch": 6.3578493076784754, + "grad_norm": 3.488080811236963, + "learning_rate": 8.845120623439855e-07, + "loss": 0.0655, + "step": 8839 + }, + { + "epoch": 6.358568602769286, + "grad_norm": 3.289398712900295, + "learning_rate": 8.842017956992292e-07, + "loss": 0.0633, + "step": 8840 + }, + { + "epoch": 6.3592878978600975, + "grad_norm": 4.731424308874869, + "learning_rate": 8.838915607383719e-07, + "loss": 0.0461, + "step": 8841 + }, + { + "epoch": 6.360007192950908, + "grad_norm": 3.2709335125531855, + "learning_rate": 8.83581357477375e-07, + "loss": 0.0375, + "step": 8842 + }, + { + "epoch": 6.360726488041719, + "grad_norm": 2.716096653903255, + "learning_rate": 8.832711859321994e-07, + "loss": 0.029, + "step": 8843 + }, + { + "epoch": 6.36144578313253, + "grad_norm": 5.627160943830431, + "learning_rate": 8.829610461188034e-07, + "loss": 0.1463, + "step": 8844 + }, + { + "epoch": 6.362165078223341, + "grad_norm": 4.596739695697162, + "learning_rate": 8.826509380531452e-07, + "loss": 0.0435, + "step": 8845 + }, + { + "epoch": 6.362884373314152, + "grad_norm": 3.6889412548726526, + "learning_rate": 8.823408617511802e-07, + "loss": 0.0438, + "step": 8846 + }, + { + "epoch": 6.363603668404963, + "grad_norm": 1.498435967016737, + "learning_rate": 8.820308172288617e-07, + "loss": 0.0077, + "step": 8847 + }, + { + "epoch": 6.364322963495774, + "grad_norm": 2.5571070418335817, + "learning_rate": 8.817208045021429e-07, + "loss": 0.02, + "step": 8848 + }, + { + "epoch": 6.365042258586585, + "grad_norm": 7.825169449372754, + "learning_rate": 8.814108235869737e-07, + "loss": 0.0317, + "step": 8849 + }, + { + "epoch": 6.365761553677396, + "grad_norm": 1.4546594683298426, + "learning_rate": 8.811008744993035e-07, + "loss": 0.0208, + "step": 8850 + }, + { + "epoch": 6.366480848768207, + "grad_norm": 2.246375998978973, + "learning_rate": 8.807909572550799e-07, + "loss": 0.0234, + "step": 8851 + }, + { + "epoch": 6.3672001438590184, + "grad_norm": 2.29033121129588, + "learning_rate": 8.804810718702486e-07, + "loss": 0.0109, + "step": 8852 + }, + { + "epoch": 6.367919438949829, + "grad_norm": 1.572150930078236, + "learning_rate": 8.801712183607537e-07, + "loss": 0.0078, + "step": 8853 + }, + { + "epoch": 6.3686387340406405, + "grad_norm": 0.2174196865072084, + "learning_rate": 8.798613967425375e-07, + "loss": 0.0008, + "step": 8854 + }, + { + "epoch": 6.369358029131451, + "grad_norm": 0.2531615524328989, + "learning_rate": 8.795516070315409e-07, + "loss": 0.0011, + "step": 8855 + }, + { + "epoch": 6.370077324222263, + "grad_norm": 1.0266172538592795, + "learning_rate": 8.792418492437037e-07, + "loss": 0.0043, + "step": 8856 + }, + { + "epoch": 6.370796619313073, + "grad_norm": 0.8515767624101608, + "learning_rate": 8.789321233949629e-07, + "loss": 0.0057, + "step": 8857 + }, + { + "epoch": 6.371515914403885, + "grad_norm": 2.3318023342016883, + "learning_rate": 8.786224295012548e-07, + "loss": 0.0265, + "step": 8858 + }, + { + "epoch": 6.372235209494695, + "grad_norm": 1.3713485552736278, + "learning_rate": 8.783127675785133e-07, + "loss": 0.0062, + "step": 8859 + }, + { + "epoch": 6.372954504585506, + "grad_norm": 2.210153811029482, + "learning_rate": 8.780031376426706e-07, + "loss": 0.0203, + "step": 8860 + }, + { + "epoch": 6.373673799676317, + "grad_norm": 0.190430397105261, + "learning_rate": 8.776935397096593e-07, + "loss": 0.0003, + "step": 8861 + }, + { + "epoch": 6.374393094767128, + "grad_norm": 0.9593889394214729, + "learning_rate": 8.773839737954074e-07, + "loss": 0.0093, + "step": 8862 + }, + { + "epoch": 6.375112389857939, + "grad_norm": 3.462511540942245, + "learning_rate": 8.770744399158431e-07, + "loss": 0.0365, + "step": 8863 + }, + { + "epoch": 6.37583168494875, + "grad_norm": 1.4806128082395744, + "learning_rate": 8.767649380868929e-07, + "loss": 0.0108, + "step": 8864 + }, + { + "epoch": 6.3765509800395614, + "grad_norm": 0.995440494285847, + "learning_rate": 8.764554683244796e-07, + "loss": 0.009, + "step": 8865 + }, + { + "epoch": 6.377270275130372, + "grad_norm": 1.6965995503087783, + "learning_rate": 8.761460306445277e-07, + "loss": 0.0131, + "step": 8866 + }, + { + "epoch": 6.3779895702211835, + "grad_norm": 3.7869682524510013, + "learning_rate": 8.758366250629575e-07, + "loss": 0.0247, + "step": 8867 + }, + { + "epoch": 6.378708865311994, + "grad_norm": 2.729912380123919, + "learning_rate": 8.755272515956889e-07, + "loss": 0.0325, + "step": 8868 + }, + { + "epoch": 6.379428160402806, + "grad_norm": 0.21290788072809544, + "learning_rate": 8.752179102586394e-07, + "loss": 0.0006, + "step": 8869 + }, + { + "epoch": 6.380147455493616, + "grad_norm": 1.9606985281134635, + "learning_rate": 8.74908601067725e-07, + "loss": 0.0222, + "step": 8870 + }, + { + "epoch": 6.380866750584428, + "grad_norm": 3.0713466653704016, + "learning_rate": 8.745993240388599e-07, + "loss": 0.0229, + "step": 8871 + }, + { + "epoch": 6.381586045675238, + "grad_norm": 4.7201237136630345, + "learning_rate": 8.742900791879581e-07, + "loss": 0.0417, + "step": 8872 + }, + { + "epoch": 6.382305340766049, + "grad_norm": 0.8787227302639952, + "learning_rate": 8.7398086653093e-07, + "loss": 0.0063, + "step": 8873 + }, + { + "epoch": 6.38302463585686, + "grad_norm": 0.03504534456731374, + "learning_rate": 8.736716860836852e-07, + "loss": 0.0001, + "step": 8874 + }, + { + "epoch": 6.383743930947671, + "grad_norm": 1.49911061222743, + "learning_rate": 8.733625378621324e-07, + "loss": 0.0069, + "step": 8875 + }, + { + "epoch": 6.384463226038482, + "grad_norm": 2.248592953308147, + "learning_rate": 8.730534218821759e-07, + "loss": 0.0256, + "step": 8876 + }, + { + "epoch": 6.385182521129293, + "grad_norm": 1.9687433666978895, + "learning_rate": 8.727443381597219e-07, + "loss": 0.0078, + "step": 8877 + }, + { + "epoch": 6.385901816220104, + "grad_norm": 2.1684186173494755, + "learning_rate": 8.724352867106728e-07, + "loss": 0.0067, + "step": 8878 + }, + { + "epoch": 6.386621111310915, + "grad_norm": 3.0439064520304506, + "learning_rate": 8.721262675509299e-07, + "loss": 0.0458, + "step": 8879 + }, + { + "epoch": 6.3873404064017265, + "grad_norm": 4.439455582763364, + "learning_rate": 8.718172806963928e-07, + "loss": 0.0516, + "step": 8880 + }, + { + "epoch": 6.388059701492537, + "grad_norm": 1.6028908941542037, + "learning_rate": 8.715083261629585e-07, + "loss": 0.018, + "step": 8881 + }, + { + "epoch": 6.388778996583349, + "grad_norm": 0.01760652859041619, + "learning_rate": 8.711994039665249e-07, + "loss": 0.0001, + "step": 8882 + }, + { + "epoch": 6.389498291674159, + "grad_norm": 2.5321592180056713, + "learning_rate": 8.708905141229858e-07, + "loss": 0.0266, + "step": 8883 + }, + { + "epoch": 6.390217586764971, + "grad_norm": 1.8394404164783509, + "learning_rate": 8.70581656648234e-07, + "loss": 0.0183, + "step": 8884 + }, + { + "epoch": 6.390936881855781, + "grad_norm": 0.37974566603703425, + "learning_rate": 8.70272831558161e-07, + "loss": 0.0006, + "step": 8885 + }, + { + "epoch": 6.391656176946593, + "grad_norm": 0.8260223131562302, + "learning_rate": 8.699640388686557e-07, + "loss": 0.0016, + "step": 8886 + }, + { + "epoch": 6.392375472037403, + "grad_norm": 2.7910516597135446, + "learning_rate": 8.696552785956071e-07, + "loss": 0.0214, + "step": 8887 + }, + { + "epoch": 6.393094767128215, + "grad_norm": 5.879446551234585, + "learning_rate": 8.693465507549015e-07, + "loss": 0.0789, + "step": 8888 + }, + { + "epoch": 6.393814062219025, + "grad_norm": 2.080797087037036, + "learning_rate": 8.690378553624223e-07, + "loss": 0.037, + "step": 8889 + }, + { + "epoch": 6.394533357309836, + "grad_norm": 2.334901431422955, + "learning_rate": 8.687291924340533e-07, + "loss": 0.028, + "step": 8890 + }, + { + "epoch": 6.395252652400647, + "grad_norm": 1.226331921262495, + "learning_rate": 8.684205619856755e-07, + "loss": 0.0042, + "step": 8891 + }, + { + "epoch": 6.395971947491458, + "grad_norm": 0.8429090812870623, + "learning_rate": 8.681119640331675e-07, + "loss": 0.0063, + "step": 8892 + }, + { + "epoch": 6.3966912425822695, + "grad_norm": 1.4364596334759265, + "learning_rate": 8.678033985924093e-07, + "loss": 0.013, + "step": 8893 + }, + { + "epoch": 6.39741053767308, + "grad_norm": 6.232695347549032, + "learning_rate": 8.674948656792755e-07, + "loss": 0.0667, + "step": 8894 + }, + { + "epoch": 6.398129832763892, + "grad_norm": 3.686855526729972, + "learning_rate": 8.671863653096417e-07, + "loss": 0.0295, + "step": 8895 + }, + { + "epoch": 6.398849127854702, + "grad_norm": 0.9857648780043771, + "learning_rate": 8.668778974993799e-07, + "loss": 0.0035, + "step": 8896 + }, + { + "epoch": 6.399568422945514, + "grad_norm": 3.044519582244946, + "learning_rate": 8.665694622643613e-07, + "loss": 0.0206, + "step": 8897 + }, + { + "epoch": 6.400287718036324, + "grad_norm": 1.4080140983171567, + "learning_rate": 8.662610596204564e-07, + "loss": 0.0095, + "step": 8898 + }, + { + "epoch": 6.401007013127136, + "grad_norm": 2.531498124051744, + "learning_rate": 8.659526895835323e-07, + "loss": 0.0098, + "step": 8899 + }, + { + "epoch": 6.401726308217946, + "grad_norm": 0.17978323441379584, + "learning_rate": 8.656443521694554e-07, + "loss": 0.0005, + "step": 8900 + }, + { + "epoch": 6.402445603308758, + "grad_norm": 0.6098183841820773, + "learning_rate": 8.653360473940903e-07, + "loss": 0.0018, + "step": 8901 + }, + { + "epoch": 6.403164898399568, + "grad_norm": 0.9373041585039072, + "learning_rate": 8.650277752732997e-07, + "loss": 0.0011, + "step": 8902 + }, + { + "epoch": 6.403884193490379, + "grad_norm": 2.4943998059225643, + "learning_rate": 8.647195358229446e-07, + "loss": 0.0127, + "step": 8903 + }, + { + "epoch": 6.40460348858119, + "grad_norm": 1.5248194075380261, + "learning_rate": 8.644113290588846e-07, + "loss": 0.0064, + "step": 8904 + }, + { + "epoch": 6.405322783672001, + "grad_norm": 0.7705211354598958, + "learning_rate": 8.641031549969775e-07, + "loss": 0.0071, + "step": 8905 + }, + { + "epoch": 6.4060420787628125, + "grad_norm": 2.5498966133348437, + "learning_rate": 8.637950136530792e-07, + "loss": 0.0256, + "step": 8906 + }, + { + "epoch": 6.406761373853623, + "grad_norm": 2.835275565558995, + "learning_rate": 8.634869050430439e-07, + "loss": 0.0381, + "step": 8907 + }, + { + "epoch": 6.407480668944435, + "grad_norm": 4.472139736247631, + "learning_rate": 8.631788291827252e-07, + "loss": 0.0418, + "step": 8908 + }, + { + "epoch": 6.408199964035245, + "grad_norm": 0.23513161265314242, + "learning_rate": 8.628707860879737e-07, + "loss": 0.0015, + "step": 8909 + }, + { + "epoch": 6.408919259126057, + "grad_norm": 1.4435590847343849, + "learning_rate": 8.625627757746385e-07, + "loss": 0.0032, + "step": 8910 + }, + { + "epoch": 6.409638554216867, + "grad_norm": 4.709766854758723, + "learning_rate": 8.622547982585678e-07, + "loss": 0.0663, + "step": 8911 + }, + { + "epoch": 6.410357849307679, + "grad_norm": 0.23802264132063838, + "learning_rate": 8.61946853555607e-07, + "loss": 0.0011, + "step": 8912 + }, + { + "epoch": 6.411077144398489, + "grad_norm": 4.7628500897710415, + "learning_rate": 8.616389416816001e-07, + "loss": 0.0465, + "step": 8913 + }, + { + "epoch": 6.411796439489301, + "grad_norm": 8.417343436991485, + "learning_rate": 8.613310626523911e-07, + "loss": 0.1095, + "step": 8914 + }, + { + "epoch": 6.412515734580111, + "grad_norm": 8.13064189030512, + "learning_rate": 8.610232164838197e-07, + "loss": 0.1145, + "step": 8915 + }, + { + "epoch": 6.413235029670923, + "grad_norm": 0.019346909599929647, + "learning_rate": 8.607154031917256e-07, + "loss": 0.0001, + "step": 8916 + }, + { + "epoch": 6.413954324761733, + "grad_norm": 7.593836282403827, + "learning_rate": 8.604076227919463e-07, + "loss": 0.0655, + "step": 8917 + }, + { + "epoch": 6.414673619852545, + "grad_norm": 0.8685402252421303, + "learning_rate": 8.600998753003177e-07, + "loss": 0.0055, + "step": 8918 + }, + { + "epoch": 6.4153929149433555, + "grad_norm": 0.09141102490529088, + "learning_rate": 8.597921607326736e-07, + "loss": 0.0001, + "step": 8919 + }, + { + "epoch": 6.416112210034166, + "grad_norm": 3.6201695245838748, + "learning_rate": 8.594844791048468e-07, + "loss": 0.0368, + "step": 8920 + }, + { + "epoch": 6.416831505124978, + "grad_norm": 2.649732393400778, + "learning_rate": 8.591768304326679e-07, + "loss": 0.0224, + "step": 8921 + }, + { + "epoch": 6.417550800215788, + "grad_norm": 0.12626737286585576, + "learning_rate": 8.588692147319659e-07, + "loss": 0.0007, + "step": 8922 + }, + { + "epoch": 6.4182700953066, + "grad_norm": 3.88031542451791, + "learning_rate": 8.585616320185679e-07, + "loss": 0.0372, + "step": 8923 + }, + { + "epoch": 6.41898939039741, + "grad_norm": 5.355883495795779, + "learning_rate": 8.582540823083005e-07, + "loss": 0.1083, + "step": 8924 + }, + { + "epoch": 6.419708685488222, + "grad_norm": 1.5167923156903176, + "learning_rate": 8.57946565616987e-07, + "loss": 0.0189, + "step": 8925 + }, + { + "epoch": 6.420427980579032, + "grad_norm": 6.070602330450886, + "learning_rate": 8.576390819604498e-07, + "loss": 0.0828, + "step": 8926 + }, + { + "epoch": 6.421147275669844, + "grad_norm": 3.3898833691698407, + "learning_rate": 8.573316313545095e-07, + "loss": 0.0408, + "step": 8927 + }, + { + "epoch": 6.421866570760654, + "grad_norm": 2.177045477983283, + "learning_rate": 8.570242138149843e-07, + "loss": 0.0237, + "step": 8928 + }, + { + "epoch": 6.422585865851466, + "grad_norm": 7.186816670352962, + "learning_rate": 8.567168293576926e-07, + "loss": 0.0744, + "step": 8929 + }, + { + "epoch": 6.423305160942276, + "grad_norm": 2.8925623453478315, + "learning_rate": 8.564094779984493e-07, + "loss": 0.0352, + "step": 8930 + }, + { + "epoch": 6.424024456033088, + "grad_norm": 0.09386189855906947, + "learning_rate": 8.561021597530683e-07, + "loss": 0.0002, + "step": 8931 + }, + { + "epoch": 6.4247437511238985, + "grad_norm": 3.9830346415053848, + "learning_rate": 8.557948746373614e-07, + "loss": 0.072, + "step": 8932 + }, + { + "epoch": 6.42546304621471, + "grad_norm": 1.7709977278296958, + "learning_rate": 8.554876226671391e-07, + "loss": 0.005, + "step": 8933 + }, + { + "epoch": 6.426182341305521, + "grad_norm": 1.0464823655223152, + "learning_rate": 8.551804038582098e-07, + "loss": 0.005, + "step": 8934 + }, + { + "epoch": 6.426901636396332, + "grad_norm": 3.781554948955761, + "learning_rate": 8.548732182263808e-07, + "loss": 0.0508, + "step": 8935 + }, + { + "epoch": 6.427620931487143, + "grad_norm": 0.32770563140194175, + "learning_rate": 8.545660657874572e-07, + "loss": 0.0004, + "step": 8936 + }, + { + "epoch": 6.428340226577953, + "grad_norm": 2.580662734572084, + "learning_rate": 8.542589465572426e-07, + "loss": 0.0442, + "step": 8937 + }, + { + "epoch": 6.429059521668765, + "grad_norm": 0.0014393917454963477, + "learning_rate": 8.539518605515387e-07, + "loss": 0.0, + "step": 8938 + }, + { + "epoch": 6.429778816759575, + "grad_norm": 1.7412538449246573, + "learning_rate": 8.53644807786145e-07, + "loss": 0.0133, + "step": 8939 + }, + { + "epoch": 6.430498111850387, + "grad_norm": 1.8127913143464833, + "learning_rate": 8.533377882768614e-07, + "loss": 0.0189, + "step": 8940 + }, + { + "epoch": 6.431217406941197, + "grad_norm": 0.41770101156032463, + "learning_rate": 8.530308020394837e-07, + "loss": 0.0014, + "step": 8941 + }, + { + "epoch": 6.431936702032009, + "grad_norm": 0.32372971227933534, + "learning_rate": 8.52723849089807e-07, + "loss": 0.0013, + "step": 8942 + }, + { + "epoch": 6.432655997122819, + "grad_norm": 5.093854571103995, + "learning_rate": 8.524169294436244e-07, + "loss": 0.084, + "step": 8943 + }, + { + "epoch": 6.433375292213631, + "grad_norm": 0.034917045123148936, + "learning_rate": 8.52110043116727e-07, + "loss": 0.0002, + "step": 8944 + }, + { + "epoch": 6.4340945873044415, + "grad_norm": 2.211581527961614, + "learning_rate": 8.518031901249057e-07, + "loss": 0.0223, + "step": 8945 + }, + { + "epoch": 6.434813882395253, + "grad_norm": 2.310987003274305, + "learning_rate": 8.514963704839484e-07, + "loss": 0.0163, + "step": 8946 + }, + { + "epoch": 6.435533177486064, + "grad_norm": 5.827613377267531, + "learning_rate": 8.511895842096416e-07, + "loss": 0.0626, + "step": 8947 + }, + { + "epoch": 6.436252472576875, + "grad_norm": 0.43743181061151953, + "learning_rate": 8.508828313177693e-07, + "loss": 0.0019, + "step": 8948 + }, + { + "epoch": 6.436971767667686, + "grad_norm": 3.5998280082545278, + "learning_rate": 8.50576111824114e-07, + "loss": 0.0208, + "step": 8949 + }, + { + "epoch": 6.437691062758496, + "grad_norm": 1.1811332770953717, + "learning_rate": 8.502694257444585e-07, + "loss": 0.005, + "step": 8950 + }, + { + "epoch": 6.438410357849308, + "grad_norm": 5.4844134784951954, + "learning_rate": 8.499627730945817e-07, + "loss": 0.0446, + "step": 8951 + }, + { + "epoch": 6.439129652940118, + "grad_norm": 0.30102313002785663, + "learning_rate": 8.49656153890261e-07, + "loss": 0.0015, + "step": 8952 + }, + { + "epoch": 6.43984894803093, + "grad_norm": 0.11316688001374867, + "learning_rate": 8.493495681472732e-07, + "loss": 0.0003, + "step": 8953 + }, + { + "epoch": 6.44056824312174, + "grad_norm": 1.9611103177681712, + "learning_rate": 8.490430158813914e-07, + "loss": 0.0166, + "step": 8954 + }, + { + "epoch": 6.441287538212552, + "grad_norm": 5.903969837736834, + "learning_rate": 8.4873649710839e-07, + "loss": 0.046, + "step": 8955 + }, + { + "epoch": 6.442006833303362, + "grad_norm": 5.493654129226573, + "learning_rate": 8.484300118440388e-07, + "loss": 0.0827, + "step": 8956 + }, + { + "epoch": 6.442726128394174, + "grad_norm": 0.5194971899101845, + "learning_rate": 8.481235601041075e-07, + "loss": 0.0021, + "step": 8957 + }, + { + "epoch": 6.4434454234849845, + "grad_norm": 2.1590536392572224, + "learning_rate": 8.478171419043632e-07, + "loss": 0.0261, + "step": 8958 + }, + { + "epoch": 6.444164718575796, + "grad_norm": 2.213600751581886, + "learning_rate": 8.475107572605719e-07, + "loss": 0.012, + "step": 8959 + }, + { + "epoch": 6.444884013666607, + "grad_norm": 1.085554908249116, + "learning_rate": 8.472044061884977e-07, + "loss": 0.0046, + "step": 8960 + }, + { + "epoch": 6.445603308757418, + "grad_norm": 0.3262649689824017, + "learning_rate": 8.468980887039026e-07, + "loss": 0.0018, + "step": 8961 + }, + { + "epoch": 6.446322603848229, + "grad_norm": 2.154388203613856, + "learning_rate": 8.465918048225474e-07, + "loss": 0.024, + "step": 8962 + }, + { + "epoch": 6.44704189893904, + "grad_norm": 0.5626380666792972, + "learning_rate": 8.462855545601907e-07, + "loss": 0.0024, + "step": 8963 + }, + { + "epoch": 6.447761194029851, + "grad_norm": 2.550592593591615, + "learning_rate": 8.459793379325899e-07, + "loss": 0.0153, + "step": 8964 + }, + { + "epoch": 6.448480489120662, + "grad_norm": 1.7094117392984753, + "learning_rate": 8.456731549554998e-07, + "loss": 0.0192, + "step": 8965 + }, + { + "epoch": 6.449199784211473, + "grad_norm": 1.5337929796751875, + "learning_rate": 8.453670056446749e-07, + "loss": 0.021, + "step": 8966 + }, + { + "epoch": 6.449919079302283, + "grad_norm": 2.7319339034478456, + "learning_rate": 8.450608900158666e-07, + "loss": 0.026, + "step": 8967 + }, + { + "epoch": 6.450638374393095, + "grad_norm": 3.3355657822259346, + "learning_rate": 8.447548080848255e-07, + "loss": 0.0387, + "step": 8968 + }, + { + "epoch": 6.451357669483905, + "grad_norm": 0.13385826735234174, + "learning_rate": 8.444487598672998e-07, + "loss": 0.0004, + "step": 8969 + }, + { + "epoch": 6.452076964574717, + "grad_norm": 0.206349847258314, + "learning_rate": 8.441427453790353e-07, + "loss": 0.0012, + "step": 8970 + }, + { + "epoch": 6.4527962596655275, + "grad_norm": 0.6583705135899217, + "learning_rate": 8.438367646357787e-07, + "loss": 0.0052, + "step": 8971 + }, + { + "epoch": 6.453515554756339, + "grad_norm": 1.0986634438184668, + "learning_rate": 8.435308176532723e-07, + "loss": 0.0148, + "step": 8972 + }, + { + "epoch": 6.45423484984715, + "grad_norm": 2.3706667641247545, + "learning_rate": 8.432249044472579e-07, + "loss": 0.0221, + "step": 8973 + }, + { + "epoch": 6.454954144937961, + "grad_norm": 0.1747583307609236, + "learning_rate": 8.429190250334749e-07, + "loss": 0.0004, + "step": 8974 + }, + { + "epoch": 6.455673440028772, + "grad_norm": 3.1163334064257606, + "learning_rate": 8.426131794276617e-07, + "loss": 0.0212, + "step": 8975 + }, + { + "epoch": 6.456392735119583, + "grad_norm": 2.4404740354018823, + "learning_rate": 8.423073676455542e-07, + "loss": 0.0195, + "step": 8976 + }, + { + "epoch": 6.457112030210394, + "grad_norm": 4.978076237478886, + "learning_rate": 8.420015897028874e-07, + "loss": 0.0569, + "step": 8977 + }, + { + "epoch": 6.457831325301205, + "grad_norm": 1.2062550207007061, + "learning_rate": 8.416958456153938e-07, + "loss": 0.0074, + "step": 8978 + }, + { + "epoch": 6.458550620392016, + "grad_norm": 2.1968007696927776, + "learning_rate": 8.413901353988045e-07, + "loss": 0.0042, + "step": 8979 + }, + { + "epoch": 6.459269915482826, + "grad_norm": 1.982444022798762, + "learning_rate": 8.41084459068849e-07, + "loss": 0.0213, + "step": 8980 + }, + { + "epoch": 6.459989210573638, + "grad_norm": 5.576606499256845, + "learning_rate": 8.40778816641254e-07, + "loss": 0.0749, + "step": 8981 + }, + { + "epoch": 6.460708505664448, + "grad_norm": 8.426091303697401, + "learning_rate": 8.404732081317469e-07, + "loss": 0.0932, + "step": 8982 + }, + { + "epoch": 6.46142780075526, + "grad_norm": 4.873821299046943, + "learning_rate": 8.401676335560509e-07, + "loss": 0.0706, + "step": 8983 + }, + { + "epoch": 6.4621470958460705, + "grad_norm": 0.3837468013100595, + "learning_rate": 8.398620929298886e-07, + "loss": 0.002, + "step": 8984 + }, + { + "epoch": 6.462866390936882, + "grad_norm": 0.061104524094953504, + "learning_rate": 8.395565862689803e-07, + "loss": 0.0003, + "step": 8985 + }, + { + "epoch": 6.463585686027693, + "grad_norm": 2.8464481318067056, + "learning_rate": 8.392511135890445e-07, + "loss": 0.0195, + "step": 8986 + }, + { + "epoch": 6.464304981118504, + "grad_norm": 5.513831698520571, + "learning_rate": 8.389456749057993e-07, + "loss": 0.0665, + "step": 8987 + }, + { + "epoch": 6.465024276209315, + "grad_norm": 1.9801266095299925, + "learning_rate": 8.386402702349596e-07, + "loss": 0.0193, + "step": 8988 + }, + { + "epoch": 6.465743571300126, + "grad_norm": 5.9338516215734485, + "learning_rate": 8.38334899592239e-07, + "loss": 0.0464, + "step": 8989 + }, + { + "epoch": 6.466462866390937, + "grad_norm": 1.5008004747814612, + "learning_rate": 8.380295629933494e-07, + "loss": 0.0043, + "step": 8990 + }, + { + "epoch": 6.467182161481748, + "grad_norm": 0.654716194899523, + "learning_rate": 8.377242604540007e-07, + "loss": 0.0023, + "step": 8991 + }, + { + "epoch": 6.467901456572559, + "grad_norm": 5.312835773650265, + "learning_rate": 8.374189919899014e-07, + "loss": 0.0835, + "step": 8992 + }, + { + "epoch": 6.46862075166337, + "grad_norm": 0.0008679643304771031, + "learning_rate": 8.371137576167583e-07, + "loss": 0.0, + "step": 8993 + }, + { + "epoch": 6.469340046754181, + "grad_norm": 0.006920763231601144, + "learning_rate": 8.368085573502759e-07, + "loss": 0.0, + "step": 8994 + }, + { + "epoch": 6.470059341844992, + "grad_norm": 1.3317932442929332, + "learning_rate": 8.365033912061574e-07, + "loss": 0.0109, + "step": 8995 + }, + { + "epoch": 6.470778636935803, + "grad_norm": 1.3098448689827236, + "learning_rate": 8.361982592001035e-07, + "loss": 0.0094, + "step": 8996 + }, + { + "epoch": 6.4714979320266135, + "grad_norm": 3.933804221198364, + "learning_rate": 8.35893161347815e-07, + "loss": 0.0535, + "step": 8997 + }, + { + "epoch": 6.472217227117425, + "grad_norm": 1.9638019884861195, + "learning_rate": 8.355880976649894e-07, + "loss": 0.0065, + "step": 8998 + }, + { + "epoch": 6.472936522208236, + "grad_norm": 0.1479192394280468, + "learning_rate": 8.352830681673223e-07, + "loss": 0.0004, + "step": 8999 + }, + { + "epoch": 6.473655817299047, + "grad_norm": 2.950809941019658, + "learning_rate": 8.349780728705082e-07, + "loss": 0.0142, + "step": 9000 + }, + { + "epoch": 6.474375112389858, + "grad_norm": 1.4646178770991831, + "learning_rate": 8.346731117902396e-07, + "loss": 0.0103, + "step": 9001 + }, + { + "epoch": 6.475094407480669, + "grad_norm": 0.02724322695386287, + "learning_rate": 8.343681849422072e-07, + "loss": 0.0001, + "step": 9002 + }, + { + "epoch": 6.47581370257148, + "grad_norm": 2.4326442460772153, + "learning_rate": 8.340632923421002e-07, + "loss": 0.0266, + "step": 9003 + }, + { + "epoch": 6.476532997662291, + "grad_norm": 4.9394904070346985, + "learning_rate": 8.33758434005606e-07, + "loss": 0.0838, + "step": 9004 + }, + { + "epoch": 6.477252292753102, + "grad_norm": 2.530402645559697, + "learning_rate": 8.3345360994841e-07, + "loss": 0.0303, + "step": 9005 + }, + { + "epoch": 6.477971587843913, + "grad_norm": 3.214950498842226, + "learning_rate": 8.331488201861966e-07, + "loss": 0.0372, + "step": 9006 + }, + { + "epoch": 6.478690882934724, + "grad_norm": 0.014158250993345585, + "learning_rate": 8.328440647346457e-07, + "loss": 0.0001, + "step": 9007 + }, + { + "epoch": 6.479410178025535, + "grad_norm": 0.06357696028851358, + "learning_rate": 8.325393436094393e-07, + "loss": 0.0002, + "step": 9008 + }, + { + "epoch": 6.480129473116346, + "grad_norm": 5.410152868118783, + "learning_rate": 8.322346568262554e-07, + "loss": 0.0564, + "step": 9009 + }, + { + "epoch": 6.480848768207157, + "grad_norm": 7.147010300469229, + "learning_rate": 8.319300044007706e-07, + "loss": 0.0958, + "step": 9010 + }, + { + "epoch": 6.481568063297968, + "grad_norm": 4.584996513572875, + "learning_rate": 8.316253863486599e-07, + "loss": 0.063, + "step": 9011 + }, + { + "epoch": 6.4822873583887795, + "grad_norm": 1.433636036138529, + "learning_rate": 8.313208026855958e-07, + "loss": 0.0168, + "step": 9012 + }, + { + "epoch": 6.48300665347959, + "grad_norm": 2.480026476925808, + "learning_rate": 8.310162534272509e-07, + "loss": 0.0205, + "step": 9013 + }, + { + "epoch": 6.483725948570401, + "grad_norm": 2.82768236704938, + "learning_rate": 8.30711738589294e-07, + "loss": 0.0152, + "step": 9014 + }, + { + "epoch": 6.484445243661212, + "grad_norm": 0.8697142418010388, + "learning_rate": 8.304072581873931e-07, + "loss": 0.0085, + "step": 9015 + }, + { + "epoch": 6.485164538752023, + "grad_norm": 0.018494592720449746, + "learning_rate": 8.301028122372143e-07, + "loss": 0.0001, + "step": 9016 + }, + { + "epoch": 6.485883833842834, + "grad_norm": 0.47017991446586543, + "learning_rate": 8.297984007544209e-07, + "loss": 0.0031, + "step": 9017 + }, + { + "epoch": 6.486603128933645, + "grad_norm": 2.232212551986779, + "learning_rate": 8.29494023754677e-07, + "loss": 0.0272, + "step": 9018 + }, + { + "epoch": 6.487322424024456, + "grad_norm": 0.9485377216997941, + "learning_rate": 8.291896812536433e-07, + "loss": 0.0052, + "step": 9019 + }, + { + "epoch": 6.488041719115267, + "grad_norm": 3.0145703349814172, + "learning_rate": 8.288853732669776e-07, + "loss": 0.0238, + "step": 9020 + }, + { + "epoch": 6.488761014206078, + "grad_norm": 3.046108183142833, + "learning_rate": 8.285810998103372e-07, + "loss": 0.04, + "step": 9021 + }, + { + "epoch": 6.489480309296889, + "grad_norm": 2.1578901793479117, + "learning_rate": 8.282768608993782e-07, + "loss": 0.0178, + "step": 9022 + }, + { + "epoch": 6.4901996043877, + "grad_norm": 2.140229202704074, + "learning_rate": 8.27972656549753e-07, + "loss": 0.0214, + "step": 9023 + }, + { + "epoch": 6.490918899478511, + "grad_norm": 7.136714202331538, + "learning_rate": 8.276684867771149e-07, + "loss": 0.0265, + "step": 9024 + }, + { + "epoch": 6.4916381945693225, + "grad_norm": 1.6562828776795715, + "learning_rate": 8.273643515971135e-07, + "loss": 0.0052, + "step": 9025 + }, + { + "epoch": 6.492357489660133, + "grad_norm": 2.0683586333712287, + "learning_rate": 8.27060251025397e-07, + "loss": 0.0051, + "step": 9026 + }, + { + "epoch": 6.493076784750944, + "grad_norm": 1.369015844474457, + "learning_rate": 8.26756185077612e-07, + "loss": 0.0057, + "step": 9027 + }, + { + "epoch": 6.493796079841755, + "grad_norm": 4.107537566357717, + "learning_rate": 8.264521537694024e-07, + "loss": 0.0529, + "step": 9028 + }, + { + "epoch": 6.494515374932566, + "grad_norm": 4.0240313708782525, + "learning_rate": 8.261481571164125e-07, + "loss": 0.0677, + "step": 9029 + }, + { + "epoch": 6.495234670023377, + "grad_norm": 3.690267364795766, + "learning_rate": 8.258441951342829e-07, + "loss": 0.0285, + "step": 9030 + }, + { + "epoch": 6.495953965114188, + "grad_norm": 3.2187383352968326, + "learning_rate": 8.255402678386529e-07, + "loss": 0.0293, + "step": 9031 + }, + { + "epoch": 6.496673260204999, + "grad_norm": 1.8787207103416212, + "learning_rate": 8.252363752451599e-07, + "loss": 0.0153, + "step": 9032 + }, + { + "epoch": 6.49739255529581, + "grad_norm": 0.012922586304003841, + "learning_rate": 8.2493251736944e-07, + "loss": 0.0001, + "step": 9033 + }, + { + "epoch": 6.498111850386621, + "grad_norm": 4.851643377132431, + "learning_rate": 8.246286942271271e-07, + "loss": 0.0634, + "step": 9034 + }, + { + "epoch": 6.498831145477432, + "grad_norm": 0.035232967368464166, + "learning_rate": 8.243249058338534e-07, + "loss": 0.0002, + "step": 9035 + }, + { + "epoch": 6.499550440568243, + "grad_norm": 2.343317663465008, + "learning_rate": 8.240211522052497e-07, + "loss": 0.0217, + "step": 9036 + }, + { + "epoch": 6.500269735659054, + "grad_norm": 4.279148056891122, + "learning_rate": 8.237174333569443e-07, + "loss": 0.0326, + "step": 9037 + }, + { + "epoch": 6.5009890307498654, + "grad_norm": 4.278933151511634, + "learning_rate": 8.234137493045631e-07, + "loss": 0.0503, + "step": 9038 + }, + { + "epoch": 6.501708325840676, + "grad_norm": 3.675013356896929, + "learning_rate": 8.231101000637333e-07, + "loss": 0.0301, + "step": 9039 + }, + { + "epoch": 6.5024276209314875, + "grad_norm": 3.4245382407384524, + "learning_rate": 8.22806485650077e-07, + "loss": 0.0243, + "step": 9040 + }, + { + "epoch": 6.503146916022298, + "grad_norm": 2.4369782053835487, + "learning_rate": 8.225029060792156e-07, + "loss": 0.0141, + "step": 9041 + }, + { + "epoch": 6.50386621111311, + "grad_norm": 1.1799935502273156, + "learning_rate": 8.221993613667691e-07, + "loss": 0.0032, + "step": 9042 + }, + { + "epoch": 6.50458550620392, + "grad_norm": 0.8514744249801934, + "learning_rate": 8.218958515283552e-07, + "loss": 0.0045, + "step": 9043 + }, + { + "epoch": 6.505304801294731, + "grad_norm": 3.8127977546568896, + "learning_rate": 8.215923765795895e-07, + "loss": 0.0485, + "step": 9044 + }, + { + "epoch": 6.506024096385542, + "grad_norm": 4.817633395172584, + "learning_rate": 8.212889365360879e-07, + "loss": 0.0761, + "step": 9045 + }, + { + "epoch": 6.506743391476353, + "grad_norm": 1.6509601906451454, + "learning_rate": 8.209855314134616e-07, + "loss": 0.0069, + "step": 9046 + }, + { + "epoch": 6.507462686567164, + "grad_norm": 3.095977917513011, + "learning_rate": 8.206821612273217e-07, + "loss": 0.0367, + "step": 9047 + }, + { + "epoch": 6.508181981657975, + "grad_norm": 0.024147160187432983, + "learning_rate": 8.203788259932773e-07, + "loss": 0.0001, + "step": 9048 + }, + { + "epoch": 6.508901276748786, + "grad_norm": 4.0554232685757485, + "learning_rate": 8.200755257269352e-07, + "loss": 0.0564, + "step": 9049 + }, + { + "epoch": 6.509620571839597, + "grad_norm": 1.5572840681862783, + "learning_rate": 8.197722604439009e-07, + "loss": 0.0096, + "step": 9050 + }, + { + "epoch": 6.5103398669304084, + "grad_norm": 10.42230184812466, + "learning_rate": 8.194690301597777e-07, + "loss": 0.1572, + "step": 9051 + }, + { + "epoch": 6.511059162021219, + "grad_norm": 0.09919915134967089, + "learning_rate": 8.191658348901679e-07, + "loss": 0.0003, + "step": 9052 + }, + { + "epoch": 6.5117784571120305, + "grad_norm": 3.5023241419540345, + "learning_rate": 8.188626746506707e-07, + "loss": 0.0321, + "step": 9053 + }, + { + "epoch": 6.512497752202841, + "grad_norm": 0.6871613359221471, + "learning_rate": 8.185595494568842e-07, + "loss": 0.003, + "step": 9054 + }, + { + "epoch": 6.513217047293653, + "grad_norm": 3.9388180106460116, + "learning_rate": 8.182564593244059e-07, + "loss": 0.0308, + "step": 9055 + }, + { + "epoch": 6.513936342384463, + "grad_norm": 2.180505878502797, + "learning_rate": 8.17953404268829e-07, + "loss": 0.0381, + "step": 9056 + }, + { + "epoch": 6.514655637475274, + "grad_norm": 3.649408581772886, + "learning_rate": 8.176503843057472e-07, + "loss": 0.0383, + "step": 9057 + }, + { + "epoch": 6.515374932566085, + "grad_norm": 2.6046540632343618, + "learning_rate": 8.173473994507508e-07, + "loss": 0.0558, + "step": 9058 + }, + { + "epoch": 6.516094227656897, + "grad_norm": 1.5047597480198458, + "learning_rate": 8.170444497194284e-07, + "loss": 0.0133, + "step": 9059 + }, + { + "epoch": 6.516813522747707, + "grad_norm": 0.25163390364836113, + "learning_rate": 8.167415351273688e-07, + "loss": 0.0008, + "step": 9060 + }, + { + "epoch": 6.517532817838518, + "grad_norm": 0.9969964202545379, + "learning_rate": 8.164386556901564e-07, + "loss": 0.0016, + "step": 9061 + }, + { + "epoch": 6.518252112929329, + "grad_norm": 1.571913562792056, + "learning_rate": 8.161358114233751e-07, + "loss": 0.0136, + "step": 9062 + }, + { + "epoch": 6.51897140802014, + "grad_norm": 5.537259099788676, + "learning_rate": 8.158330023426068e-07, + "loss": 0.0887, + "step": 9063 + }, + { + "epoch": 6.5196907031109514, + "grad_norm": 0.5900406779450229, + "learning_rate": 8.155302284634319e-07, + "loss": 0.0017, + "step": 9064 + }, + { + "epoch": 6.520409998201762, + "grad_norm": 3.0379571508336034, + "learning_rate": 8.152274898014279e-07, + "loss": 0.0437, + "step": 9065 + }, + { + "epoch": 6.5211292932925735, + "grad_norm": 0.007852050632606564, + "learning_rate": 8.149247863721716e-07, + "loss": 0.0, + "step": 9066 + }, + { + "epoch": 6.521848588383384, + "grad_norm": 0.5279246378294468, + "learning_rate": 8.146221181912378e-07, + "loss": 0.0044, + "step": 9067 + }, + { + "epoch": 6.522567883474196, + "grad_norm": 0.018690246732179155, + "learning_rate": 8.143194852741992e-07, + "loss": 0.0001, + "step": 9068 + }, + { + "epoch": 6.523287178565006, + "grad_norm": 7.7681703031251255, + "learning_rate": 8.140168876366267e-07, + "loss": 0.0748, + "step": 9069 + }, + { + "epoch": 6.524006473655818, + "grad_norm": 0.1418162363657988, + "learning_rate": 8.137143252940889e-07, + "loss": 0.0004, + "step": 9070 + }, + { + "epoch": 6.524725768746628, + "grad_norm": 4.456061254440884, + "learning_rate": 8.134117982621546e-07, + "loss": 0.0478, + "step": 9071 + }, + { + "epoch": 6.52544506383744, + "grad_norm": 0.03521379002937794, + "learning_rate": 8.131093065563883e-07, + "loss": 0.0001, + "step": 9072 + }, + { + "epoch": 6.52616435892825, + "grad_norm": 2.186774218967586, + "learning_rate": 8.128068501923541e-07, + "loss": 0.003, + "step": 9073 + }, + { + "epoch": 6.526883654019061, + "grad_norm": 0.1870835233910017, + "learning_rate": 8.125044291856137e-07, + "loss": 0.0006, + "step": 9074 + }, + { + "epoch": 6.527602949109872, + "grad_norm": 2.8882096922754377, + "learning_rate": 8.122020435517268e-07, + "loss": 0.0288, + "step": 9075 + }, + { + "epoch": 6.528322244200683, + "grad_norm": 2.1787871809181674, + "learning_rate": 8.118996933062528e-07, + "loss": 0.0118, + "step": 9076 + }, + { + "epoch": 6.5290415392914944, + "grad_norm": 3.643487746070624, + "learning_rate": 8.115973784647474e-07, + "loss": 0.0564, + "step": 9077 + }, + { + "epoch": 6.529760834382305, + "grad_norm": 0.7162299731987567, + "learning_rate": 8.112950990427659e-07, + "loss": 0.006, + "step": 9078 + }, + { + "epoch": 6.5304801294731165, + "grad_norm": 1.0724774505979167, + "learning_rate": 8.1099285505586e-07, + "loss": 0.0048, + "step": 9079 + }, + { + "epoch": 6.531199424563927, + "grad_norm": 5.072828376599335, + "learning_rate": 8.106906465195805e-07, + "loss": 0.0827, + "step": 9080 + }, + { + "epoch": 6.531918719654739, + "grad_norm": 1.6812705808955324, + "learning_rate": 8.103884734494779e-07, + "loss": 0.012, + "step": 9081 + }, + { + "epoch": 6.532638014745549, + "grad_norm": 0.1432516864671047, + "learning_rate": 8.100863358610991e-07, + "loss": 0.0004, + "step": 9082 + }, + { + "epoch": 6.533357309836361, + "grad_norm": 3.072589143816585, + "learning_rate": 8.097842337699892e-07, + "loss": 0.0091, + "step": 9083 + }, + { + "epoch": 6.534076604927171, + "grad_norm": 1.464949989494499, + "learning_rate": 8.094821671916921e-07, + "loss": 0.0099, + "step": 9084 + }, + { + "epoch": 6.534795900017983, + "grad_norm": 0.7224658706299573, + "learning_rate": 8.091801361417499e-07, + "loss": 0.0014, + "step": 9085 + }, + { + "epoch": 6.535515195108793, + "grad_norm": 0.055936977327535255, + "learning_rate": 8.088781406357015e-07, + "loss": 0.0002, + "step": 9086 + }, + { + "epoch": 6.536234490199604, + "grad_norm": 2.098690826251373, + "learning_rate": 8.085761806890869e-07, + "loss": 0.0054, + "step": 9087 + }, + { + "epoch": 6.536953785290415, + "grad_norm": 5.022263765939293, + "learning_rate": 8.082742563174414e-07, + "loss": 0.0334, + "step": 9088 + }, + { + "epoch": 6.537673080381227, + "grad_norm": 4.1610920342493225, + "learning_rate": 8.079723675362998e-07, + "loss": 0.0469, + "step": 9089 + }, + { + "epoch": 6.538392375472037, + "grad_norm": 1.3541422978185296, + "learning_rate": 8.076705143611944e-07, + "loss": 0.0061, + "step": 9090 + }, + { + "epoch": 6.539111670562848, + "grad_norm": 1.684094918850468, + "learning_rate": 8.073686968076568e-07, + "loss": 0.01, + "step": 9091 + }, + { + "epoch": 6.5398309656536595, + "grad_norm": 2.345854181603881, + "learning_rate": 8.070669148912155e-07, + "loss": 0.0325, + "step": 9092 + }, + { + "epoch": 6.54055026074447, + "grad_norm": 5.690988563911193, + "learning_rate": 8.067651686273975e-07, + "loss": 0.0571, + "step": 9093 + }, + { + "epoch": 6.541269555835282, + "grad_norm": 3.1371835249714835, + "learning_rate": 8.064634580317288e-07, + "loss": 0.0357, + "step": 9094 + }, + { + "epoch": 6.541988850926092, + "grad_norm": 0.1311713426390841, + "learning_rate": 8.061617831197327e-07, + "loss": 0.0004, + "step": 9095 + }, + { + "epoch": 6.542708146016904, + "grad_norm": 2.5182358098507005, + "learning_rate": 8.058601439069302e-07, + "loss": 0.0275, + "step": 9096 + }, + { + "epoch": 6.543427441107714, + "grad_norm": 8.751079710612775, + "learning_rate": 8.055585404088424e-07, + "loss": 0.0453, + "step": 9097 + }, + { + "epoch": 6.544146736198526, + "grad_norm": 0.20819883000753434, + "learning_rate": 8.052569726409867e-07, + "loss": 0.0006, + "step": 9098 + }, + { + "epoch": 6.544866031289336, + "grad_norm": 1.619031555089714, + "learning_rate": 8.049554406188794e-07, + "loss": 0.0241, + "step": 9099 + }, + { + "epoch": 6.545585326380148, + "grad_norm": 0.02433748179842984, + "learning_rate": 8.046539443580349e-07, + "loss": 0.0002, + "step": 9100 + }, + { + "epoch": 6.546304621470958, + "grad_norm": 2.525801970413565, + "learning_rate": 8.043524838739647e-07, + "loss": 0.0418, + "step": 9101 + }, + { + "epoch": 6.54702391656177, + "grad_norm": 2.4484060445339377, + "learning_rate": 8.040510591821812e-07, + "loss": 0.0213, + "step": 9102 + }, + { + "epoch": 6.54774321165258, + "grad_norm": 2.133294945029812, + "learning_rate": 8.037496702981922e-07, + "loss": 0.0062, + "step": 9103 + }, + { + "epoch": 6.548462506743391, + "grad_norm": 2.551420437111838, + "learning_rate": 8.034483172375051e-07, + "loss": 0.027, + "step": 9104 + }, + { + "epoch": 6.5491818018342025, + "grad_norm": 0.25217014427514894, + "learning_rate": 8.031470000156249e-07, + "loss": 0.0013, + "step": 9105 + }, + { + "epoch": 6.549901096925013, + "grad_norm": 0.025034031121782433, + "learning_rate": 8.028457186480547e-07, + "loss": 0.0001, + "step": 9106 + }, + { + "epoch": 6.550620392015825, + "grad_norm": 2.4860366374398484, + "learning_rate": 8.02544473150296e-07, + "loss": 0.0239, + "step": 9107 + }, + { + "epoch": 6.551339687106635, + "grad_norm": 2.9936080599262733, + "learning_rate": 8.022432635378488e-07, + "loss": 0.051, + "step": 9108 + }, + { + "epoch": 6.552058982197447, + "grad_norm": 1.0401497179213475, + "learning_rate": 8.019420898262103e-07, + "loss": 0.0066, + "step": 9109 + }, + { + "epoch": 6.552778277288257, + "grad_norm": 0.047921396843807834, + "learning_rate": 8.016409520308768e-07, + "loss": 0.0002, + "step": 9110 + }, + { + "epoch": 6.553497572379069, + "grad_norm": 0.12255280546948558, + "learning_rate": 8.013398501673421e-07, + "loss": 0.0003, + "step": 9111 + }, + { + "epoch": 6.554216867469879, + "grad_norm": 0.738605322552749, + "learning_rate": 8.010387842510981e-07, + "loss": 0.0051, + "step": 9112 + }, + { + "epoch": 6.554936162560691, + "grad_norm": 2.341616433280017, + "learning_rate": 8.007377542976363e-07, + "loss": 0.0125, + "step": 9113 + }, + { + "epoch": 6.555655457651501, + "grad_norm": 3.5121224360728953, + "learning_rate": 8.004367603224445e-07, + "loss": 0.0438, + "step": 9114 + }, + { + "epoch": 6.556374752742313, + "grad_norm": 1.7975218214489652, + "learning_rate": 8.001358023410094e-07, + "loss": 0.0098, + "step": 9115 + }, + { + "epoch": 6.557094047833123, + "grad_norm": 3.4878595753606683, + "learning_rate": 7.998348803688158e-07, + "loss": 0.0274, + "step": 9116 + }, + { + "epoch": 6.557813342923935, + "grad_norm": 2.5179530627798767, + "learning_rate": 7.995339944213462e-07, + "loss": 0.0277, + "step": 9117 + }, + { + "epoch": 6.5585326380147455, + "grad_norm": 2.1884938383833865, + "learning_rate": 7.992331445140826e-07, + "loss": 0.0251, + "step": 9118 + }, + { + "epoch": 6.559251933105557, + "grad_norm": 1.9440703564005861, + "learning_rate": 7.989323306625041e-07, + "loss": 0.0244, + "step": 9119 + }, + { + "epoch": 6.559971228196368, + "grad_norm": 3.892452142767073, + "learning_rate": 7.986315528820878e-07, + "loss": 0.0465, + "step": 9120 + }, + { + "epoch": 6.560690523287178, + "grad_norm": 1.2230506758515454, + "learning_rate": 7.983308111883094e-07, + "loss": 0.0029, + "step": 9121 + }, + { + "epoch": 6.56140981837799, + "grad_norm": 1.816290762732067, + "learning_rate": 7.980301055966423e-07, + "loss": 0.0041, + "step": 9122 + }, + { + "epoch": 6.5621291134688, + "grad_norm": 0.004836245304332215, + "learning_rate": 7.977294361225587e-07, + "loss": 0.0, + "step": 9123 + }, + { + "epoch": 6.562848408559612, + "grad_norm": 2.3402817990934066, + "learning_rate": 7.974288027815285e-07, + "loss": 0.0417, + "step": 9124 + }, + { + "epoch": 6.563567703650422, + "grad_norm": 1.0386239105569515, + "learning_rate": 7.971282055890195e-07, + "loss": 0.016, + "step": 9125 + }, + { + "epoch": 6.564286998741234, + "grad_norm": 0.013622239514545184, + "learning_rate": 7.968276445604983e-07, + "loss": 0.0001, + "step": 9126 + }, + { + "epoch": 6.565006293832044, + "grad_norm": 3.843303833660665, + "learning_rate": 7.965271197114284e-07, + "loss": 0.0466, + "step": 9127 + }, + { + "epoch": 6.565725588922856, + "grad_norm": 4.718840596239944, + "learning_rate": 7.96226631057274e-07, + "loss": 0.0815, + "step": 9128 + }, + { + "epoch": 6.566444884013666, + "grad_norm": 0.12274027723285953, + "learning_rate": 7.959261786134946e-07, + "loss": 0.0002, + "step": 9129 + }, + { + "epoch": 6.567164179104478, + "grad_norm": 2.0008014797326705, + "learning_rate": 7.956257623955495e-07, + "loss": 0.0327, + "step": 9130 + }, + { + "epoch": 6.5678834741952885, + "grad_norm": 1.643762221475917, + "learning_rate": 7.953253824188953e-07, + "loss": 0.0089, + "step": 9131 + }, + { + "epoch": 6.5686027692861, + "grad_norm": 4.515995083344554, + "learning_rate": 7.950250386989873e-07, + "loss": 0.1168, + "step": 9132 + }, + { + "epoch": 6.569322064376911, + "grad_norm": 2.1254989520345906, + "learning_rate": 7.947247312512776e-07, + "loss": 0.0201, + "step": 9133 + }, + { + "epoch": 6.570041359467721, + "grad_norm": 3.8970398749650776, + "learning_rate": 7.944244600912196e-07, + "loss": 0.0289, + "step": 9134 + }, + { + "epoch": 6.570760654558533, + "grad_norm": 2.9320708048886996, + "learning_rate": 7.941242252342614e-07, + "loss": 0.0289, + "step": 9135 + }, + { + "epoch": 6.571479949649344, + "grad_norm": 6.424454264592999, + "learning_rate": 7.938240266958513e-07, + "loss": 0.0533, + "step": 9136 + }, + { + "epoch": 6.572199244740155, + "grad_norm": 6.784785167946617, + "learning_rate": 7.93523864491435e-07, + "loss": 0.1202, + "step": 9137 + }, + { + "epoch": 6.572918539830965, + "grad_norm": 1.1267268603397185, + "learning_rate": 7.93223738636455e-07, + "loss": 0.0121, + "step": 9138 + }, + { + "epoch": 6.573637834921777, + "grad_norm": 3.4653280710661627, + "learning_rate": 7.929236491463549e-07, + "loss": 0.0321, + "step": 9139 + }, + { + "epoch": 6.574357130012587, + "grad_norm": 1.3838383057903176, + "learning_rate": 7.926235960365743e-07, + "loss": 0.0062, + "step": 9140 + }, + { + "epoch": 6.575076425103399, + "grad_norm": 1.5157971796360263, + "learning_rate": 7.923235793225513e-07, + "loss": 0.0217, + "step": 9141 + }, + { + "epoch": 6.575795720194209, + "grad_norm": 0.1805287950454141, + "learning_rate": 7.920235990197226e-07, + "loss": 0.0009, + "step": 9142 + }, + { + "epoch": 6.576515015285021, + "grad_norm": 1.034272828521846, + "learning_rate": 7.917236551435219e-07, + "loss": 0.0084, + "step": 9143 + }, + { + "epoch": 6.5772343103758315, + "grad_norm": 3.6381272809991927, + "learning_rate": 7.914237477093831e-07, + "loss": 0.0369, + "step": 9144 + }, + { + "epoch": 6.577953605466643, + "grad_norm": 4.460190460352121, + "learning_rate": 7.911238767327363e-07, + "loss": 0.0804, + "step": 9145 + }, + { + "epoch": 6.578672900557454, + "grad_norm": 0.04618929857267571, + "learning_rate": 7.908240422290105e-07, + "loss": 0.0001, + "step": 9146 + }, + { + "epoch": 6.579392195648265, + "grad_norm": 2.0659714181593487, + "learning_rate": 7.905242442136326e-07, + "loss": 0.0185, + "step": 9147 + }, + { + "epoch": 6.580111490739076, + "grad_norm": 5.541518648984874, + "learning_rate": 7.902244827020271e-07, + "loss": 0.052, + "step": 9148 + }, + { + "epoch": 6.580830785829887, + "grad_norm": 8.448873415137106, + "learning_rate": 7.899247577096187e-07, + "loss": 0.09, + "step": 9149 + }, + { + "epoch": 6.581550080920698, + "grad_norm": 2.8163495233162834, + "learning_rate": 7.896250692518284e-07, + "loss": 0.044, + "step": 9150 + }, + { + "epoch": 6.582269376011508, + "grad_norm": 1.4958786407677807, + "learning_rate": 7.893254173440748e-07, + "loss": 0.0209, + "step": 9151 + }, + { + "epoch": 6.58298867110232, + "grad_norm": 4.0350835647342596, + "learning_rate": 7.890258020017762e-07, + "loss": 0.0373, + "step": 9152 + }, + { + "epoch": 6.58370796619313, + "grad_norm": 2.4893816571398335, + "learning_rate": 7.887262232403479e-07, + "loss": 0.0245, + "step": 9153 + }, + { + "epoch": 6.584427261283942, + "grad_norm": 1.8796288189036094, + "learning_rate": 7.884266810752037e-07, + "loss": 0.0191, + "step": 9154 + }, + { + "epoch": 6.585146556374752, + "grad_norm": 4.267236101140209, + "learning_rate": 7.881271755217564e-07, + "loss": 0.0855, + "step": 9155 + }, + { + "epoch": 6.585865851465564, + "grad_norm": 6.148147030236824, + "learning_rate": 7.878277065954155e-07, + "loss": 0.0711, + "step": 9156 + }, + { + "epoch": 6.5865851465563745, + "grad_norm": 3.0229081265424544, + "learning_rate": 7.875282743115893e-07, + "loss": 0.0476, + "step": 9157 + }, + { + "epoch": 6.587304441647186, + "grad_norm": 2.3482806668838463, + "learning_rate": 7.872288786856841e-07, + "loss": 0.0215, + "step": 9158 + }, + { + "epoch": 6.588023736737997, + "grad_norm": 1.065012954551068, + "learning_rate": 7.869295197331036e-07, + "loss": 0.0079, + "step": 9159 + }, + { + "epoch": 6.588743031828808, + "grad_norm": 8.615995696965959, + "learning_rate": 7.866301974692518e-07, + "loss": 0.1008, + "step": 9160 + }, + { + "epoch": 6.589462326919619, + "grad_norm": 3.4682828009624593, + "learning_rate": 7.863309119095284e-07, + "loss": 0.0546, + "step": 9161 + }, + { + "epoch": 6.59018162201043, + "grad_norm": 1.7217107716272346, + "learning_rate": 7.860316630693327e-07, + "loss": 0.0084, + "step": 9162 + }, + { + "epoch": 6.590900917101241, + "grad_norm": 3.913433116992014, + "learning_rate": 7.85732450964061e-07, + "loss": 0.0749, + "step": 9163 + }, + { + "epoch": 6.591620212192051, + "grad_norm": 2.886645201227442, + "learning_rate": 7.854332756091087e-07, + "loss": 0.0254, + "step": 9164 + }, + { + "epoch": 6.592339507282863, + "grad_norm": 1.5262574846542398, + "learning_rate": 7.851341370198688e-07, + "loss": 0.0149, + "step": 9165 + }, + { + "epoch": 6.593058802373674, + "grad_norm": 1.249384398686335, + "learning_rate": 7.848350352117324e-07, + "loss": 0.0038, + "step": 9166 + }, + { + "epoch": 6.593778097464485, + "grad_norm": 2.7905924238575737, + "learning_rate": 7.845359702000889e-07, + "loss": 0.0269, + "step": 9167 + }, + { + "epoch": 6.594497392555295, + "grad_norm": 7.177392424332668, + "learning_rate": 7.842369420003257e-07, + "loss": 0.1073, + "step": 9168 + }, + { + "epoch": 6.595216687646107, + "grad_norm": 0.31316262868089223, + "learning_rate": 7.839379506278275e-07, + "loss": 0.0007, + "step": 9169 + }, + { + "epoch": 6.5959359827369175, + "grad_norm": 2.565639809684226, + "learning_rate": 7.836389960979797e-07, + "loss": 0.0035, + "step": 9170 + }, + { + "epoch": 6.596655277827729, + "grad_norm": 2.07391579765218, + "learning_rate": 7.833400784261628e-07, + "loss": 0.0168, + "step": 9171 + }, + { + "epoch": 6.59737457291854, + "grad_norm": 0.8278743027981645, + "learning_rate": 7.830411976277571e-07, + "loss": 0.0022, + "step": 9172 + }, + { + "epoch": 6.598093868009351, + "grad_norm": 3.371580564934196, + "learning_rate": 7.827423537181404e-07, + "loss": 0.0279, + "step": 9173 + }, + { + "epoch": 6.598813163100162, + "grad_norm": 2.6309459792210794, + "learning_rate": 7.824435467126886e-07, + "loss": 0.0247, + "step": 9174 + }, + { + "epoch": 6.599532458190973, + "grad_norm": 4.632684958823881, + "learning_rate": 7.821447766267752e-07, + "loss": 0.0524, + "step": 9175 + }, + { + "epoch": 6.600251753281784, + "grad_norm": 3.750031860014004, + "learning_rate": 7.818460434757741e-07, + "loss": 0.0519, + "step": 9176 + }, + { + "epoch": 6.600971048372595, + "grad_norm": 3.9400994949789276, + "learning_rate": 7.815473472750545e-07, + "loss": 0.0865, + "step": 9177 + }, + { + "epoch": 6.601690343463406, + "grad_norm": 3.8631631781368214, + "learning_rate": 7.812486880399853e-07, + "loss": 0.0806, + "step": 9178 + }, + { + "epoch": 6.602409638554217, + "grad_norm": 3.118380623663954, + "learning_rate": 7.809500657859327e-07, + "loss": 0.0172, + "step": 9179 + }, + { + "epoch": 6.603128933645028, + "grad_norm": 1.9756137338336994, + "learning_rate": 7.806514805282615e-07, + "loss": 0.0162, + "step": 9180 + }, + { + "epoch": 6.603848228735838, + "grad_norm": 0.355364971024024, + "learning_rate": 7.803529322823343e-07, + "loss": 0.0029, + "step": 9181 + }, + { + "epoch": 6.60456752382665, + "grad_norm": 1.7958043428896433, + "learning_rate": 7.800544210635121e-07, + "loss": 0.0254, + "step": 9182 + }, + { + "epoch": 6.6052868189174605, + "grad_norm": 4.7655101994494125, + "learning_rate": 7.797559468871538e-07, + "loss": 0.0814, + "step": 9183 + }, + { + "epoch": 6.606006114008272, + "grad_norm": 2.3575149822407635, + "learning_rate": 7.794575097686164e-07, + "loss": 0.0172, + "step": 9184 + }, + { + "epoch": 6.606725409099083, + "grad_norm": 3.455858869179822, + "learning_rate": 7.791591097232542e-07, + "loss": 0.0214, + "step": 9185 + }, + { + "epoch": 6.607444704189894, + "grad_norm": 2.0615865657396473, + "learning_rate": 7.788607467664221e-07, + "loss": 0.0116, + "step": 9186 + }, + { + "epoch": 6.608163999280705, + "grad_norm": 2.6112725861738366, + "learning_rate": 7.785624209134703e-07, + "loss": 0.0202, + "step": 9187 + }, + { + "epoch": 6.608883294371516, + "grad_norm": 0.8209195640876248, + "learning_rate": 7.782641321797484e-07, + "loss": 0.0029, + "step": 9188 + }, + { + "epoch": 6.609602589462327, + "grad_norm": 0.016731093231987133, + "learning_rate": 7.779658805806039e-07, + "loss": 0.0, + "step": 9189 + }, + { + "epoch": 6.610321884553138, + "grad_norm": 1.2403348475015386, + "learning_rate": 7.776676661313817e-07, + "loss": 0.0076, + "step": 9190 + }, + { + "epoch": 6.611041179643949, + "grad_norm": 2.3232787037652107, + "learning_rate": 7.773694888474268e-07, + "loss": 0.0234, + "step": 9191 + }, + { + "epoch": 6.61176047473476, + "grad_norm": 2.5699303868830174, + "learning_rate": 7.770713487440803e-07, + "loss": 0.0173, + "step": 9192 + }, + { + "epoch": 6.612479769825571, + "grad_norm": 1.1714391754981597, + "learning_rate": 7.767732458366818e-07, + "loss": 0.0103, + "step": 9193 + }, + { + "epoch": 6.613199064916382, + "grad_norm": 1.0447056086113435, + "learning_rate": 7.764751801405694e-07, + "loss": 0.0058, + "step": 9194 + }, + { + "epoch": 6.613918360007193, + "grad_norm": 0.366854815796995, + "learning_rate": 7.761771516710791e-07, + "loss": 0.0065, + "step": 9195 + }, + { + "epoch": 6.614637655098004, + "grad_norm": 3.475088139880463, + "learning_rate": 7.75879160443545e-07, + "loss": 0.0362, + "step": 9196 + }, + { + "epoch": 6.615356950188815, + "grad_norm": 4.2596596021231905, + "learning_rate": 7.755812064732993e-07, + "loss": 0.0617, + "step": 9197 + }, + { + "epoch": 6.616076245279626, + "grad_norm": 0.004009900481763295, + "learning_rate": 7.75283289775672e-07, + "loss": 0.0, + "step": 9198 + }, + { + "epoch": 6.616795540370437, + "grad_norm": 5.043167297397006, + "learning_rate": 7.749854103659918e-07, + "loss": 0.0533, + "step": 9199 + }, + { + "epoch": 6.617514835461248, + "grad_norm": 4.869106460612118, + "learning_rate": 7.746875682595851e-07, + "loss": 0.0456, + "step": 9200 + }, + { + "epoch": 6.618234130552059, + "grad_norm": 0.02750597220746227, + "learning_rate": 7.743897634717754e-07, + "loss": 0.0001, + "step": 9201 + }, + { + "epoch": 6.61895342564287, + "grad_norm": 2.260767859058794, + "learning_rate": 7.740919960178869e-07, + "loss": 0.0213, + "step": 9202 + }, + { + "epoch": 6.619672720733681, + "grad_norm": 0.4964628715306744, + "learning_rate": 7.737942659132393e-07, + "loss": 0.0022, + "step": 9203 + }, + { + "epoch": 6.620392015824492, + "grad_norm": 2.6642894546031237, + "learning_rate": 7.734965731731518e-07, + "loss": 0.0345, + "step": 9204 + }, + { + "epoch": 6.621111310915303, + "grad_norm": 4.234238267880308, + "learning_rate": 7.731989178129409e-07, + "loss": 0.0754, + "step": 9205 + }, + { + "epoch": 6.621830606006114, + "grad_norm": 2.2516807685944644, + "learning_rate": 7.729012998479208e-07, + "loss": 0.0345, + "step": 9206 + }, + { + "epoch": 6.622549901096925, + "grad_norm": 2.4962226520677744, + "learning_rate": 7.726037192934058e-07, + "loss": 0.0305, + "step": 9207 + }, + { + "epoch": 6.623269196187736, + "grad_norm": 1.4811978543267024, + "learning_rate": 7.723061761647066e-07, + "loss": 0.0195, + "step": 9208 + }, + { + "epoch": 6.623988491278547, + "grad_norm": 1.3240038025338545, + "learning_rate": 7.720086704771322e-07, + "loss": 0.011, + "step": 9209 + }, + { + "epoch": 6.624707786369358, + "grad_norm": 2.5960680891114514, + "learning_rate": 7.717112022459894e-07, + "loss": 0.0256, + "step": 9210 + }, + { + "epoch": 6.625427081460169, + "grad_norm": 1.1515772671109032, + "learning_rate": 7.714137714865831e-07, + "loss": 0.0122, + "step": 9211 + }, + { + "epoch": 6.62614637655098, + "grad_norm": 0.4974156994767069, + "learning_rate": 7.711163782142178e-07, + "loss": 0.002, + "step": 9212 + }, + { + "epoch": 6.6268656716417915, + "grad_norm": 3.4824858015429014, + "learning_rate": 7.708190224441942e-07, + "loss": 0.0573, + "step": 9213 + }, + { + "epoch": 6.627584966732602, + "grad_norm": 1.8214087803048777, + "learning_rate": 7.705217041918122e-07, + "loss": 0.0127, + "step": 9214 + }, + { + "epoch": 6.628304261823413, + "grad_norm": 0.01985904814995546, + "learning_rate": 7.70224423472369e-07, + "loss": 0.0001, + "step": 9215 + }, + { + "epoch": 6.629023556914224, + "grad_norm": 1.8543248228526918, + "learning_rate": 7.699271803011603e-07, + "loss": 0.0085, + "step": 9216 + }, + { + "epoch": 6.629742852005035, + "grad_norm": 2.158672340879328, + "learning_rate": 7.69629974693479e-07, + "loss": 0.0492, + "step": 9217 + }, + { + "epoch": 6.630462147095846, + "grad_norm": 2.4050366223512767, + "learning_rate": 7.693328066646185e-07, + "loss": 0.0258, + "step": 9218 + }, + { + "epoch": 6.631181442186657, + "grad_norm": 1.5137893035587797, + "learning_rate": 7.690356762298674e-07, + "loss": 0.0175, + "step": 9219 + }, + { + "epoch": 6.631900737277468, + "grad_norm": 2.6504485079669844, + "learning_rate": 7.687385834045142e-07, + "loss": 0.0369, + "step": 9220 + }, + { + "epoch": 6.632620032368279, + "grad_norm": 0.025941179885917917, + "learning_rate": 7.684415282038444e-07, + "loss": 0.0001, + "step": 9221 + }, + { + "epoch": 6.63333932745909, + "grad_norm": 2.4812286198716778, + "learning_rate": 7.681445106431423e-07, + "loss": 0.029, + "step": 9222 + }, + { + "epoch": 6.634058622549901, + "grad_norm": 1.7437988093872, + "learning_rate": 7.678475307376898e-07, + "loss": 0.0097, + "step": 9223 + }, + { + "epoch": 6.6347779176407125, + "grad_norm": 1.0707972460183572, + "learning_rate": 7.675505885027673e-07, + "loss": 0.0044, + "step": 9224 + }, + { + "epoch": 6.635497212731523, + "grad_norm": 3.7146789168537424, + "learning_rate": 7.672536839536525e-07, + "loss": 0.0362, + "step": 9225 + }, + { + "epoch": 6.6362165078223345, + "grad_norm": 3.057872199559811, + "learning_rate": 7.669568171056221e-07, + "loss": 0.0196, + "step": 9226 + }, + { + "epoch": 6.636935802913145, + "grad_norm": 0.9886899566261532, + "learning_rate": 7.666599879739497e-07, + "loss": 0.0055, + "step": 9227 + }, + { + "epoch": 6.637655098003956, + "grad_norm": 0.5891238187964615, + "learning_rate": 7.663631965739089e-07, + "loss": 0.0026, + "step": 9228 + }, + { + "epoch": 6.638374393094767, + "grad_norm": 1.0309529069953904, + "learning_rate": 7.660664429207692e-07, + "loss": 0.0046, + "step": 9229 + }, + { + "epoch": 6.639093688185578, + "grad_norm": 1.0028216228461202, + "learning_rate": 7.657697270297996e-07, + "loss": 0.0053, + "step": 9230 + }, + { + "epoch": 6.639812983276389, + "grad_norm": 2.825765297999436, + "learning_rate": 7.654730489162665e-07, + "loss": 0.0452, + "step": 9231 + }, + { + "epoch": 6.6405322783672, + "grad_norm": 1.9086415407362105, + "learning_rate": 7.651764085954338e-07, + "loss": 0.0117, + "step": 9232 + }, + { + "epoch": 6.641251573458011, + "grad_norm": 5.713526185958292, + "learning_rate": 7.648798060825655e-07, + "loss": 0.0728, + "step": 9233 + }, + { + "epoch": 6.641970868548822, + "grad_norm": 2.7185819623718834, + "learning_rate": 7.645832413929215e-07, + "loss": 0.0142, + "step": 9234 + }, + { + "epoch": 6.642690163639633, + "grad_norm": 3.3996485862296746, + "learning_rate": 7.642867145417607e-07, + "loss": 0.0413, + "step": 9235 + }, + { + "epoch": 6.643409458730444, + "grad_norm": 0.1574302340325019, + "learning_rate": 7.639902255443401e-07, + "loss": 0.0007, + "step": 9236 + }, + { + "epoch": 6.6441287538212555, + "grad_norm": 5.146737961477, + "learning_rate": 7.636937744159143e-07, + "loss": 0.0417, + "step": 9237 + }, + { + "epoch": 6.644848048912066, + "grad_norm": 0.0960424557780923, + "learning_rate": 7.633973611717363e-07, + "loss": 0.0001, + "step": 9238 + }, + { + "epoch": 6.6455673440028775, + "grad_norm": 0.40216217931734444, + "learning_rate": 7.631009858270572e-07, + "loss": 0.0018, + "step": 9239 + }, + { + "epoch": 6.646286639093688, + "grad_norm": 5.905324334961775, + "learning_rate": 7.628046483971262e-07, + "loss": 0.1209, + "step": 9240 + }, + { + "epoch": 6.647005934184499, + "grad_norm": 0.06522250109637673, + "learning_rate": 7.625083488971899e-07, + "loss": 0.0002, + "step": 9241 + }, + { + "epoch": 6.64772522927531, + "grad_norm": 0.025625530174102323, + "learning_rate": 7.622120873424936e-07, + "loss": 0.0001, + "step": 9242 + }, + { + "epoch": 6.648444524366122, + "grad_norm": 2.2386531342420626, + "learning_rate": 7.6191586374828e-07, + "loss": 0.0093, + "step": 9243 + }, + { + "epoch": 6.649163819456932, + "grad_norm": 4.398638423959456, + "learning_rate": 7.616196781297918e-07, + "loss": 0.0538, + "step": 9244 + }, + { + "epoch": 6.649883114547743, + "grad_norm": 2.3221308589907768, + "learning_rate": 7.61323530502267e-07, + "loss": 0.0325, + "step": 9245 + }, + { + "epoch": 6.650602409638554, + "grad_norm": 1.896467011638651, + "learning_rate": 7.610274208809433e-07, + "loss": 0.0145, + "step": 9246 + }, + { + "epoch": 6.651321704729365, + "grad_norm": 1.7244694277007888, + "learning_rate": 7.607313492810559e-07, + "loss": 0.0187, + "step": 9247 + }, + { + "epoch": 6.652040999820176, + "grad_norm": 2.3617015376496098, + "learning_rate": 7.60435315717838e-07, + "loss": 0.0235, + "step": 9248 + }, + { + "epoch": 6.652760294910987, + "grad_norm": 0.4379451553652231, + "learning_rate": 7.601393202065218e-07, + "loss": 0.0031, + "step": 9249 + }, + { + "epoch": 6.6534795900017984, + "grad_norm": 0.22718821637992365, + "learning_rate": 7.598433627623365e-07, + "loss": 0.0003, + "step": 9250 + }, + { + "epoch": 6.654198885092609, + "grad_norm": 5.517635034536187, + "learning_rate": 7.595474434005094e-07, + "loss": 0.0606, + "step": 9251 + }, + { + "epoch": 6.6549181801834205, + "grad_norm": 6.373603245531257, + "learning_rate": 7.592515621362661e-07, + "loss": 0.0713, + "step": 9252 + }, + { + "epoch": 6.655637475274231, + "grad_norm": 0.6993899534972993, + "learning_rate": 7.589557189848304e-07, + "loss": 0.0049, + "step": 9253 + }, + { + "epoch": 6.656356770365043, + "grad_norm": 0.6167609797959132, + "learning_rate": 7.586599139614237e-07, + "loss": 0.0012, + "step": 9254 + }, + { + "epoch": 6.657076065455853, + "grad_norm": 1.872083290735303, + "learning_rate": 7.583641470812659e-07, + "loss": 0.0106, + "step": 9255 + }, + { + "epoch": 6.657795360546665, + "grad_norm": 0.02490169908951131, + "learning_rate": 7.580684183595746e-07, + "loss": 0.0001, + "step": 9256 + }, + { + "epoch": 6.658514655637475, + "grad_norm": 0.28840168108170083, + "learning_rate": 7.577727278115656e-07, + "loss": 0.0022, + "step": 9257 + }, + { + "epoch": 6.659233950728286, + "grad_norm": 3.412956396746199, + "learning_rate": 7.574770754524521e-07, + "loss": 0.0265, + "step": 9258 + }, + { + "epoch": 6.659953245819097, + "grad_norm": 2.381591860263674, + "learning_rate": 7.571814612974472e-07, + "loss": 0.0085, + "step": 9259 + }, + { + "epoch": 6.660672540909908, + "grad_norm": 2.855636087880378, + "learning_rate": 7.568858853617599e-07, + "loss": 0.035, + "step": 9260 + }, + { + "epoch": 6.661391836000719, + "grad_norm": 1.8055210802033927, + "learning_rate": 7.565903476605984e-07, + "loss": 0.009, + "step": 9261 + }, + { + "epoch": 6.66211113109153, + "grad_norm": 0.3604263120246064, + "learning_rate": 7.562948482091686e-07, + "loss": 0.001, + "step": 9262 + }, + { + "epoch": 6.6628304261823414, + "grad_norm": 2.2735896485455838, + "learning_rate": 7.559993870226742e-07, + "loss": 0.0225, + "step": 9263 + }, + { + "epoch": 6.663549721273152, + "grad_norm": 4.054764729257557, + "learning_rate": 7.557039641163169e-07, + "loss": 0.0226, + "step": 9264 + }, + { + "epoch": 6.6642690163639635, + "grad_norm": 4.66191974492267, + "learning_rate": 7.554085795052979e-07, + "loss": 0.0334, + "step": 9265 + }, + { + "epoch": 6.664988311454774, + "grad_norm": 5.1161060779089444, + "learning_rate": 7.551132332048144e-07, + "loss": 0.0432, + "step": 9266 + }, + { + "epoch": 6.665707606545586, + "grad_norm": 2.762227134825095, + "learning_rate": 7.548179252300628e-07, + "loss": 0.0365, + "step": 9267 + }, + { + "epoch": 6.666426901636396, + "grad_norm": 0.1917402270053323, + "learning_rate": 7.545226555962376e-07, + "loss": 0.0011, + "step": 9268 + }, + { + "epoch": 6.667146196727208, + "grad_norm": 4.383706687654783, + "learning_rate": 7.542274243185294e-07, + "loss": 0.0171, + "step": 9269 + }, + { + "epoch": 6.667865491818018, + "grad_norm": 5.929178672596262, + "learning_rate": 7.539322314121298e-07, + "loss": 0.0617, + "step": 9270 + }, + { + "epoch": 6.66858478690883, + "grad_norm": 1.91152096526662, + "learning_rate": 7.536370768922266e-07, + "loss": 0.0232, + "step": 9271 + }, + { + "epoch": 6.66930408199964, + "grad_norm": 3.57138456201653, + "learning_rate": 7.533419607740059e-07, + "loss": 0.0404, + "step": 9272 + }, + { + "epoch": 6.670023377090452, + "grad_norm": 1.3684523563873565, + "learning_rate": 7.530468830726521e-07, + "loss": 0.0299, + "step": 9273 + }, + { + "epoch": 6.670742672181262, + "grad_norm": 3.707834330455445, + "learning_rate": 7.527518438033469e-07, + "loss": 0.0127, + "step": 9274 + }, + { + "epoch": 6.671461967272073, + "grad_norm": 0.4606069814144959, + "learning_rate": 7.524568429812717e-07, + "loss": 0.0012, + "step": 9275 + }, + { + "epoch": 6.6721812623628844, + "grad_norm": 4.000293259590899, + "learning_rate": 7.52161880621604e-07, + "loss": 0.0784, + "step": 9276 + }, + { + "epoch": 6.672900557453695, + "grad_norm": 3.8821066573071326, + "learning_rate": 7.518669567395205e-07, + "loss": 0.037, + "step": 9277 + }, + { + "epoch": 6.6736198525445065, + "grad_norm": 0.4050982929718099, + "learning_rate": 7.515720713501953e-07, + "loss": 0.0012, + "step": 9278 + }, + { + "epoch": 6.674339147635317, + "grad_norm": 0.15702945059757664, + "learning_rate": 7.512772244688002e-07, + "loss": 0.0001, + "step": 9279 + }, + { + "epoch": 6.675058442726129, + "grad_norm": 0.010881879744723849, + "learning_rate": 7.509824161105076e-07, + "loss": 0.0001, + "step": 9280 + }, + { + "epoch": 6.675777737816939, + "grad_norm": 4.537493035185573, + "learning_rate": 7.50687646290484e-07, + "loss": 0.0231, + "step": 9281 + }, + { + "epoch": 6.676497032907751, + "grad_norm": 4.750805664649021, + "learning_rate": 7.503929150238961e-07, + "loss": 0.0582, + "step": 9282 + }, + { + "epoch": 6.677216327998561, + "grad_norm": 1.9588017851796558, + "learning_rate": 7.500982223259088e-07, + "loss": 0.0401, + "step": 9283 + }, + { + "epoch": 6.677935623089373, + "grad_norm": 1.4466331370940126, + "learning_rate": 7.498035682116844e-07, + "loss": 0.0109, + "step": 9284 + }, + { + "epoch": 6.678654918180183, + "grad_norm": 0.04795264254829809, + "learning_rate": 7.495089526963827e-07, + "loss": 0.0001, + "step": 9285 + }, + { + "epoch": 6.679374213270995, + "grad_norm": 12.670770756641044, + "learning_rate": 7.492143757951634e-07, + "loss": 0.046, + "step": 9286 + }, + { + "epoch": 6.680093508361805, + "grad_norm": 5.835191453600034, + "learning_rate": 7.489198375231824e-07, + "loss": 0.0436, + "step": 9287 + }, + { + "epoch": 6.680812803452616, + "grad_norm": 3.3141604577256265, + "learning_rate": 7.486253378955942e-07, + "loss": 0.0226, + "step": 9288 + }, + { + "epoch": 6.6815320985434274, + "grad_norm": 1.7428361361186078, + "learning_rate": 7.483308769275515e-07, + "loss": 0.0233, + "step": 9289 + }, + { + "epoch": 6.682251393634239, + "grad_norm": 4.510797165339182, + "learning_rate": 7.48036454634204e-07, + "loss": 0.0887, + "step": 9290 + }, + { + "epoch": 6.6829706887250495, + "grad_norm": 0.21774690703429317, + "learning_rate": 7.477420710307017e-07, + "loss": 0.0019, + "step": 9291 + }, + { + "epoch": 6.68368998381586, + "grad_norm": 11.931935906493692, + "learning_rate": 7.474477261321902e-07, + "loss": 0.0752, + "step": 9292 + }, + { + "epoch": 6.684409278906672, + "grad_norm": 1.4135191014791237, + "learning_rate": 7.471534199538143e-07, + "loss": 0.0151, + "step": 9293 + }, + { + "epoch": 6.685128573997482, + "grad_norm": 0.022883092318833113, + "learning_rate": 7.468591525107166e-07, + "loss": 0.0001, + "step": 9294 + }, + { + "epoch": 6.685847869088294, + "grad_norm": 0.04134297311394368, + "learning_rate": 7.465649238180374e-07, + "loss": 0.0002, + "step": 9295 + }, + { + "epoch": 6.686567164179104, + "grad_norm": 4.056626006646951, + "learning_rate": 7.462707338909158e-07, + "loss": 0.0574, + "step": 9296 + }, + { + "epoch": 6.687286459269916, + "grad_norm": 1.3182808532599992, + "learning_rate": 7.459765827444878e-07, + "loss": 0.0065, + "step": 9297 + }, + { + "epoch": 6.688005754360726, + "grad_norm": 3.424774020550793, + "learning_rate": 7.456824703938883e-07, + "loss": 0.0248, + "step": 9298 + }, + { + "epoch": 6.688725049451538, + "grad_norm": 0.4913841822060778, + "learning_rate": 7.4538839685425e-07, + "loss": 0.0019, + "step": 9299 + }, + { + "epoch": 6.689444344542348, + "grad_norm": 1.3367428743808711, + "learning_rate": 7.450943621407026e-07, + "loss": 0.0122, + "step": 9300 + }, + { + "epoch": 6.69016363963316, + "grad_norm": 2.6322449785473756, + "learning_rate": 7.448003662683763e-07, + "loss": 0.0081, + "step": 9301 + }, + { + "epoch": 6.69088293472397, + "grad_norm": 4.095802912537301, + "learning_rate": 7.445064092523966e-07, + "loss": 0.0178, + "step": 9302 + }, + { + "epoch": 6.691602229814782, + "grad_norm": 2.2290817746139266, + "learning_rate": 7.442124911078885e-07, + "loss": 0.0216, + "step": 9303 + }, + { + "epoch": 6.6923215249055925, + "grad_norm": 1.529087879105561, + "learning_rate": 7.439186118499745e-07, + "loss": 0.011, + "step": 9304 + }, + { + "epoch": 6.693040819996403, + "grad_norm": 7.1350785250810445, + "learning_rate": 7.436247714937751e-07, + "loss": 0.0978, + "step": 9305 + }, + { + "epoch": 6.693760115087215, + "grad_norm": 1.1806255418859448, + "learning_rate": 7.433309700544085e-07, + "loss": 0.0041, + "step": 9306 + }, + { + "epoch": 6.694479410178025, + "grad_norm": 0.016251873604279234, + "learning_rate": 7.430372075469923e-07, + "loss": 0.0001, + "step": 9307 + }, + { + "epoch": 6.695198705268837, + "grad_norm": 1.3243923072115749, + "learning_rate": 7.427434839866406e-07, + "loss": 0.0253, + "step": 9308 + }, + { + "epoch": 6.695918000359647, + "grad_norm": 5.479041383730956, + "learning_rate": 7.424497993884659e-07, + "loss": 0.0803, + "step": 9309 + }, + { + "epoch": 6.696637295450459, + "grad_norm": 1.161744582846034, + "learning_rate": 7.421561537675789e-07, + "loss": 0.0104, + "step": 9310 + }, + { + "epoch": 6.697356590541269, + "grad_norm": 4.280130751660438, + "learning_rate": 7.418625471390881e-07, + "loss": 0.0327, + "step": 9311 + }, + { + "epoch": 6.698075885632081, + "grad_norm": 2.8995309577054673, + "learning_rate": 7.415689795181002e-07, + "loss": 0.0272, + "step": 9312 + }, + { + "epoch": 6.698795180722891, + "grad_norm": 2.8449225011664834, + "learning_rate": 7.412754509197197e-07, + "loss": 0.019, + "step": 9313 + }, + { + "epoch": 6.699514475813703, + "grad_norm": 1.046673397224266, + "learning_rate": 7.409819613590491e-07, + "loss": 0.0123, + "step": 9314 + }, + { + "epoch": 6.700233770904513, + "grad_norm": 2.4394640470708744, + "learning_rate": 7.40688510851189e-07, + "loss": 0.0445, + "step": 9315 + }, + { + "epoch": 6.700953065995325, + "grad_norm": 7.487269457060803, + "learning_rate": 7.403950994112374e-07, + "loss": 0.0418, + "step": 9316 + }, + { + "epoch": 6.7016723610861355, + "grad_norm": 5.248625740120604, + "learning_rate": 7.401017270542919e-07, + "loss": 0.0293, + "step": 9317 + }, + { + "epoch": 6.702391656176946, + "grad_norm": 0.18764638452686874, + "learning_rate": 7.398083937954465e-07, + "loss": 0.0014, + "step": 9318 + }, + { + "epoch": 6.703110951267758, + "grad_norm": 2.018403837600847, + "learning_rate": 7.395150996497938e-07, + "loss": 0.0283, + "step": 9319 + }, + { + "epoch": 6.703830246358569, + "grad_norm": 6.5970274436695355, + "learning_rate": 7.392218446324241e-07, + "loss": 0.1167, + "step": 9320 + }, + { + "epoch": 6.70454954144938, + "grad_norm": 0.19060058193051965, + "learning_rate": 7.389286287584255e-07, + "loss": 0.0006, + "step": 9321 + }, + { + "epoch": 6.70526883654019, + "grad_norm": 0.8598279647487471, + "learning_rate": 7.386354520428857e-07, + "loss": 0.0052, + "step": 9322 + }, + { + "epoch": 6.705988131631002, + "grad_norm": 1.833249437935175, + "learning_rate": 7.383423145008882e-07, + "loss": 0.0138, + "step": 9323 + }, + { + "epoch": 6.706707426721812, + "grad_norm": 0.5595740823372187, + "learning_rate": 7.380492161475155e-07, + "loss": 0.0048, + "step": 9324 + }, + { + "epoch": 6.707426721812624, + "grad_norm": 2.1745665112050516, + "learning_rate": 7.377561569978486e-07, + "loss": 0.026, + "step": 9325 + }, + { + "epoch": 6.708146016903434, + "grad_norm": 1.8030080987157304, + "learning_rate": 7.374631370669653e-07, + "loss": 0.0161, + "step": 9326 + }, + { + "epoch": 6.708865311994246, + "grad_norm": 0.3317860017561948, + "learning_rate": 7.371701563699422e-07, + "loss": 0.0012, + "step": 9327 + }, + { + "epoch": 6.709584607085056, + "grad_norm": 0.6060571621020399, + "learning_rate": 7.368772149218535e-07, + "loss": 0.0053, + "step": 9328 + }, + { + "epoch": 6.710303902175868, + "grad_norm": 0.7408035942343373, + "learning_rate": 7.36584312737772e-07, + "loss": 0.0047, + "step": 9329 + }, + { + "epoch": 6.7110231972666785, + "grad_norm": 3.55637608898205, + "learning_rate": 7.362914498327676e-07, + "loss": 0.0553, + "step": 9330 + }, + { + "epoch": 6.71174249235749, + "grad_norm": 1.9651727002967563, + "learning_rate": 7.359986262219088e-07, + "loss": 0.0247, + "step": 9331 + }, + { + "epoch": 6.712461787448301, + "grad_norm": 4.019873266417743, + "learning_rate": 7.357058419202613e-07, + "loss": 0.0244, + "step": 9332 + }, + { + "epoch": 6.713181082539112, + "grad_norm": 0.7596981787203991, + "learning_rate": 7.354130969428903e-07, + "loss": 0.0025, + "step": 9333 + }, + { + "epoch": 6.713900377629923, + "grad_norm": 4.447813921295416, + "learning_rate": 7.351203913048578e-07, + "loss": 0.0175, + "step": 9334 + }, + { + "epoch": 6.714619672720733, + "grad_norm": 5.3026621536265575, + "learning_rate": 7.348277250212238e-07, + "loss": 0.0708, + "step": 9335 + }, + { + "epoch": 6.715338967811545, + "grad_norm": 0.8013365209391727, + "learning_rate": 7.345350981070466e-07, + "loss": 0.0062, + "step": 9336 + }, + { + "epoch": 6.716058262902356, + "grad_norm": 2.6377886399391217, + "learning_rate": 7.342425105773818e-07, + "loss": 0.0174, + "step": 9337 + }, + { + "epoch": 6.716777557993167, + "grad_norm": 6.452312123278984, + "learning_rate": 7.339499624472847e-07, + "loss": 0.1268, + "step": 9338 + }, + { + "epoch": 6.717496853083977, + "grad_norm": 3.9346962237446137, + "learning_rate": 7.336574537318072e-07, + "loss": 0.0581, + "step": 9339 + }, + { + "epoch": 6.718216148174789, + "grad_norm": 3.6293729197100046, + "learning_rate": 7.333649844459985e-07, + "loss": 0.0553, + "step": 9340 + }, + { + "epoch": 6.718935443265599, + "grad_norm": 5.053751330890065, + "learning_rate": 7.330725546049071e-07, + "loss": 0.0874, + "step": 9341 + }, + { + "epoch": 6.719654738356411, + "grad_norm": 2.1578526438066703, + "learning_rate": 7.327801642235785e-07, + "loss": 0.0278, + "step": 9342 + }, + { + "epoch": 6.7203740334472215, + "grad_norm": 2.9620948068950543, + "learning_rate": 7.32487813317058e-07, + "loss": 0.0363, + "step": 9343 + }, + { + "epoch": 6.721093328538033, + "grad_norm": 2.247195140065091, + "learning_rate": 7.321955019003865e-07, + "loss": 0.0182, + "step": 9344 + }, + { + "epoch": 6.721812623628844, + "grad_norm": 0.6194243286948617, + "learning_rate": 7.319032299886045e-07, + "loss": 0.0049, + "step": 9345 + }, + { + "epoch": 6.722531918719655, + "grad_norm": 2.056635928527957, + "learning_rate": 7.316109975967496e-07, + "loss": 0.01, + "step": 9346 + }, + { + "epoch": 6.723251213810466, + "grad_norm": 0.04077116550585425, + "learning_rate": 7.313188047398579e-07, + "loss": 0.0001, + "step": 9347 + }, + { + "epoch": 6.723970508901277, + "grad_norm": 3.867708739468298, + "learning_rate": 7.310266514329623e-07, + "loss": 0.0342, + "step": 9348 + }, + { + "epoch": 6.724689803992088, + "grad_norm": 3.9131545703634845, + "learning_rate": 7.307345376910961e-07, + "loss": 0.0249, + "step": 9349 + }, + { + "epoch": 6.725409099082899, + "grad_norm": 0.041786669932742056, + "learning_rate": 7.304424635292881e-07, + "loss": 0.0001, + "step": 9350 + }, + { + "epoch": 6.72612839417371, + "grad_norm": 0.04552444885912063, + "learning_rate": 7.301504289625663e-07, + "loss": 0.0001, + "step": 9351 + }, + { + "epoch": 6.72684768926452, + "grad_norm": 2.857317015085587, + "learning_rate": 7.298584340059565e-07, + "loss": 0.0287, + "step": 9352 + }, + { + "epoch": 6.727566984355332, + "grad_norm": 0.8059133220174901, + "learning_rate": 7.295664786744822e-07, + "loss": 0.0054, + "step": 9353 + }, + { + "epoch": 6.728286279446142, + "grad_norm": 0.11491876421033155, + "learning_rate": 7.292745629831651e-07, + "loss": 0.0005, + "step": 9354 + }, + { + "epoch": 6.729005574536954, + "grad_norm": 0.006246010270944295, + "learning_rate": 7.289826869470246e-07, + "loss": 0.0, + "step": 9355 + }, + { + "epoch": 6.7297248696277645, + "grad_norm": 2.3415009186685913, + "learning_rate": 7.286908505810783e-07, + "loss": 0.0281, + "step": 9356 + }, + { + "epoch": 6.730444164718576, + "grad_norm": 2.044832462269974, + "learning_rate": 7.283990539003418e-07, + "loss": 0.009, + "step": 9357 + }, + { + "epoch": 6.731163459809387, + "grad_norm": 0.22770260476865353, + "learning_rate": 7.281072969198278e-07, + "loss": 0.0003, + "step": 9358 + }, + { + "epoch": 6.731882754900198, + "grad_norm": 2.9780049153879595, + "learning_rate": 7.278155796545491e-07, + "loss": 0.0346, + "step": 9359 + }, + { + "epoch": 6.732602049991009, + "grad_norm": 0.9630451036642091, + "learning_rate": 7.275239021195143e-07, + "loss": 0.0038, + "step": 9360 + }, + { + "epoch": 6.73332134508182, + "grad_norm": 2.248754803487241, + "learning_rate": 7.272322643297308e-07, + "loss": 0.0069, + "step": 9361 + }, + { + "epoch": 6.734040640172631, + "grad_norm": 0.014061456433342771, + "learning_rate": 7.269406663002036e-07, + "loss": 0.0, + "step": 9362 + }, + { + "epoch": 6.734759935263442, + "grad_norm": 2.2915956211815627, + "learning_rate": 7.26649108045936e-07, + "loss": 0.0301, + "step": 9363 + }, + { + "epoch": 6.735479230354253, + "grad_norm": 3.765037633142931, + "learning_rate": 7.263575895819297e-07, + "loss": 0.0611, + "step": 9364 + }, + { + "epoch": 6.736198525445063, + "grad_norm": 4.600622016510284, + "learning_rate": 7.260661109231836e-07, + "loss": 0.0579, + "step": 9365 + }, + { + "epoch": 6.736917820535875, + "grad_norm": 1.4848719926005178, + "learning_rate": 7.257746720846946e-07, + "loss": 0.0118, + "step": 9366 + }, + { + "epoch": 6.737637115626686, + "grad_norm": 2.1109314440876243, + "learning_rate": 7.25483273081458e-07, + "loss": 0.0085, + "step": 9367 + }, + { + "epoch": 6.738356410717497, + "grad_norm": 5.517811248319581, + "learning_rate": 7.251919139284664e-07, + "loss": 0.0459, + "step": 9368 + }, + { + "epoch": 6.7390757058083075, + "grad_norm": 3.122570095780809, + "learning_rate": 7.249005946407112e-07, + "loss": 0.0064, + "step": 9369 + }, + { + "epoch": 6.739795000899119, + "grad_norm": 3.1571844806650136, + "learning_rate": 7.246093152331808e-07, + "loss": 0.0247, + "step": 9370 + }, + { + "epoch": 6.74051429598993, + "grad_norm": 1.561634828802724, + "learning_rate": 7.243180757208623e-07, + "loss": 0.0216, + "step": 9371 + }, + { + "epoch": 6.741233591080741, + "grad_norm": 0.37846121113712106, + "learning_rate": 7.240268761187406e-07, + "loss": 0.0012, + "step": 9372 + }, + { + "epoch": 6.741952886171552, + "grad_norm": 1.8840044450549067, + "learning_rate": 7.237357164417983e-07, + "loss": 0.0149, + "step": 9373 + }, + { + "epoch": 6.742672181262363, + "grad_norm": 6.248529838791128, + "learning_rate": 7.234445967050155e-07, + "loss": 0.0927, + "step": 9374 + }, + { + "epoch": 6.743391476353174, + "grad_norm": 3.539397027458527, + "learning_rate": 7.23153516923372e-07, + "loss": 0.0175, + "step": 9375 + }, + { + "epoch": 6.744110771443985, + "grad_norm": 0.01579547600482638, + "learning_rate": 7.228624771118438e-07, + "loss": 0.0, + "step": 9376 + }, + { + "epoch": 6.744830066534796, + "grad_norm": 4.262804643956493, + "learning_rate": 7.225714772854052e-07, + "loss": 0.0255, + "step": 9377 + }, + { + "epoch": 6.745549361625607, + "grad_norm": 3.815066219305703, + "learning_rate": 7.222805174590289e-07, + "loss": 0.0559, + "step": 9378 + }, + { + "epoch": 6.746268656716418, + "grad_norm": 5.950945102386406, + "learning_rate": 7.219895976476846e-07, + "loss": 0.069, + "step": 9379 + }, + { + "epoch": 6.746987951807229, + "grad_norm": 0.03916793716413172, + "learning_rate": 7.216987178663419e-07, + "loss": 0.0002, + "step": 9380 + }, + { + "epoch": 6.74770724689804, + "grad_norm": 1.9644194333893057, + "learning_rate": 7.214078781299664e-07, + "loss": 0.0099, + "step": 9381 + }, + { + "epoch": 6.7484265419888505, + "grad_norm": 0.7088639676195897, + "learning_rate": 7.211170784535224e-07, + "loss": 0.0058, + "step": 9382 + }, + { + "epoch": 6.749145837079662, + "grad_norm": 0.04337170392726917, + "learning_rate": 7.20826318851972e-07, + "loss": 0.0002, + "step": 9383 + }, + { + "epoch": 6.749865132170473, + "grad_norm": 0.8852390127564654, + "learning_rate": 7.205355993402753e-07, + "loss": 0.0037, + "step": 9384 + }, + { + "epoch": 6.750584427261284, + "grad_norm": 0.7029820116615056, + "learning_rate": 7.202449199333903e-07, + "loss": 0.0053, + "step": 9385 + }, + { + "epoch": 6.751303722352095, + "grad_norm": 3.546449336943123, + "learning_rate": 7.19954280646273e-07, + "loss": 0.0325, + "step": 9386 + }, + { + "epoch": 6.752023017442906, + "grad_norm": 11.890362245344578, + "learning_rate": 7.196636814938772e-07, + "loss": 0.2338, + "step": 9387 + }, + { + "epoch": 6.752742312533717, + "grad_norm": 1.646446999759009, + "learning_rate": 7.193731224911549e-07, + "loss": 0.0083, + "step": 9388 + }, + { + "epoch": 6.753461607624528, + "grad_norm": 2.183176126150383, + "learning_rate": 7.190826036530556e-07, + "loss": 0.0317, + "step": 9389 + }, + { + "epoch": 6.754180902715339, + "grad_norm": 0.751828219577722, + "learning_rate": 7.187921249945269e-07, + "loss": 0.0053, + "step": 9390 + }, + { + "epoch": 6.75490019780615, + "grad_norm": 1.3099259452399739, + "learning_rate": 7.185016865305152e-07, + "loss": 0.004, + "step": 9391 + }, + { + "epoch": 6.755619492896961, + "grad_norm": 1.5017502772394196, + "learning_rate": 7.182112882759637e-07, + "loss": 0.0068, + "step": 9392 + }, + { + "epoch": 6.756338787987772, + "grad_norm": 0.58731972598284, + "learning_rate": 7.179209302458136e-07, + "loss": 0.0011, + "step": 9393 + }, + { + "epoch": 6.757058083078583, + "grad_norm": 11.427512903135618, + "learning_rate": 7.176306124550047e-07, + "loss": 0.081, + "step": 9394 + }, + { + "epoch": 6.7577773781693935, + "grad_norm": 0.06366123726731962, + "learning_rate": 7.173403349184735e-07, + "loss": 0.0002, + "step": 9395 + }, + { + "epoch": 6.758496673260205, + "grad_norm": 1.9291416789219265, + "learning_rate": 7.170500976511567e-07, + "loss": 0.036, + "step": 9396 + }, + { + "epoch": 6.7592159683510165, + "grad_norm": 3.0662774073423433, + "learning_rate": 7.167599006679868e-07, + "loss": 0.0337, + "step": 9397 + }, + { + "epoch": 6.759935263441827, + "grad_norm": 1.3581851771281874, + "learning_rate": 7.164697439838953e-07, + "loss": 0.0127, + "step": 9398 + }, + { + "epoch": 6.760654558532638, + "grad_norm": 4.5882455398741655, + "learning_rate": 7.161796276138106e-07, + "loss": 0.0817, + "step": 9399 + }, + { + "epoch": 6.761373853623449, + "grad_norm": 2.37429675740209, + "learning_rate": 7.158895515726593e-07, + "loss": 0.0216, + "step": 9400 + }, + { + "epoch": 6.76209314871426, + "grad_norm": 0.43570644420295107, + "learning_rate": 7.155995158753676e-07, + "loss": 0.006, + "step": 9401 + }, + { + "epoch": 6.762812443805071, + "grad_norm": 3.4337232514824274, + "learning_rate": 7.153095205368579e-07, + "loss": 0.0248, + "step": 9402 + }, + { + "epoch": 6.763531738895882, + "grad_norm": 2.352694339287455, + "learning_rate": 7.150195655720508e-07, + "loss": 0.0182, + "step": 9403 + }, + { + "epoch": 6.764251033986693, + "grad_norm": 1.9615738091404755, + "learning_rate": 7.14729650995865e-07, + "loss": 0.027, + "step": 9404 + }, + { + "epoch": 6.764970329077504, + "grad_norm": 6.131426254145296, + "learning_rate": 7.144397768232165e-07, + "loss": 0.0777, + "step": 9405 + }, + { + "epoch": 6.765689624168315, + "grad_norm": 3.6488440452040765, + "learning_rate": 7.141499430690212e-07, + "loss": 0.0385, + "step": 9406 + }, + { + "epoch": 6.766408919259126, + "grad_norm": 1.3781798389425872, + "learning_rate": 7.138601497481908e-07, + "loss": 0.0135, + "step": 9407 + }, + { + "epoch": 6.767128214349937, + "grad_norm": 3.045649499517548, + "learning_rate": 7.135703968756357e-07, + "loss": 0.0288, + "step": 9408 + }, + { + "epoch": 6.767847509440748, + "grad_norm": 1.7570373749880486, + "learning_rate": 7.132806844662643e-07, + "loss": 0.0132, + "step": 9409 + }, + { + "epoch": 6.7685668045315595, + "grad_norm": 4.860557117113826, + "learning_rate": 7.129910125349825e-07, + "loss": 0.0699, + "step": 9410 + }, + { + "epoch": 6.76928609962237, + "grad_norm": 1.4369368880996662, + "learning_rate": 7.12701381096695e-07, + "loss": 0.0042, + "step": 9411 + }, + { + "epoch": 6.770005394713181, + "grad_norm": 2.934246243052716, + "learning_rate": 7.124117901663032e-07, + "loss": 0.0175, + "step": 9412 + }, + { + "epoch": 6.770724689803992, + "grad_norm": 3.0301673412914383, + "learning_rate": 7.121222397587074e-07, + "loss": 0.0365, + "step": 9413 + }, + { + "epoch": 6.771443984894804, + "grad_norm": 2.4136493804967096, + "learning_rate": 7.118327298888056e-07, + "loss": 0.0154, + "step": 9414 + }, + { + "epoch": 6.772163279985614, + "grad_norm": 5.842995853972884, + "learning_rate": 7.115432605714934e-07, + "loss": 0.0531, + "step": 9415 + }, + { + "epoch": 6.772882575076425, + "grad_norm": 1.024534351096606, + "learning_rate": 7.112538318216639e-07, + "loss": 0.0053, + "step": 9416 + }, + { + "epoch": 6.773601870167236, + "grad_norm": 2.018440855604963, + "learning_rate": 7.1096444365421e-07, + "loss": 0.0245, + "step": 9417 + }, + { + "epoch": 6.774321165258047, + "grad_norm": 3.5877216911955525, + "learning_rate": 7.106750960840206e-07, + "loss": 0.0543, + "step": 9418 + }, + { + "epoch": 6.775040460348858, + "grad_norm": 0.025383327980933627, + "learning_rate": 7.10385789125983e-07, + "loss": 0.0001, + "step": 9419 + }, + { + "epoch": 6.775759755439669, + "grad_norm": 4.82801750020902, + "learning_rate": 7.100965227949826e-07, + "loss": 0.0666, + "step": 9420 + }, + { + "epoch": 6.77647905053048, + "grad_norm": 1.120797256214227, + "learning_rate": 7.098072971059023e-07, + "loss": 0.0054, + "step": 9421 + }, + { + "epoch": 6.777198345621291, + "grad_norm": 4.829016411484128, + "learning_rate": 7.095181120736244e-07, + "loss": 0.0797, + "step": 9422 + }, + { + "epoch": 6.7779176407121025, + "grad_norm": 2.1537575856084357, + "learning_rate": 7.092289677130271e-07, + "loss": 0.0291, + "step": 9423 + }, + { + "epoch": 6.778636935802913, + "grad_norm": 4.439588179644254, + "learning_rate": 7.089398640389875e-07, + "loss": 0.0662, + "step": 9424 + }, + { + "epoch": 6.7793562308937245, + "grad_norm": 2.497204017797739, + "learning_rate": 7.086508010663808e-07, + "loss": 0.0233, + "step": 9425 + }, + { + "epoch": 6.780075525984535, + "grad_norm": 3.353935350557472, + "learning_rate": 7.083617788100795e-07, + "loss": 0.0532, + "step": 9426 + }, + { + "epoch": 6.780794821075347, + "grad_norm": 6.628124705202478, + "learning_rate": 7.080727972849541e-07, + "loss": 0.0995, + "step": 9427 + }, + { + "epoch": 6.781514116166157, + "grad_norm": 3.1223091384352126, + "learning_rate": 7.077838565058737e-07, + "loss": 0.0114, + "step": 9428 + }, + { + "epoch": 6.782233411256968, + "grad_norm": 0.05222877577771055, + "learning_rate": 7.074949564877045e-07, + "loss": 0.0003, + "step": 9429 + }, + { + "epoch": 6.782952706347779, + "grad_norm": 2.24523980534879, + "learning_rate": 7.072060972453111e-07, + "loss": 0.0157, + "step": 9430 + }, + { + "epoch": 6.78367200143859, + "grad_norm": 0.06002401666728625, + "learning_rate": 7.069172787935548e-07, + "loss": 0.0002, + "step": 9431 + }, + { + "epoch": 6.784391296529401, + "grad_norm": 0.06782892940390409, + "learning_rate": 7.066285011472977e-07, + "loss": 0.0001, + "step": 9432 + }, + { + "epoch": 6.785110591620212, + "grad_norm": 0.057145276703741706, + "learning_rate": 7.063397643213968e-07, + "loss": 0.0004, + "step": 9433 + }, + { + "epoch": 6.785829886711023, + "grad_norm": 2.8171169180217044, + "learning_rate": 7.060510683307081e-07, + "loss": 0.0311, + "step": 9434 + }, + { + "epoch": 6.786549181801834, + "grad_norm": 2.55041092228719, + "learning_rate": 7.057624131900858e-07, + "loss": 0.0261, + "step": 9435 + }, + { + "epoch": 6.7872684768926455, + "grad_norm": 2.1382672895014823, + "learning_rate": 7.054737989143816e-07, + "loss": 0.0169, + "step": 9436 + }, + { + "epoch": 6.787987771983456, + "grad_norm": 2.7812247666646956, + "learning_rate": 7.051852255184446e-07, + "loss": 0.0181, + "step": 9437 + }, + { + "epoch": 6.7887070670742675, + "grad_norm": 2.452080266612999, + "learning_rate": 7.048966930171235e-07, + "loss": 0.0211, + "step": 9438 + }, + { + "epoch": 6.789426362165078, + "grad_norm": 2.4355984367455457, + "learning_rate": 7.046082014252635e-07, + "loss": 0.0201, + "step": 9439 + }, + { + "epoch": 6.79014565725589, + "grad_norm": 2.6342903256075374, + "learning_rate": 7.043197507577075e-07, + "loss": 0.0171, + "step": 9440 + }, + { + "epoch": 6.7908649523467, + "grad_norm": 1.7133249045329155, + "learning_rate": 7.040313410292972e-07, + "loss": 0.0259, + "step": 9441 + }, + { + "epoch": 6.791584247437511, + "grad_norm": 4.317041008921396, + "learning_rate": 7.037429722548717e-07, + "loss": 0.0408, + "step": 9442 + }, + { + "epoch": 6.792303542528322, + "grad_norm": 2.5584654426525097, + "learning_rate": 7.034546444492682e-07, + "loss": 0.0239, + "step": 9443 + }, + { + "epoch": 6.793022837619134, + "grad_norm": 0.8645566425849917, + "learning_rate": 7.031663576273215e-07, + "loss": 0.003, + "step": 9444 + }, + { + "epoch": 6.793742132709944, + "grad_norm": 2.072462389735559, + "learning_rate": 7.028781118038644e-07, + "loss": 0.0278, + "step": 9445 + }, + { + "epoch": 6.794461427800755, + "grad_norm": 3.9578504575355393, + "learning_rate": 7.025899069937278e-07, + "loss": 0.0333, + "step": 9446 + }, + { + "epoch": 6.795180722891566, + "grad_norm": 3.982429151442624, + "learning_rate": 7.023017432117397e-07, + "loss": 0.04, + "step": 9447 + }, + { + "epoch": 6.795900017982377, + "grad_norm": 0.7542057989541002, + "learning_rate": 7.020136204727278e-07, + "loss": 0.0043, + "step": 9448 + }, + { + "epoch": 6.7966193130731885, + "grad_norm": 2.693552603050783, + "learning_rate": 7.01725538791516e-07, + "loss": 0.0129, + "step": 9449 + }, + { + "epoch": 6.797338608163999, + "grad_norm": 1.4848418979523297, + "learning_rate": 7.014374981829264e-07, + "loss": 0.0232, + "step": 9450 + }, + { + "epoch": 6.7980579032548105, + "grad_norm": 2.3338386731459413, + "learning_rate": 7.011494986617795e-07, + "loss": 0.0254, + "step": 9451 + }, + { + "epoch": 6.798777198345621, + "grad_norm": 2.2182688656573926, + "learning_rate": 7.008615402428926e-07, + "loss": 0.0063, + "step": 9452 + }, + { + "epoch": 6.799496493436433, + "grad_norm": 0.042405753630824404, + "learning_rate": 7.005736229410829e-07, + "loss": 0.0001, + "step": 9453 + }, + { + "epoch": 6.800215788527243, + "grad_norm": 0.09248860067657824, + "learning_rate": 7.002857467711636e-07, + "loss": 0.0003, + "step": 9454 + }, + { + "epoch": 6.800935083618055, + "grad_norm": 2.4382381627570857, + "learning_rate": 6.999979117479466e-07, + "loss": 0.0334, + "step": 9455 + }, + { + "epoch": 6.801654378708865, + "grad_norm": 5.70710993391149, + "learning_rate": 6.997101178862413e-07, + "loss": 0.0436, + "step": 9456 + }, + { + "epoch": 6.802373673799677, + "grad_norm": 0.044254021210479104, + "learning_rate": 6.994223652008559e-07, + "loss": 0.0001, + "step": 9457 + }, + { + "epoch": 6.803092968890487, + "grad_norm": 0.2070623605786642, + "learning_rate": 6.991346537065941e-07, + "loss": 0.0005, + "step": 9458 + }, + { + "epoch": 6.803812263981298, + "grad_norm": 0.19838497822579995, + "learning_rate": 6.988469834182606e-07, + "loss": 0.0011, + "step": 9459 + }, + { + "epoch": 6.804531559072109, + "grad_norm": 5.910301521026819, + "learning_rate": 6.985593543506564e-07, + "loss": 0.0384, + "step": 9460 + }, + { + "epoch": 6.80525085416292, + "grad_norm": 0.09088116187387213, + "learning_rate": 6.982717665185803e-07, + "loss": 0.0003, + "step": 9461 + }, + { + "epoch": 6.8059701492537314, + "grad_norm": 0.2576845019642027, + "learning_rate": 6.979842199368291e-07, + "loss": 0.001, + "step": 9462 + }, + { + "epoch": 6.806689444344542, + "grad_norm": 2.1982664049161964, + "learning_rate": 6.97696714620197e-07, + "loss": 0.0229, + "step": 9463 + }, + { + "epoch": 6.8074087394353535, + "grad_norm": 3.740140156635895, + "learning_rate": 6.974092505834781e-07, + "loss": 0.0474, + "step": 9464 + }, + { + "epoch": 6.808128034526164, + "grad_norm": 0.14656309359276892, + "learning_rate": 6.971218278414622e-07, + "loss": 0.0005, + "step": 9465 + }, + { + "epoch": 6.808847329616976, + "grad_norm": 2.4249508030527966, + "learning_rate": 6.968344464089374e-07, + "loss": 0.0187, + "step": 9466 + }, + { + "epoch": 6.809566624707786, + "grad_norm": 0.9707932455336749, + "learning_rate": 6.965471063006906e-07, + "loss": 0.0041, + "step": 9467 + }, + { + "epoch": 6.810285919798598, + "grad_norm": 0.06927555215200416, + "learning_rate": 6.962598075315047e-07, + "loss": 0.0002, + "step": 9468 + }, + { + "epoch": 6.811005214889408, + "grad_norm": 0.11200895252036543, + "learning_rate": 6.959725501161633e-07, + "loss": 0.0006, + "step": 9469 + }, + { + "epoch": 6.81172450998022, + "grad_norm": 1.237716378581375, + "learning_rate": 6.956853340694463e-07, + "loss": 0.0126, + "step": 9470 + }, + { + "epoch": 6.81244380507103, + "grad_norm": 4.332295226594724, + "learning_rate": 6.953981594061302e-07, + "loss": 0.0501, + "step": 9471 + }, + { + "epoch": 6.813163100161841, + "grad_norm": 3.720909421125435, + "learning_rate": 6.951110261409912e-07, + "loss": 0.0518, + "step": 9472 + }, + { + "epoch": 6.813882395252652, + "grad_norm": 0.10274662739060642, + "learning_rate": 6.948239342888021e-07, + "loss": 0.0001, + "step": 9473 + }, + { + "epoch": 6.814601690343464, + "grad_norm": 0.4840652443162129, + "learning_rate": 6.945368838643359e-07, + "loss": 0.0008, + "step": 9474 + }, + { + "epoch": 6.8153209854342744, + "grad_norm": 0.03209657180672942, + "learning_rate": 6.942498748823607e-07, + "loss": 0.0001, + "step": 9475 + }, + { + "epoch": 6.816040280525085, + "grad_norm": 3.484442044885247, + "learning_rate": 6.939629073576441e-07, + "loss": 0.0439, + "step": 9476 + }, + { + "epoch": 6.8167595756158965, + "grad_norm": 1.8013990950942431, + "learning_rate": 6.936759813049507e-07, + "loss": 0.0031, + "step": 9477 + }, + { + "epoch": 6.817478870706707, + "grad_norm": 2.409776652739454, + "learning_rate": 6.933890967390437e-07, + "loss": 0.0229, + "step": 9478 + }, + { + "epoch": 6.818198165797519, + "grad_norm": 1.8722464990592167, + "learning_rate": 6.931022536746828e-07, + "loss": 0.0273, + "step": 9479 + }, + { + "epoch": 6.818917460888329, + "grad_norm": 0.43873271948833475, + "learning_rate": 6.928154521266282e-07, + "loss": 0.0016, + "step": 9480 + }, + { + "epoch": 6.819636755979141, + "grad_norm": 2.30618775785641, + "learning_rate": 6.925286921096355e-07, + "loss": 0.0356, + "step": 9481 + }, + { + "epoch": 6.820356051069951, + "grad_norm": 1.6736941943666799, + "learning_rate": 6.922419736384589e-07, + "loss": 0.0199, + "step": 9482 + }, + { + "epoch": 6.821075346160763, + "grad_norm": 0.05142948412536048, + "learning_rate": 6.919552967278511e-07, + "loss": 0.0002, + "step": 9483 + }, + { + "epoch": 6.821794641251573, + "grad_norm": 3.368093593467535, + "learning_rate": 6.916686613925616e-07, + "loss": 0.0368, + "step": 9484 + }, + { + "epoch": 6.822513936342385, + "grad_norm": 3.6671947292989424, + "learning_rate": 6.913820676473384e-07, + "loss": 0.0078, + "step": 9485 + }, + { + "epoch": 6.823233231433195, + "grad_norm": 0.9311595041362651, + "learning_rate": 6.910955155069272e-07, + "loss": 0.0114, + "step": 9486 + }, + { + "epoch": 6.823952526524007, + "grad_norm": 4.824179772011184, + "learning_rate": 6.908090049860719e-07, + "loss": 0.1005, + "step": 9487 + }, + { + "epoch": 6.8246718216148174, + "grad_norm": 2.2074141897271553, + "learning_rate": 6.905225360995136e-07, + "loss": 0.0245, + "step": 9488 + }, + { + "epoch": 6.825391116705628, + "grad_norm": 1.661359277880932, + "learning_rate": 6.902361088619914e-07, + "loss": 0.0168, + "step": 9489 + }, + { + "epoch": 6.8261104117964395, + "grad_norm": 3.8578801500024196, + "learning_rate": 6.899497232882433e-07, + "loss": 0.0479, + "step": 9490 + }, + { + "epoch": 6.826829706887251, + "grad_norm": 1.4928380512249801, + "learning_rate": 6.896633793930042e-07, + "loss": 0.0175, + "step": 9491 + }, + { + "epoch": 6.827549001978062, + "grad_norm": 3.35009835151914, + "learning_rate": 6.893770771910065e-07, + "loss": 0.0546, + "step": 9492 + }, + { + "epoch": 6.828268297068872, + "grad_norm": 2.546839485543675, + "learning_rate": 6.890908166969812e-07, + "loss": 0.0184, + "step": 9493 + }, + { + "epoch": 6.828987592159684, + "grad_norm": 2.501530353871296, + "learning_rate": 6.888045979256565e-07, + "loss": 0.0168, + "step": 9494 + }, + { + "epoch": 6.829706887250494, + "grad_norm": 0.9796695855465727, + "learning_rate": 6.885184208917596e-07, + "loss": 0.007, + "step": 9495 + }, + { + "epoch": 6.830426182341306, + "grad_norm": 0.30489368947749473, + "learning_rate": 6.882322856100146e-07, + "loss": 0.0007, + "step": 9496 + }, + { + "epoch": 6.831145477432116, + "grad_norm": 2.227728116959272, + "learning_rate": 6.879461920951435e-07, + "loss": 0.015, + "step": 9497 + }, + { + "epoch": 6.831864772522928, + "grad_norm": 2.2328572729874145, + "learning_rate": 6.876601403618663e-07, + "loss": 0.0126, + "step": 9498 + }, + { + "epoch": 6.832584067613738, + "grad_norm": 5.183540785618823, + "learning_rate": 6.873741304249008e-07, + "loss": 0.0839, + "step": 9499 + }, + { + "epoch": 6.83330336270455, + "grad_norm": 0.10637601015362137, + "learning_rate": 6.870881622989629e-07, + "loss": 0.0003, + "step": 9500 + }, + { + "epoch": 6.8340226577953604, + "grad_norm": 1.2442898223385002, + "learning_rate": 6.868022359987661e-07, + "loss": 0.0127, + "step": 9501 + }, + { + "epoch": 6.834741952886172, + "grad_norm": 1.3361625181079408, + "learning_rate": 6.865163515390216e-07, + "loss": 0.0122, + "step": 9502 + }, + { + "epoch": 6.8354612479769825, + "grad_norm": 0.11821580957032561, + "learning_rate": 6.862305089344391e-07, + "loss": 0.0002, + "step": 9503 + }, + { + "epoch": 6.836180543067794, + "grad_norm": 0.9534147504164407, + "learning_rate": 6.859447081997252e-07, + "loss": 0.0057, + "step": 9504 + }, + { + "epoch": 6.836899838158605, + "grad_norm": 2.244467907051712, + "learning_rate": 6.856589493495848e-07, + "loss": 0.0224, + "step": 9505 + }, + { + "epoch": 6.837619133249415, + "grad_norm": 0.14900549100617136, + "learning_rate": 6.853732323987214e-07, + "loss": 0.0005, + "step": 9506 + }, + { + "epoch": 6.838338428340227, + "grad_norm": 3.174471405089122, + "learning_rate": 6.850875573618352e-07, + "loss": 0.0297, + "step": 9507 + }, + { + "epoch": 6.839057723431037, + "grad_norm": 1.636703072969135, + "learning_rate": 6.848019242536248e-07, + "loss": 0.0122, + "step": 9508 + }, + { + "epoch": 6.839777018521849, + "grad_norm": 0.011346553919882582, + "learning_rate": 6.845163330887863e-07, + "loss": 0.0, + "step": 9509 + }, + { + "epoch": 6.840496313612659, + "grad_norm": 0.004399115619544973, + "learning_rate": 6.842307838820136e-07, + "loss": 0.0, + "step": 9510 + }, + { + "epoch": 6.841215608703471, + "grad_norm": 1.7503737542542743, + "learning_rate": 6.839452766479996e-07, + "loss": 0.0237, + "step": 9511 + }, + { + "epoch": 6.841934903794281, + "grad_norm": 0.02102275189464944, + "learning_rate": 6.836598114014336e-07, + "loss": 0.0, + "step": 9512 + }, + { + "epoch": 6.842654198885093, + "grad_norm": 0.050366428411115716, + "learning_rate": 6.833743881570034e-07, + "loss": 0.0002, + "step": 9513 + }, + { + "epoch": 6.843373493975903, + "grad_norm": 2.0228478158443846, + "learning_rate": 6.830890069293944e-07, + "loss": 0.0085, + "step": 9514 + }, + { + "epoch": 6.844092789066715, + "grad_norm": 3.599710392890551, + "learning_rate": 6.828036677332902e-07, + "loss": 0.0104, + "step": 9515 + }, + { + "epoch": 6.8448120841575255, + "grad_norm": 3.0365100356009487, + "learning_rate": 6.825183705833717e-07, + "loss": 0.0499, + "step": 9516 + }, + { + "epoch": 6.845531379248337, + "grad_norm": 4.587813532713674, + "learning_rate": 6.82233115494318e-07, + "loss": 0.0517, + "step": 9517 + }, + { + "epoch": 6.846250674339148, + "grad_norm": 3.158419219616889, + "learning_rate": 6.819479024808062e-07, + "loss": 0.0242, + "step": 9518 + }, + { + "epoch": 6.846969969429958, + "grad_norm": 1.4595582452887963, + "learning_rate": 6.816627315575108e-07, + "loss": 0.0049, + "step": 9519 + }, + { + "epoch": 6.84768926452077, + "grad_norm": 1.0100461009374384, + "learning_rate": 6.813776027391044e-07, + "loss": 0.0018, + "step": 9520 + }, + { + "epoch": 6.848408559611581, + "grad_norm": 0.010510362829603903, + "learning_rate": 6.810925160402568e-07, + "loss": 0.0001, + "step": 9521 + }, + { + "epoch": 6.849127854702392, + "grad_norm": 0.984461300875884, + "learning_rate": 6.808074714756375e-07, + "loss": 0.0095, + "step": 9522 + }, + { + "epoch": 6.849847149793202, + "grad_norm": 1.2621741412960477, + "learning_rate": 6.805224690599119e-07, + "loss": 0.0082, + "step": 9523 + }, + { + "epoch": 6.850566444884014, + "grad_norm": 2.91411068747113, + "learning_rate": 6.802375088077435e-07, + "loss": 0.0077, + "step": 9524 + }, + { + "epoch": 6.851285739974824, + "grad_norm": 0.38591760792892305, + "learning_rate": 6.799525907337946e-07, + "loss": 0.0008, + "step": 9525 + }, + { + "epoch": 6.852005035065636, + "grad_norm": 2.1965360160359353, + "learning_rate": 6.796677148527236e-07, + "loss": 0.0121, + "step": 9526 + }, + { + "epoch": 6.852724330156446, + "grad_norm": 4.787956440885891, + "learning_rate": 6.793828811791896e-07, + "loss": 0.0517, + "step": 9527 + }, + { + "epoch": 6.853443625247258, + "grad_norm": 2.2987605615171174, + "learning_rate": 6.790980897278467e-07, + "loss": 0.0162, + "step": 9528 + }, + { + "epoch": 6.8541629203380685, + "grad_norm": 4.901009781338586, + "learning_rate": 6.788133405133487e-07, + "loss": 0.0796, + "step": 9529 + }, + { + "epoch": 6.85488221542888, + "grad_norm": 0.5076744551250423, + "learning_rate": 6.785286335503455e-07, + "loss": 0.0041, + "step": 9530 + }, + { + "epoch": 6.855601510519691, + "grad_norm": 0.19306745424712193, + "learning_rate": 6.782439688534854e-07, + "loss": 0.0004, + "step": 9531 + }, + { + "epoch": 6.856320805610502, + "grad_norm": 3.06482060348981, + "learning_rate": 6.779593464374163e-07, + "loss": 0.0409, + "step": 9532 + }, + { + "epoch": 6.857040100701313, + "grad_norm": 1.8450013747611929, + "learning_rate": 6.776747663167819e-07, + "loss": 0.0173, + "step": 9533 + }, + { + "epoch": 6.857759395792124, + "grad_norm": 2.726958610216243, + "learning_rate": 6.773902285062243e-07, + "loss": 0.0212, + "step": 9534 + }, + { + "epoch": 6.858478690882935, + "grad_norm": 1.2442638492673244, + "learning_rate": 6.771057330203833e-07, + "loss": 0.0058, + "step": 9535 + }, + { + "epoch": 6.859197985973745, + "grad_norm": 3.0476510546799283, + "learning_rate": 6.768212798738966e-07, + "loss": 0.0442, + "step": 9536 + }, + { + "epoch": 6.859917281064557, + "grad_norm": 2.6432891978400863, + "learning_rate": 6.765368690814005e-07, + "loss": 0.0277, + "step": 9537 + }, + { + "epoch": 6.860636576155367, + "grad_norm": 0.021096059940317175, + "learning_rate": 6.76252500657528e-07, + "loss": 0.0001, + "step": 9538 + }, + { + "epoch": 6.861355871246179, + "grad_norm": 3.5455783573368436, + "learning_rate": 6.759681746169104e-07, + "loss": 0.0331, + "step": 9539 + }, + { + "epoch": 6.862075166336989, + "grad_norm": 0.05090905317895362, + "learning_rate": 6.756838909741769e-07, + "loss": 0.0002, + "step": 9540 + }, + { + "epoch": 6.862794461427801, + "grad_norm": 1.7029079610926767, + "learning_rate": 6.75399649743954e-07, + "loss": 0.0168, + "step": 9541 + }, + { + "epoch": 6.8635137565186115, + "grad_norm": 2.8731549097670888, + "learning_rate": 6.75115450940867e-07, + "loss": 0.0466, + "step": 9542 + }, + { + "epoch": 6.864233051609423, + "grad_norm": 0.1481956401785415, + "learning_rate": 6.748312945795379e-07, + "loss": 0.0003, + "step": 9543 + }, + { + "epoch": 6.864952346700234, + "grad_norm": 1.4790521930739675, + "learning_rate": 6.745471806745872e-07, + "loss": 0.0204, + "step": 9544 + }, + { + "epoch": 6.865671641791045, + "grad_norm": 0.02862472181031118, + "learning_rate": 6.74263109240633e-07, + "loss": 0.0001, + "step": 9545 + }, + { + "epoch": 6.866390936881856, + "grad_norm": 3.5296929810847324, + "learning_rate": 6.739790802922916e-07, + "loss": 0.009, + "step": 9546 + }, + { + "epoch": 6.867110231972667, + "grad_norm": 5.442667630171063, + "learning_rate": 6.736950938441758e-07, + "loss": 0.045, + "step": 9547 + }, + { + "epoch": 6.867829527063478, + "grad_norm": 2.7943335403994567, + "learning_rate": 6.734111499108987e-07, + "loss": 0.0184, + "step": 9548 + }, + { + "epoch": 6.868548822154288, + "grad_norm": 1.59173190570458, + "learning_rate": 6.731272485070689e-07, + "loss": 0.0074, + "step": 9549 + }, + { + "epoch": 6.8692681172451, + "grad_norm": 0.18491045759426195, + "learning_rate": 6.728433896472937e-07, + "loss": 0.001, + "step": 9550 + }, + { + "epoch": 6.869987412335911, + "grad_norm": 4.522715951434985, + "learning_rate": 6.72559573346178e-07, + "loss": 0.0366, + "step": 9551 + }, + { + "epoch": 6.870706707426722, + "grad_norm": 1.8937888786075212, + "learning_rate": 6.722757996183243e-07, + "loss": 0.0179, + "step": 9552 + }, + { + "epoch": 6.871426002517532, + "grad_norm": 2.5501421807618585, + "learning_rate": 6.719920684783343e-07, + "loss": 0.0273, + "step": 9553 + }, + { + "epoch": 6.872145297608344, + "grad_norm": 4.264925263818095, + "learning_rate": 6.717083799408059e-07, + "loss": 0.0616, + "step": 9554 + }, + { + "epoch": 6.8728645926991545, + "grad_norm": 0.10256495341485758, + "learning_rate": 6.714247340203353e-07, + "loss": 0.0005, + "step": 9555 + }, + { + "epoch": 6.873583887789966, + "grad_norm": 2.651340369716094, + "learning_rate": 6.711411307315169e-07, + "loss": 0.0157, + "step": 9556 + }, + { + "epoch": 6.874303182880777, + "grad_norm": 1.3883287941938869, + "learning_rate": 6.708575700889421e-07, + "loss": 0.0061, + "step": 9557 + }, + { + "epoch": 6.875022477971588, + "grad_norm": 0.19088582413708563, + "learning_rate": 6.705740521072009e-07, + "loss": 0.0013, + "step": 9558 + }, + { + "epoch": 6.875741773062399, + "grad_norm": 1.1439158442594792, + "learning_rate": 6.702905768008806e-07, + "loss": 0.008, + "step": 9559 + }, + { + "epoch": 6.87646106815321, + "grad_norm": 3.495077623786986, + "learning_rate": 6.700071441845669e-07, + "loss": 0.0345, + "step": 9560 + }, + { + "epoch": 6.877180363244021, + "grad_norm": 4.1184176295259, + "learning_rate": 6.697237542728424e-07, + "loss": 0.0527, + "step": 9561 + }, + { + "epoch": 6.877899658334832, + "grad_norm": 2.9950716217793376, + "learning_rate": 6.694404070802885e-07, + "loss": 0.0329, + "step": 9562 + }, + { + "epoch": 6.878618953425643, + "grad_norm": 1.9956477368723815, + "learning_rate": 6.691571026214829e-07, + "loss": 0.023, + "step": 9563 + }, + { + "epoch": 6.879338248516454, + "grad_norm": 4.216292503955705, + "learning_rate": 6.688738409110036e-07, + "loss": 0.0828, + "step": 9564 + }, + { + "epoch": 6.880057543607265, + "grad_norm": 1.9566421602645654, + "learning_rate": 6.685906219634239e-07, + "loss": 0.0172, + "step": 9565 + }, + { + "epoch": 6.880776838698075, + "grad_norm": 3.6755326902701873, + "learning_rate": 6.683074457933163e-07, + "loss": 0.0632, + "step": 9566 + }, + { + "epoch": 6.881496133788887, + "grad_norm": 3.2931851577474753, + "learning_rate": 6.680243124152507e-07, + "loss": 0.0092, + "step": 9567 + }, + { + "epoch": 6.882215428879698, + "grad_norm": 4.269679572421442, + "learning_rate": 6.677412218437941e-07, + "loss": 0.0615, + "step": 9568 + }, + { + "epoch": 6.882934723970509, + "grad_norm": 0.4696373279589194, + "learning_rate": 6.674581740935131e-07, + "loss": 0.0007, + "step": 9569 + }, + { + "epoch": 6.88365401906132, + "grad_norm": 0.7264819806926565, + "learning_rate": 6.671751691789706e-07, + "loss": 0.0036, + "step": 9570 + }, + { + "epoch": 6.884373314152131, + "grad_norm": 0.05758271880247469, + "learning_rate": 6.668922071147278e-07, + "loss": 0.0002, + "step": 9571 + }, + { + "epoch": 6.885092609242942, + "grad_norm": 0.010638972776093833, + "learning_rate": 6.666092879153432e-07, + "loss": 0.0, + "step": 9572 + }, + { + "epoch": 6.885811904333753, + "grad_norm": 0.1412943050664434, + "learning_rate": 6.663264115953737e-07, + "loss": 0.0005, + "step": 9573 + }, + { + "epoch": 6.886531199424564, + "grad_norm": 5.474701353978498, + "learning_rate": 6.660435781693738e-07, + "loss": 0.0963, + "step": 9574 + }, + { + "epoch": 6.887250494515375, + "grad_norm": 0.010284388820532623, + "learning_rate": 6.657607876518959e-07, + "loss": 0.0, + "step": 9575 + }, + { + "epoch": 6.887969789606186, + "grad_norm": 1.8625459443010866, + "learning_rate": 6.654780400574899e-07, + "loss": 0.0173, + "step": 9576 + }, + { + "epoch": 6.888689084696997, + "grad_norm": 1.5154252088654732, + "learning_rate": 6.651953354007036e-07, + "loss": 0.0127, + "step": 9577 + }, + { + "epoch": 6.889408379787808, + "grad_norm": 0.9428311241402649, + "learning_rate": 6.649126736960821e-07, + "loss": 0.0043, + "step": 9578 + }, + { + "epoch": 6.890127674878619, + "grad_norm": 1.8499323394938465, + "learning_rate": 6.646300549581703e-07, + "loss": 0.0201, + "step": 9579 + }, + { + "epoch": 6.89084696996943, + "grad_norm": 0.2738857773989984, + "learning_rate": 6.643474792015085e-07, + "loss": 0.0019, + "step": 9580 + }, + { + "epoch": 6.891566265060241, + "grad_norm": 4.449727479546476, + "learning_rate": 6.640649464406358e-07, + "loss": 0.0179, + "step": 9581 + }, + { + "epoch": 6.892285560151052, + "grad_norm": 1.0990601196275487, + "learning_rate": 6.637824566900891e-07, + "loss": 0.008, + "step": 9582 + }, + { + "epoch": 6.893004855241863, + "grad_norm": 2.5194498118148996, + "learning_rate": 6.635000099644028e-07, + "loss": 0.0243, + "step": 9583 + }, + { + "epoch": 6.893724150332674, + "grad_norm": 3.652765958413593, + "learning_rate": 6.632176062781092e-07, + "loss": 0.0521, + "step": 9584 + }, + { + "epoch": 6.894443445423485, + "grad_norm": 2.858932711033515, + "learning_rate": 6.629352456457389e-07, + "loss": 0.0168, + "step": 9585 + }, + { + "epoch": 6.895162740514296, + "grad_norm": 3.926575681877727, + "learning_rate": 6.626529280818198e-07, + "loss": 0.0536, + "step": 9586 + }, + { + "epoch": 6.895882035605107, + "grad_norm": 0.016656132525300196, + "learning_rate": 6.623706536008774e-07, + "loss": 0.0001, + "step": 9587 + }, + { + "epoch": 6.896601330695918, + "grad_norm": 1.8402339090039799, + "learning_rate": 6.620884222174359e-07, + "loss": 0.0073, + "step": 9588 + }, + { + "epoch": 6.897320625786729, + "grad_norm": 4.4386585316132345, + "learning_rate": 6.618062339460148e-07, + "loss": 0.0598, + "step": 9589 + }, + { + "epoch": 6.89803992087754, + "grad_norm": 3.935050236345519, + "learning_rate": 6.61524088801135e-07, + "loss": 0.0347, + "step": 9590 + }, + { + "epoch": 6.898759215968351, + "grad_norm": 0.6490173094453943, + "learning_rate": 6.612419867973126e-07, + "loss": 0.0058, + "step": 9591 + }, + { + "epoch": 6.899478511059162, + "grad_norm": 0.1316601291324939, + "learning_rate": 6.609599279490625e-07, + "loss": 0.0002, + "step": 9592 + }, + { + "epoch": 6.900197806149973, + "grad_norm": 4.923261233821945, + "learning_rate": 6.606779122708969e-07, + "loss": 0.0884, + "step": 9593 + }, + { + "epoch": 6.900917101240784, + "grad_norm": 2.7850649488810033, + "learning_rate": 6.603959397773257e-07, + "loss": 0.0209, + "step": 9594 + }, + { + "epoch": 6.901636396331595, + "grad_norm": 3.179151030400151, + "learning_rate": 6.601140104828576e-07, + "loss": 0.0235, + "step": 9595 + }, + { + "epoch": 6.902355691422406, + "grad_norm": 6.423954120075853, + "learning_rate": 6.598321244019981e-07, + "loss": 0.043, + "step": 9596 + }, + { + "epoch": 6.903074986513217, + "grad_norm": 2.47137486969841, + "learning_rate": 6.595502815492507e-07, + "loss": 0.0271, + "step": 9597 + }, + { + "epoch": 6.9037942816040285, + "grad_norm": 0.5541800608477188, + "learning_rate": 6.592684819391169e-07, + "loss": 0.0008, + "step": 9598 + }, + { + "epoch": 6.904513576694839, + "grad_norm": 1.5422729661010472, + "learning_rate": 6.589867255860946e-07, + "loss": 0.0239, + "step": 9599 + }, + { + "epoch": 6.90523287178565, + "grad_norm": 1.4624973288996423, + "learning_rate": 6.587050125046826e-07, + "loss": 0.0083, + "step": 9600 + }, + { + "epoch": 6.905952166876461, + "grad_norm": 3.5401994731782427, + "learning_rate": 6.584233427093747e-07, + "loss": 0.0328, + "step": 9601 + }, + { + "epoch": 6.906671461967272, + "grad_norm": 0.25859812876691435, + "learning_rate": 6.581417162146628e-07, + "loss": 0.0004, + "step": 9602 + }, + { + "epoch": 6.907390757058083, + "grad_norm": 2.2495531979935195, + "learning_rate": 6.578601330350375e-07, + "loss": 0.0428, + "step": 9603 + }, + { + "epoch": 6.908110052148894, + "grad_norm": 1.198524107590265, + "learning_rate": 6.575785931849861e-07, + "loss": 0.0146, + "step": 9604 + }, + { + "epoch": 6.908829347239705, + "grad_norm": 4.632097716473925, + "learning_rate": 6.572970966789955e-07, + "loss": 0.0258, + "step": 9605 + }, + { + "epoch": 6.909548642330516, + "grad_norm": 2.521241463890767, + "learning_rate": 6.570156435315484e-07, + "loss": 0.0291, + "step": 9606 + }, + { + "epoch": 6.910267937421327, + "grad_norm": 2.3608812512701114, + "learning_rate": 6.567342337571264e-07, + "loss": 0.0322, + "step": 9607 + }, + { + "epoch": 6.910987232512138, + "grad_norm": 3.917356018963479, + "learning_rate": 6.564528673702082e-07, + "loss": 0.0664, + "step": 9608 + }, + { + "epoch": 6.9117065276029495, + "grad_norm": 1.7229404740701324, + "learning_rate": 6.56171544385271e-07, + "loss": 0.0282, + "step": 9609 + }, + { + "epoch": 6.91242582269376, + "grad_norm": 4.970198115729434, + "learning_rate": 6.558902648167885e-07, + "loss": 0.0755, + "step": 9610 + }, + { + "epoch": 6.9131451177845715, + "grad_norm": 4.532665348787438, + "learning_rate": 6.556090286792343e-07, + "loss": 0.0574, + "step": 9611 + }, + { + "epoch": 6.913864412875382, + "grad_norm": 3.240231118807355, + "learning_rate": 6.553278359870778e-07, + "loss": 0.0319, + "step": 9612 + }, + { + "epoch": 6.914583707966193, + "grad_norm": 0.46290770631184047, + "learning_rate": 6.550466867547867e-07, + "loss": 0.0005, + "step": 9613 + }, + { + "epoch": 6.915303003057004, + "grad_norm": 2.5771738775899027, + "learning_rate": 6.547655809968271e-07, + "loss": 0.0531, + "step": 9614 + }, + { + "epoch": 6.916022298147815, + "grad_norm": 0.06663407099329895, + "learning_rate": 6.544845187276622e-07, + "loss": 0.0001, + "step": 9615 + }, + { + "epoch": 6.916741593238626, + "grad_norm": 2.247154604311651, + "learning_rate": 6.542034999617529e-07, + "loss": 0.0134, + "step": 9616 + }, + { + "epoch": 6.917460888329437, + "grad_norm": 0.782355995383442, + "learning_rate": 6.539225247135581e-07, + "loss": 0.0072, + "step": 9617 + }, + { + "epoch": 6.918180183420248, + "grad_norm": 2.293934135645733, + "learning_rate": 6.53641592997535e-07, + "loss": 0.0137, + "step": 9618 + }, + { + "epoch": 6.918899478511059, + "grad_norm": 3.1675541421472433, + "learning_rate": 6.533607048281374e-07, + "loss": 0.0442, + "step": 9619 + }, + { + "epoch": 6.91961877360187, + "grad_norm": 0.06606395116039768, + "learning_rate": 6.530798602198173e-07, + "loss": 0.0002, + "step": 9620 + }, + { + "epoch": 6.920338068692681, + "grad_norm": 1.553566435231719, + "learning_rate": 6.527990591870257e-07, + "loss": 0.0177, + "step": 9621 + }, + { + "epoch": 6.9210573637834925, + "grad_norm": 0.06128461225008258, + "learning_rate": 6.525183017442097e-07, + "loss": 0.0002, + "step": 9622 + }, + { + "epoch": 6.921776658874303, + "grad_norm": 3.13320022757644, + "learning_rate": 6.522375879058149e-07, + "loss": 0.0337, + "step": 9623 + }, + { + "epoch": 6.9224959539651145, + "grad_norm": 0.007090267994506725, + "learning_rate": 6.519569176862843e-07, + "loss": 0.0, + "step": 9624 + }, + { + "epoch": 6.923215249055925, + "grad_norm": 0.966354244999809, + "learning_rate": 6.516762911000584e-07, + "loss": 0.0051, + "step": 9625 + }, + { + "epoch": 6.923934544146736, + "grad_norm": 8.507955689780061, + "learning_rate": 6.513957081615771e-07, + "loss": 0.1157, + "step": 9626 + }, + { + "epoch": 6.924653839237547, + "grad_norm": 0.02579147560233767, + "learning_rate": 6.511151688852761e-07, + "loss": 0.0001, + "step": 9627 + }, + { + "epoch": 6.925373134328359, + "grad_norm": 0.42041969903740073, + "learning_rate": 6.508346732855902e-07, + "loss": 0.0022, + "step": 9628 + }, + { + "epoch": 6.926092429419169, + "grad_norm": 2.185888763595296, + "learning_rate": 6.505542213769506e-07, + "loss": 0.0175, + "step": 9629 + }, + { + "epoch": 6.92681172450998, + "grad_norm": 0.8445446946920983, + "learning_rate": 6.502738131737879e-07, + "loss": 0.0084, + "step": 9630 + }, + { + "epoch": 6.927531019600791, + "grad_norm": 2.103423865514565, + "learning_rate": 6.499934486905288e-07, + "loss": 0.0135, + "step": 9631 + }, + { + "epoch": 6.928250314691602, + "grad_norm": 1.8148900417392313, + "learning_rate": 6.497131279415991e-07, + "loss": 0.0158, + "step": 9632 + }, + { + "epoch": 6.928969609782413, + "grad_norm": 2.6386855093392225, + "learning_rate": 6.494328509414216e-07, + "loss": 0.0352, + "step": 9633 + }, + { + "epoch": 6.929688904873224, + "grad_norm": 2.0428204389628757, + "learning_rate": 6.491526177044169e-07, + "loss": 0.009, + "step": 9634 + }, + { + "epoch": 6.9304081999640355, + "grad_norm": 1.631169899400338, + "learning_rate": 6.488724282450037e-07, + "loss": 0.0059, + "step": 9635 + }, + { + "epoch": 6.931127495054846, + "grad_norm": 0.1478297685767874, + "learning_rate": 6.485922825775978e-07, + "loss": 0.0004, + "step": 9636 + }, + { + "epoch": 6.9318467901456575, + "grad_norm": 1.627232033544253, + "learning_rate": 6.483121807166139e-07, + "loss": 0.0222, + "step": 9637 + }, + { + "epoch": 6.932566085236468, + "grad_norm": 0.1538152238067802, + "learning_rate": 6.480321226764634e-07, + "loss": 0.0003, + "step": 9638 + }, + { + "epoch": 6.93328538032728, + "grad_norm": 6.343316928320887, + "learning_rate": 6.477521084715558e-07, + "loss": 0.1243, + "step": 9639 + }, + { + "epoch": 6.93400467541809, + "grad_norm": 1.3161336964628794, + "learning_rate": 6.474721381162985e-07, + "loss": 0.0141, + "step": 9640 + }, + { + "epoch": 6.934723970508902, + "grad_norm": 2.9906443227435853, + "learning_rate": 6.471922116250953e-07, + "loss": 0.0298, + "step": 9641 + }, + { + "epoch": 6.935443265599712, + "grad_norm": 0.8727156268847963, + "learning_rate": 6.469123290123507e-07, + "loss": 0.0017, + "step": 9642 + }, + { + "epoch": 6.936162560690523, + "grad_norm": 0.15339532253950078, + "learning_rate": 6.466324902924644e-07, + "loss": 0.0008, + "step": 9643 + }, + { + "epoch": 6.936881855781334, + "grad_norm": 3.852883115408068, + "learning_rate": 6.463526954798343e-07, + "loss": 0.0359, + "step": 9644 + }, + { + "epoch": 6.937601150872146, + "grad_norm": 4.229482014822658, + "learning_rate": 6.460729445888567e-07, + "loss": 0.0347, + "step": 9645 + }, + { + "epoch": 6.938320445962956, + "grad_norm": 3.818227047641267, + "learning_rate": 6.45793237633925e-07, + "loss": 0.0302, + "step": 9646 + }, + { + "epoch": 6.939039741053767, + "grad_norm": 0.46368174171981996, + "learning_rate": 6.455135746294307e-07, + "loss": 0.0024, + "step": 9647 + }, + { + "epoch": 6.9397590361445785, + "grad_norm": 3.9746098894110165, + "learning_rate": 6.452339555897632e-07, + "loss": 0.0351, + "step": 9648 + }, + { + "epoch": 6.940478331235389, + "grad_norm": 0.03478881722989724, + "learning_rate": 6.449543805293091e-07, + "loss": 0.0002, + "step": 9649 + }, + { + "epoch": 6.9411976263262005, + "grad_norm": 3.6118552099120262, + "learning_rate": 6.446748494624531e-07, + "loss": 0.0542, + "step": 9650 + }, + { + "epoch": 6.941916921417011, + "grad_norm": 0.04659945039597893, + "learning_rate": 6.443953624035775e-07, + "loss": 0.0001, + "step": 9651 + }, + { + "epoch": 6.942636216507823, + "grad_norm": 1.7196588247502351, + "learning_rate": 6.44115919367062e-07, + "loss": 0.0207, + "step": 9652 + }, + { + "epoch": 6.943355511598633, + "grad_norm": 0.9065143237731321, + "learning_rate": 6.438365203672853e-07, + "loss": 0.007, + "step": 9653 + }, + { + "epoch": 6.944074806689445, + "grad_norm": 2.2989644331630634, + "learning_rate": 6.435571654186226e-07, + "loss": 0.0152, + "step": 9654 + }, + { + "epoch": 6.944794101780255, + "grad_norm": 0.1866866449331477, + "learning_rate": 6.432778545354473e-07, + "loss": 0.0009, + "step": 9655 + }, + { + "epoch": 6.945513396871067, + "grad_norm": 3.5262938833335333, + "learning_rate": 6.429985877321304e-07, + "loss": 0.0456, + "step": 9656 + }, + { + "epoch": 6.946232691961877, + "grad_norm": 0.6108637127941808, + "learning_rate": 6.427193650230398e-07, + "loss": 0.0049, + "step": 9657 + }, + { + "epoch": 6.946951987052689, + "grad_norm": 0.7529937692923163, + "learning_rate": 6.424401864225434e-07, + "loss": 0.0011, + "step": 9658 + }, + { + "epoch": 6.947671282143499, + "grad_norm": 3.3095885181002633, + "learning_rate": 6.421610519450049e-07, + "loss": 0.0438, + "step": 9659 + }, + { + "epoch": 6.94839057723431, + "grad_norm": 2.43177832447918, + "learning_rate": 6.418819616047865e-07, + "loss": 0.0347, + "step": 9660 + }, + { + "epoch": 6.9491098723251215, + "grad_norm": 1.3837627508636992, + "learning_rate": 6.416029154162473e-07, + "loss": 0.0063, + "step": 9661 + }, + { + "epoch": 6.949829167415932, + "grad_norm": 1.6116138238000521, + "learning_rate": 6.413239133937441e-07, + "loss": 0.0181, + "step": 9662 + }, + { + "epoch": 6.9505484625067435, + "grad_norm": 0.005613815086436319, + "learning_rate": 6.410449555516338e-07, + "loss": 0.0, + "step": 9663 + }, + { + "epoch": 6.951267757597554, + "grad_norm": 3.0452861268411566, + "learning_rate": 6.407660419042682e-07, + "loss": 0.071, + "step": 9664 + }, + { + "epoch": 6.951987052688366, + "grad_norm": 2.244764895700189, + "learning_rate": 6.40487172465998e-07, + "loss": 0.0219, + "step": 9665 + }, + { + "epoch": 6.952706347779176, + "grad_norm": 0.045953375684657835, + "learning_rate": 6.402083472511719e-07, + "loss": 0.0002, + "step": 9666 + }, + { + "epoch": 6.953425642869988, + "grad_norm": 6.158389921753417, + "learning_rate": 6.399295662741349e-07, + "loss": 0.1323, + "step": 9667 + }, + { + "epoch": 6.954144937960798, + "grad_norm": 0.03685703776961801, + "learning_rate": 6.396508295492322e-07, + "loss": 0.0002, + "step": 9668 + }, + { + "epoch": 6.95486423305161, + "grad_norm": 3.809390659535929, + "learning_rate": 6.393721370908044e-07, + "loss": 0.0635, + "step": 9669 + }, + { + "epoch": 6.95558352814242, + "grad_norm": 3.7102800499556, + "learning_rate": 6.39093488913191e-07, + "loss": 0.0293, + "step": 9670 + }, + { + "epoch": 6.956302823233232, + "grad_norm": 0.12179313440948113, + "learning_rate": 6.388148850307289e-07, + "loss": 0.0003, + "step": 9671 + }, + { + "epoch": 6.957022118324042, + "grad_norm": 2.273047380407653, + "learning_rate": 6.385363254577528e-07, + "loss": 0.0221, + "step": 9672 + }, + { + "epoch": 6.957741413414853, + "grad_norm": 0.9895092491683082, + "learning_rate": 6.382578102085948e-07, + "loss": 0.0049, + "step": 9673 + }, + { + "epoch": 6.9584607085056644, + "grad_norm": 2.914646906296391, + "learning_rate": 6.379793392975854e-07, + "loss": 0.045, + "step": 9674 + }, + { + "epoch": 6.959180003596476, + "grad_norm": 3.6325483212397236, + "learning_rate": 6.377009127390522e-07, + "loss": 0.0238, + "step": 9675 + }, + { + "epoch": 6.9598992986872865, + "grad_norm": 5.353957172308345, + "learning_rate": 6.374225305473207e-07, + "loss": 0.0507, + "step": 9676 + }, + { + "epoch": 6.960618593778097, + "grad_norm": 2.101849678777899, + "learning_rate": 6.371441927367142e-07, + "loss": 0.0184, + "step": 9677 + }, + { + "epoch": 6.961337888868909, + "grad_norm": 3.719445611803904, + "learning_rate": 6.368658993215531e-07, + "loss": 0.0797, + "step": 9678 + }, + { + "epoch": 6.962057183959719, + "grad_norm": 5.6340513618909664, + "learning_rate": 6.365876503161574e-07, + "loss": 0.0345, + "step": 9679 + }, + { + "epoch": 6.962776479050531, + "grad_norm": 2.0644594402008476, + "learning_rate": 6.363094457348427e-07, + "loss": 0.0195, + "step": 9680 + }, + { + "epoch": 6.963495774141341, + "grad_norm": 2.3988439793932934, + "learning_rate": 6.36031285591923e-07, + "loss": 0.0216, + "step": 9681 + }, + { + "epoch": 6.964215069232153, + "grad_norm": 2.3864723325911457, + "learning_rate": 6.357531699017104e-07, + "loss": 0.0216, + "step": 9682 + }, + { + "epoch": 6.964934364322963, + "grad_norm": 4.004629689205305, + "learning_rate": 6.354750986785138e-07, + "loss": 0.0305, + "step": 9683 + }, + { + "epoch": 6.965653659413775, + "grad_norm": 0.02437404772495766, + "learning_rate": 6.351970719366418e-07, + "loss": 0.0001, + "step": 9684 + }, + { + "epoch": 6.966372954504585, + "grad_norm": 1.2048697249739206, + "learning_rate": 6.349190896903982e-07, + "loss": 0.012, + "step": 9685 + }, + { + "epoch": 6.967092249595397, + "grad_norm": 0.7864130706520598, + "learning_rate": 6.346411519540861e-07, + "loss": 0.0073, + "step": 9686 + }, + { + "epoch": 6.9678115446862074, + "grad_norm": 0.008929274311701724, + "learning_rate": 6.343632587420059e-07, + "loss": 0.0, + "step": 9687 + }, + { + "epoch": 6.968530839777019, + "grad_norm": 1.8962622808716085, + "learning_rate": 6.340854100684557e-07, + "loss": 0.0162, + "step": 9688 + }, + { + "epoch": 6.9692501348678295, + "grad_norm": 3.295920122494271, + "learning_rate": 6.338076059477311e-07, + "loss": 0.032, + "step": 9689 + }, + { + "epoch": 6.96996942995864, + "grad_norm": 1.1513722739890686, + "learning_rate": 6.335298463941257e-07, + "loss": 0.0119, + "step": 9690 + }, + { + "epoch": 6.970688725049452, + "grad_norm": 0.08223739552715056, + "learning_rate": 6.332521314219309e-07, + "loss": 0.0002, + "step": 9691 + }, + { + "epoch": 6.971408020140262, + "grad_norm": 3.2532988202692588, + "learning_rate": 6.329744610454353e-07, + "loss": 0.0302, + "step": 9692 + }, + { + "epoch": 6.972127315231074, + "grad_norm": 2.8317608762407502, + "learning_rate": 6.326968352789258e-07, + "loss": 0.0299, + "step": 9693 + }, + { + "epoch": 6.972846610321884, + "grad_norm": 2.151510749045081, + "learning_rate": 6.32419254136686e-07, + "loss": 0.0278, + "step": 9694 + }, + { + "epoch": 6.973565905412696, + "grad_norm": 0.05789621752922867, + "learning_rate": 6.321417176329991e-07, + "loss": 0.0001, + "step": 9695 + }, + { + "epoch": 6.974285200503506, + "grad_norm": 4.0191137660240805, + "learning_rate": 6.318642257821443e-07, + "loss": 0.0401, + "step": 9696 + }, + { + "epoch": 6.975004495594318, + "grad_norm": 2.0970419795907564, + "learning_rate": 6.315867785983991e-07, + "loss": 0.0142, + "step": 9697 + }, + { + "epoch": 6.975723790685128, + "grad_norm": 4.5888039057682635, + "learning_rate": 6.313093760960386e-07, + "loss": 0.0308, + "step": 9698 + }, + { + "epoch": 6.97644308577594, + "grad_norm": 4.50608075721665, + "learning_rate": 6.310320182893352e-07, + "loss": 0.0491, + "step": 9699 + }, + { + "epoch": 6.9771623808667504, + "grad_norm": 1.5610212396464214, + "learning_rate": 6.307547051925602e-07, + "loss": 0.0118, + "step": 9700 + }, + { + "epoch": 6.977881675957562, + "grad_norm": 6.203633242878014, + "learning_rate": 6.304774368199816e-07, + "loss": 0.1183, + "step": 9701 + }, + { + "epoch": 6.9786009710483725, + "grad_norm": 3.059698058382572, + "learning_rate": 6.302002131858653e-07, + "loss": 0.0351, + "step": 9702 + }, + { + "epoch": 6.979320266139183, + "grad_norm": 3.510843026612664, + "learning_rate": 6.299230343044751e-07, + "loss": 0.0458, + "step": 9703 + }, + { + "epoch": 6.980039561229995, + "grad_norm": 2.0824707330523524, + "learning_rate": 6.296459001900719e-07, + "loss": 0.0267, + "step": 9704 + }, + { + "epoch": 6.980758856320806, + "grad_norm": 4.4935838349339825, + "learning_rate": 6.293688108569151e-07, + "loss": 0.0307, + "step": 9705 + }, + { + "epoch": 6.981478151411617, + "grad_norm": 2.80523392945201, + "learning_rate": 6.290917663192613e-07, + "loss": 0.0548, + "step": 9706 + }, + { + "epoch": 6.982197446502427, + "grad_norm": 6.981841534227392, + "learning_rate": 6.288147665913648e-07, + "loss": 0.1168, + "step": 9707 + }, + { + "epoch": 6.982916741593239, + "grad_norm": 1.6426983241090523, + "learning_rate": 6.285378116874781e-07, + "loss": 0.0238, + "step": 9708 + }, + { + "epoch": 6.983636036684049, + "grad_norm": 2.7896079080779668, + "learning_rate": 6.282609016218503e-07, + "loss": 0.0142, + "step": 9709 + }, + { + "epoch": 6.984355331774861, + "grad_norm": 2.4148307201624535, + "learning_rate": 6.279840364087298e-07, + "loss": 0.0368, + "step": 9710 + }, + { + "epoch": 6.985074626865671, + "grad_norm": 2.9636113044874857, + "learning_rate": 6.277072160623615e-07, + "loss": 0.0205, + "step": 9711 + }, + { + "epoch": 6.985793921956483, + "grad_norm": 2.6026723924095427, + "learning_rate": 6.274304405969881e-07, + "loss": 0.0302, + "step": 9712 + }, + { + "epoch": 6.9865132170472934, + "grad_norm": 4.6564458571837175, + "learning_rate": 6.271537100268503e-07, + "loss": 0.0757, + "step": 9713 + }, + { + "epoch": 6.987232512138105, + "grad_norm": 2.6764969980228392, + "learning_rate": 6.268770243661864e-07, + "loss": 0.0141, + "step": 9714 + }, + { + "epoch": 6.9879518072289155, + "grad_norm": 1.976308244440464, + "learning_rate": 6.266003836292316e-07, + "loss": 0.033, + "step": 9715 + }, + { + "epoch": 6.988671102319727, + "grad_norm": 1.555510900891112, + "learning_rate": 6.263237878302211e-07, + "loss": 0.0162, + "step": 9716 + }, + { + "epoch": 6.989390397410538, + "grad_norm": 6.598339999204931, + "learning_rate": 6.260472369833852e-07, + "loss": 0.1134, + "step": 9717 + }, + { + "epoch": 6.990109692501349, + "grad_norm": 2.7208094063703676, + "learning_rate": 6.257707311029532e-07, + "loss": 0.0254, + "step": 9718 + }, + { + "epoch": 6.99082898759216, + "grad_norm": 2.7954264922235397, + "learning_rate": 6.254942702031522e-07, + "loss": 0.0161, + "step": 9719 + }, + { + "epoch": 6.99154828268297, + "grad_norm": 0.015098171121178238, + "learning_rate": 6.252178542982052e-07, + "loss": 0.0, + "step": 9720 + }, + { + "epoch": 6.992267577773782, + "grad_norm": 0.25959309262557134, + "learning_rate": 6.249414834023357e-07, + "loss": 0.0008, + "step": 9721 + }, + { + "epoch": 6.992986872864593, + "grad_norm": 3.1484121799476994, + "learning_rate": 6.24665157529763e-07, + "loss": 0.0497, + "step": 9722 + }, + { + "epoch": 6.993706167955404, + "grad_norm": 0.7019378881419247, + "learning_rate": 6.243888766947045e-07, + "loss": 0.0061, + "step": 9723 + }, + { + "epoch": 6.994425463046214, + "grad_norm": 0.16849662881824215, + "learning_rate": 6.241126409113754e-07, + "loss": 0.0009, + "step": 9724 + }, + { + "epoch": 6.995144758137026, + "grad_norm": 1.8330161255594535, + "learning_rate": 6.238364501939881e-07, + "loss": 0.0096, + "step": 9725 + }, + { + "epoch": 6.995864053227836, + "grad_norm": 3.826417873222989, + "learning_rate": 6.235603045567539e-07, + "loss": 0.0494, + "step": 9726 + }, + { + "epoch": 6.996583348318648, + "grad_norm": 0.6409083859373588, + "learning_rate": 6.232842040138806e-07, + "loss": 0.0027, + "step": 9727 + }, + { + "epoch": 6.9973026434094585, + "grad_norm": 2.425480338905601, + "learning_rate": 6.23008148579574e-07, + "loss": 0.0041, + "step": 9728 + }, + { + "epoch": 6.99802193850027, + "grad_norm": 1.1661395098920293, + "learning_rate": 6.227321382680375e-07, + "loss": 0.0151, + "step": 9729 + }, + { + "epoch": 6.998741233591081, + "grad_norm": 0.609560619486359, + "learning_rate": 6.224561730934723e-07, + "loss": 0.0056, + "step": 9730 + }, + { + "epoch": 6.999460528681892, + "grad_norm": 1.8038280068611348, + "learning_rate": 6.221802530700779e-07, + "loss": 0.0066, + "step": 9731 + }, + { + "epoch": 7.000179823772703, + "grad_norm": 4.114373050138204, + "learning_rate": 6.219043782120509e-07, + "loss": 0.0365, + "step": 9732 + }, + { + "epoch": 7.000899118863514, + "grad_norm": 1.3665505377050755, + "learning_rate": 6.216285485335845e-07, + "loss": 0.0058, + "step": 9733 + }, + { + "epoch": 7.001618413954325, + "grad_norm": 1.2030470761623517, + "learning_rate": 6.213527640488713e-07, + "loss": 0.005, + "step": 9734 + }, + { + "epoch": 7.002337709045136, + "grad_norm": 1.177528560286943, + "learning_rate": 6.210770247721007e-07, + "loss": 0.0072, + "step": 9735 + }, + { + "epoch": 7.003057004135947, + "grad_norm": 1.5489354966774196, + "learning_rate": 6.208013307174597e-07, + "loss": 0.0228, + "step": 9736 + }, + { + "epoch": 7.003776299226757, + "grad_norm": 5.877010011332632, + "learning_rate": 6.205256818991341e-07, + "loss": 0.0618, + "step": 9737 + }, + { + "epoch": 7.004495594317569, + "grad_norm": 0.65390196044332, + "learning_rate": 6.202500783313059e-07, + "loss": 0.0054, + "step": 9738 + }, + { + "epoch": 7.005214889408379, + "grad_norm": 1.9652460539781427, + "learning_rate": 6.199745200281558e-07, + "loss": 0.0204, + "step": 9739 + }, + { + "epoch": 7.005934184499191, + "grad_norm": 5.350525595288654, + "learning_rate": 6.196990070038613e-07, + "loss": 0.0498, + "step": 9740 + }, + { + "epoch": 7.0066534795900015, + "grad_norm": 2.2995517576483104, + "learning_rate": 6.194235392725975e-07, + "loss": 0.0141, + "step": 9741 + }, + { + "epoch": 7.007372774680813, + "grad_norm": 2.7269996666606295, + "learning_rate": 6.191481168485394e-07, + "loss": 0.0411, + "step": 9742 + }, + { + "epoch": 7.008092069771624, + "grad_norm": 0.3834127382090928, + "learning_rate": 6.188727397458566e-07, + "loss": 0.0021, + "step": 9743 + }, + { + "epoch": 7.008811364862435, + "grad_norm": 1.8683486582124447, + "learning_rate": 6.185974079787182e-07, + "loss": 0.0093, + "step": 9744 + }, + { + "epoch": 7.009530659953246, + "grad_norm": 1.5150989782750033, + "learning_rate": 6.183221215612905e-07, + "loss": 0.0135, + "step": 9745 + }, + { + "epoch": 7.010249955044057, + "grad_norm": 0.8786936974282168, + "learning_rate": 6.180468805077372e-07, + "loss": 0.0049, + "step": 9746 + }, + { + "epoch": 7.010969250134868, + "grad_norm": 0.013581926109053693, + "learning_rate": 6.177716848322203e-07, + "loss": 0.0001, + "step": 9747 + }, + { + "epoch": 7.011688545225679, + "grad_norm": 0.25987265429038253, + "learning_rate": 6.174965345488989e-07, + "loss": 0.0006, + "step": 9748 + }, + { + "epoch": 7.01240784031649, + "grad_norm": 1.4016854400482874, + "learning_rate": 6.172214296719298e-07, + "loss": 0.016, + "step": 9749 + }, + { + "epoch": 7.013127135407301, + "grad_norm": 5.348321708035791, + "learning_rate": 6.169463702154681e-07, + "loss": 0.0874, + "step": 9750 + }, + { + "epoch": 7.013846430498112, + "grad_norm": 6.269216222339803, + "learning_rate": 6.16671356193665e-07, + "loss": 0.07, + "step": 9751 + }, + { + "epoch": 7.014565725588922, + "grad_norm": 0.3689410456776705, + "learning_rate": 6.163963876206719e-07, + "loss": 0.0019, + "step": 9752 + }, + { + "epoch": 7.015285020679734, + "grad_norm": 0.5368022816202745, + "learning_rate": 6.161214645106359e-07, + "loss": 0.0039, + "step": 9753 + }, + { + "epoch": 7.0160043157705445, + "grad_norm": 0.3953816191068485, + "learning_rate": 6.15846586877702e-07, + "loss": 0.0022, + "step": 9754 + }, + { + "epoch": 7.016723610861356, + "grad_norm": 3.567178329443589, + "learning_rate": 6.155717547360133e-07, + "loss": 0.0451, + "step": 9755 + }, + { + "epoch": 7.017442905952167, + "grad_norm": 0.353429397688412, + "learning_rate": 6.152969680997104e-07, + "loss": 0.0028, + "step": 9756 + }, + { + "epoch": 7.018162201042978, + "grad_norm": 0.21338580293209683, + "learning_rate": 6.150222269829309e-07, + "loss": 0.0007, + "step": 9757 + }, + { + "epoch": 7.018881496133789, + "grad_norm": 4.6521121219133414, + "learning_rate": 6.147475313998121e-07, + "loss": 0.0273, + "step": 9758 + }, + { + "epoch": 7.0196007912246, + "grad_norm": 5.5987357076771085, + "learning_rate": 6.144728813644867e-07, + "loss": 0.0519, + "step": 9759 + }, + { + "epoch": 7.020320086315411, + "grad_norm": 1.6690932222422843, + "learning_rate": 6.14198276891086e-07, + "loss": 0.0112, + "step": 9760 + }, + { + "epoch": 7.021039381406222, + "grad_norm": 0.053044068593894086, + "learning_rate": 6.139237179937391e-07, + "loss": 0.0004, + "step": 9761 + }, + { + "epoch": 7.021758676497033, + "grad_norm": 0.2793856852966488, + "learning_rate": 6.136492046865723e-07, + "loss": 0.0005, + "step": 9762 + }, + { + "epoch": 7.022477971587844, + "grad_norm": 2.3230873324825976, + "learning_rate": 6.133747369837098e-07, + "loss": 0.016, + "step": 9763 + }, + { + "epoch": 7.023197266678655, + "grad_norm": 1.4173650185788802, + "learning_rate": 6.131003148992735e-07, + "loss": 0.0067, + "step": 9764 + }, + { + "epoch": 7.023916561769466, + "grad_norm": 11.582972002030052, + "learning_rate": 6.12825938447383e-07, + "loss": 0.3121, + "step": 9765 + }, + { + "epoch": 7.024635856860277, + "grad_norm": 0.021870159250693622, + "learning_rate": 6.125516076421552e-07, + "loss": 0.0001, + "step": 9766 + }, + { + "epoch": 7.0253551519510875, + "grad_norm": 0.8743789172263322, + "learning_rate": 6.122773224977045e-07, + "loss": 0.0027, + "step": 9767 + }, + { + "epoch": 7.026074447041899, + "grad_norm": 0.22950432154271774, + "learning_rate": 6.120030830281447e-07, + "loss": 0.0006, + "step": 9768 + }, + { + "epoch": 7.02679374213271, + "grad_norm": 2.590214306689126, + "learning_rate": 6.11728889247585e-07, + "loss": 0.0161, + "step": 9769 + }, + { + "epoch": 7.027513037223521, + "grad_norm": 2.2860067902196883, + "learning_rate": 6.114547411701332e-07, + "loss": 0.0209, + "step": 9770 + }, + { + "epoch": 7.028232332314332, + "grad_norm": 2.1960689244871947, + "learning_rate": 6.111806388098947e-07, + "loss": 0.0177, + "step": 9771 + }, + { + "epoch": 7.028951627405143, + "grad_norm": 0.2797102538316058, + "learning_rate": 6.109065821809721e-07, + "loss": 0.0012, + "step": 9772 + }, + { + "epoch": 7.029670922495954, + "grad_norm": 2.0707625945223325, + "learning_rate": 6.106325712974672e-07, + "loss": 0.0164, + "step": 9773 + }, + { + "epoch": 7.030390217586765, + "grad_norm": 2.3611832476860335, + "learning_rate": 6.103586061734776e-07, + "loss": 0.0053, + "step": 9774 + }, + { + "epoch": 7.031109512677576, + "grad_norm": 0.9439936109094192, + "learning_rate": 6.100846868230996e-07, + "loss": 0.0053, + "step": 9775 + }, + { + "epoch": 7.031828807768387, + "grad_norm": 1.9266717113407623, + "learning_rate": 6.098108132604265e-07, + "loss": 0.0195, + "step": 9776 + }, + { + "epoch": 7.032548102859198, + "grad_norm": 1.679402996731176, + "learning_rate": 6.095369854995498e-07, + "loss": 0.0035, + "step": 9777 + }, + { + "epoch": 7.033267397950009, + "grad_norm": 0.48121142057598676, + "learning_rate": 6.092632035545586e-07, + "loss": 0.0017, + "step": 9778 + }, + { + "epoch": 7.03398669304082, + "grad_norm": 2.5119742877049056, + "learning_rate": 6.08989467439539e-07, + "loss": 0.0375, + "step": 9779 + }, + { + "epoch": 7.034705988131631, + "grad_norm": 1.0281419813001658, + "learning_rate": 6.087157771685754e-07, + "loss": 0.0101, + "step": 9780 + }, + { + "epoch": 7.035425283222442, + "grad_norm": 1.485451233792959, + "learning_rate": 6.084421327557496e-07, + "loss": 0.0132, + "step": 9781 + }, + { + "epoch": 7.036144578313253, + "grad_norm": 0.8923443099508634, + "learning_rate": 6.081685342151413e-07, + "loss": 0.0029, + "step": 9782 + }, + { + "epoch": 7.036863873404064, + "grad_norm": 1.013362186402358, + "learning_rate": 6.078949815608268e-07, + "loss": 0.0032, + "step": 9783 + }, + { + "epoch": 7.037583168494875, + "grad_norm": 0.1513677360280852, + "learning_rate": 6.076214748068821e-07, + "loss": 0.0004, + "step": 9784 + }, + { + "epoch": 7.038302463585686, + "grad_norm": 0.03740888172910649, + "learning_rate": 6.073480139673792e-07, + "loss": 0.0001, + "step": 9785 + }, + { + "epoch": 7.039021758676497, + "grad_norm": 1.549699569662118, + "learning_rate": 6.07074599056388e-07, + "loss": 0.0136, + "step": 9786 + }, + { + "epoch": 7.039741053767308, + "grad_norm": 1.3731901752819964, + "learning_rate": 6.06801230087976e-07, + "loss": 0.0085, + "step": 9787 + }, + { + "epoch": 7.040460348858119, + "grad_norm": 0.04063734848204765, + "learning_rate": 6.065279070762081e-07, + "loss": 0.0001, + "step": 9788 + }, + { + "epoch": 7.04117964394893, + "grad_norm": 3.5876945099075135, + "learning_rate": 6.062546300351485e-07, + "loss": 0.0198, + "step": 9789 + }, + { + "epoch": 7.041898939039741, + "grad_norm": 0.00845680920043748, + "learning_rate": 6.05981398978857e-07, + "loss": 0.0, + "step": 9790 + }, + { + "epoch": 7.042618234130552, + "grad_norm": 1.7516232451572995, + "learning_rate": 6.057082139213925e-07, + "loss": 0.0141, + "step": 9791 + }, + { + "epoch": 7.043337529221363, + "grad_norm": 0.019563133654776545, + "learning_rate": 6.054350748768098e-07, + "loss": 0.0001, + "step": 9792 + }, + { + "epoch": 7.044056824312174, + "grad_norm": 2.488040776063752, + "learning_rate": 6.051619818591621e-07, + "loss": 0.0184, + "step": 9793 + }, + { + "epoch": 7.044776119402985, + "grad_norm": 2.894855281501834, + "learning_rate": 6.04888934882502e-07, + "loss": 0.0485, + "step": 9794 + }, + { + "epoch": 7.0454954144937965, + "grad_norm": 2.369774065215541, + "learning_rate": 6.046159339608774e-07, + "loss": 0.0071, + "step": 9795 + }, + { + "epoch": 7.046214709584607, + "grad_norm": 0.5552719332684772, + "learning_rate": 6.04342979108335e-07, + "loss": 0.002, + "step": 9796 + }, + { + "epoch": 7.0469340046754185, + "grad_norm": 4.564016177820445, + "learning_rate": 6.040700703389184e-07, + "loss": 0.0098, + "step": 9797 + }, + { + "epoch": 7.047653299766229, + "grad_norm": 0.5055012036389469, + "learning_rate": 6.037972076666691e-07, + "loss": 0.0023, + "step": 9798 + }, + { + "epoch": 7.04837259485704, + "grad_norm": 6.137378117737867, + "learning_rate": 6.035243911056271e-07, + "loss": 0.0986, + "step": 9799 + }, + { + "epoch": 7.049091889947851, + "grad_norm": 3.5392386304356798, + "learning_rate": 6.032516206698288e-07, + "loss": 0.0175, + "step": 9800 + }, + { + "epoch": 7.049811185038662, + "grad_norm": 1.4492380806182459, + "learning_rate": 6.02978896373309e-07, + "loss": 0.0069, + "step": 9801 + }, + { + "epoch": 7.050530480129473, + "grad_norm": 1.9946345026974133, + "learning_rate": 6.027062182300995e-07, + "loss": 0.0203, + "step": 9802 + }, + { + "epoch": 7.051249775220284, + "grad_norm": 0.5416908788480933, + "learning_rate": 6.024335862542304e-07, + "loss": 0.0024, + "step": 9803 + }, + { + "epoch": 7.051969070311095, + "grad_norm": 0.5835200473148132, + "learning_rate": 6.021610004597287e-07, + "loss": 0.0016, + "step": 9804 + }, + { + "epoch": 7.052688365401906, + "grad_norm": 3.6295412598846597, + "learning_rate": 6.018884608606199e-07, + "loss": 0.0565, + "step": 9805 + }, + { + "epoch": 7.053407660492717, + "grad_norm": 0.7519892655186614, + "learning_rate": 6.016159674709262e-07, + "loss": 0.0019, + "step": 9806 + }, + { + "epoch": 7.054126955583528, + "grad_norm": 0.753304692493354, + "learning_rate": 6.01343520304668e-07, + "loss": 0.0048, + "step": 9807 + }, + { + "epoch": 7.0548462506743395, + "grad_norm": 0.9369233501154034, + "learning_rate": 6.010711193758633e-07, + "loss": 0.0046, + "step": 9808 + }, + { + "epoch": 7.05556554576515, + "grad_norm": 1.6537225928520494, + "learning_rate": 6.007987646985269e-07, + "loss": 0.0146, + "step": 9809 + }, + { + "epoch": 7.0562848408559615, + "grad_norm": 0.027350609330344368, + "learning_rate": 6.005264562866731e-07, + "loss": 0.0001, + "step": 9810 + }, + { + "epoch": 7.057004135946772, + "grad_norm": 4.239265932444658, + "learning_rate": 6.002541941543122e-07, + "loss": 0.0251, + "step": 9811 + }, + { + "epoch": 7.057723431037584, + "grad_norm": 1.5279913438043238, + "learning_rate": 5.999819783154524e-07, + "loss": 0.0054, + "step": 9812 + }, + { + "epoch": 7.058442726128394, + "grad_norm": 4.013339068917343, + "learning_rate": 5.997098087840997e-07, + "loss": 0.0325, + "step": 9813 + }, + { + "epoch": 7.059162021219205, + "grad_norm": 1.0126572004524181, + "learning_rate": 5.994376855742574e-07, + "loss": 0.0028, + "step": 9814 + }, + { + "epoch": 7.059881316310016, + "grad_norm": 0.5148122898800161, + "learning_rate": 5.991656086999273e-07, + "loss": 0.0013, + "step": 9815 + }, + { + "epoch": 7.060600611400827, + "grad_norm": 1.4691036785439193, + "learning_rate": 5.988935781751083e-07, + "loss": 0.0121, + "step": 9816 + }, + { + "epoch": 7.061319906491638, + "grad_norm": 0.07018330578326569, + "learning_rate": 5.986215940137964e-07, + "loss": 0.0003, + "step": 9817 + }, + { + "epoch": 7.062039201582449, + "grad_norm": 0.040856975239102235, + "learning_rate": 5.983496562299857e-07, + "loss": 0.0003, + "step": 9818 + }, + { + "epoch": 7.06275849667326, + "grad_norm": 1.4164009759688014, + "learning_rate": 5.980777648376682e-07, + "loss": 0.006, + "step": 9819 + }, + { + "epoch": 7.063477791764071, + "grad_norm": 2.373744862064544, + "learning_rate": 5.978059198508329e-07, + "loss": 0.0221, + "step": 9820 + }, + { + "epoch": 7.0641970868548825, + "grad_norm": 2.711481519574826, + "learning_rate": 5.975341212834669e-07, + "loss": 0.0366, + "step": 9821 + }, + { + "epoch": 7.064916381945693, + "grad_norm": 1.6648793594875908, + "learning_rate": 5.972623691495543e-07, + "loss": 0.0165, + "step": 9822 + }, + { + "epoch": 7.0656356770365045, + "grad_norm": 4.135003086508846, + "learning_rate": 5.969906634630778e-07, + "loss": 0.0302, + "step": 9823 + }, + { + "epoch": 7.066354972127315, + "grad_norm": 1.834011848906889, + "learning_rate": 5.967190042380166e-07, + "loss": 0.0163, + "step": 9824 + }, + { + "epoch": 7.067074267218127, + "grad_norm": 1.3692425742975858, + "learning_rate": 5.964473914883479e-07, + "loss": 0.0126, + "step": 9825 + }, + { + "epoch": 7.067793562308937, + "grad_norm": 6.324118774345091, + "learning_rate": 5.961758252280475e-07, + "loss": 0.0457, + "step": 9826 + }, + { + "epoch": 7.068512857399749, + "grad_norm": 2.3177232415966027, + "learning_rate": 5.959043054710877e-07, + "loss": 0.0339, + "step": 9827 + }, + { + "epoch": 7.069232152490559, + "grad_norm": 1.298498577601237, + "learning_rate": 5.956328322314382e-07, + "loss": 0.0133, + "step": 9828 + }, + { + "epoch": 7.06995144758137, + "grad_norm": 0.12951579738456803, + "learning_rate": 5.953614055230672e-07, + "loss": 0.0003, + "step": 9829 + }, + { + "epoch": 7.070670742672181, + "grad_norm": 2.3663452687719038, + "learning_rate": 5.950900253599393e-07, + "loss": 0.0078, + "step": 9830 + }, + { + "epoch": 7.071390037762992, + "grad_norm": 3.335353817728755, + "learning_rate": 5.948186917560188e-07, + "loss": 0.0175, + "step": 9831 + }, + { + "epoch": 7.072109332853803, + "grad_norm": 3.232970626698506, + "learning_rate": 5.945474047252655e-07, + "loss": 0.042, + "step": 9832 + }, + { + "epoch": 7.072828627944614, + "grad_norm": 0.012128805927072167, + "learning_rate": 5.942761642816378e-07, + "loss": 0.0001, + "step": 9833 + }, + { + "epoch": 7.0735479230354255, + "grad_norm": 0.3246883488204007, + "learning_rate": 5.940049704390912e-07, + "loss": 0.0018, + "step": 9834 + }, + { + "epoch": 7.074267218126236, + "grad_norm": 0.22464383374525052, + "learning_rate": 5.937338232115794e-07, + "loss": 0.0013, + "step": 9835 + }, + { + "epoch": 7.0749865132170475, + "grad_norm": 1.3833514544905234, + "learning_rate": 5.934627226130533e-07, + "loss": 0.0094, + "step": 9836 + }, + { + "epoch": 7.075705808307858, + "grad_norm": 1.3783970770707854, + "learning_rate": 5.931916686574616e-07, + "loss": 0.0049, + "step": 9837 + }, + { + "epoch": 7.07642510339867, + "grad_norm": 0.9931066386385917, + "learning_rate": 5.929206613587503e-07, + "loss": 0.0041, + "step": 9838 + }, + { + "epoch": 7.07714439848948, + "grad_norm": 2.7289868338261196, + "learning_rate": 5.926497007308633e-07, + "loss": 0.0168, + "step": 9839 + }, + { + "epoch": 7.077863693580292, + "grad_norm": 1.4368568935870967, + "learning_rate": 5.923787867877414e-07, + "loss": 0.01, + "step": 9840 + }, + { + "epoch": 7.078582988671102, + "grad_norm": 0.07550632921520924, + "learning_rate": 5.921079195433249e-07, + "loss": 0.0003, + "step": 9841 + }, + { + "epoch": 7.079302283761914, + "grad_norm": 0.05271639989624794, + "learning_rate": 5.918370990115496e-07, + "loss": 0.0002, + "step": 9842 + }, + { + "epoch": 7.080021578852724, + "grad_norm": 0.5502220229082071, + "learning_rate": 5.915663252063497e-07, + "loss": 0.0052, + "step": 9843 + }, + { + "epoch": 7.080740873943535, + "grad_norm": 1.8325274399656852, + "learning_rate": 5.912955981416571e-07, + "loss": 0.014, + "step": 9844 + }, + { + "epoch": 7.081460169034346, + "grad_norm": 1.962960645299671, + "learning_rate": 5.910249178314013e-07, + "loss": 0.0064, + "step": 9845 + }, + { + "epoch": 7.082179464125157, + "grad_norm": 0.005149597332949507, + "learning_rate": 5.907542842895084e-07, + "loss": 0.0, + "step": 9846 + }, + { + "epoch": 7.0828987592159685, + "grad_norm": 4.228946698612718, + "learning_rate": 5.904836975299044e-07, + "loss": 0.0442, + "step": 9847 + }, + { + "epoch": 7.083618054306779, + "grad_norm": 3.695507394217208, + "learning_rate": 5.902131575665108e-07, + "loss": 0.0233, + "step": 9848 + }, + { + "epoch": 7.0843373493975905, + "grad_norm": 1.6626487774057466, + "learning_rate": 5.899426644132471e-07, + "loss": 0.0151, + "step": 9849 + }, + { + "epoch": 7.085056644488401, + "grad_norm": 0.03164205384152862, + "learning_rate": 5.896722180840316e-07, + "loss": 0.0001, + "step": 9850 + }, + { + "epoch": 7.085775939579213, + "grad_norm": 2.9107882467720714, + "learning_rate": 5.894018185927774e-07, + "loss": 0.0194, + "step": 9851 + }, + { + "epoch": 7.086495234670023, + "grad_norm": 2.9760819666641987, + "learning_rate": 5.891314659533986e-07, + "loss": 0.0135, + "step": 9852 + }, + { + "epoch": 7.087214529760835, + "grad_norm": 0.10491673029969581, + "learning_rate": 5.888611601798049e-07, + "loss": 0.0002, + "step": 9853 + }, + { + "epoch": 7.087933824851645, + "grad_norm": 0.4756174692950594, + "learning_rate": 5.88590901285904e-07, + "loss": 0.0037, + "step": 9854 + }, + { + "epoch": 7.088653119942457, + "grad_norm": 0.11315353660601975, + "learning_rate": 5.883206892856012e-07, + "loss": 0.0005, + "step": 9855 + }, + { + "epoch": 7.089372415033267, + "grad_norm": 1.9075653421249503, + "learning_rate": 5.880505241927987e-07, + "loss": 0.0152, + "step": 9856 + }, + { + "epoch": 7.090091710124079, + "grad_norm": 1.8318379560621805, + "learning_rate": 5.877804060213983e-07, + "loss": 0.0194, + "step": 9857 + }, + { + "epoch": 7.090811005214889, + "grad_norm": 0.0071586586768308666, + "learning_rate": 5.875103347852972e-07, + "loss": 0.0, + "step": 9858 + }, + { + "epoch": 7.091530300305701, + "grad_norm": 0.8940520946933789, + "learning_rate": 5.872403104983912e-07, + "loss": 0.0019, + "step": 9859 + }, + { + "epoch": 7.0922495953965115, + "grad_norm": 0.6155018021822648, + "learning_rate": 5.869703331745735e-07, + "loss": 0.0013, + "step": 9860 + }, + { + "epoch": 7.092968890487322, + "grad_norm": 2.5178486802208755, + "learning_rate": 5.867004028277345e-07, + "loss": 0.0501, + "step": 9861 + }, + { + "epoch": 7.0936881855781335, + "grad_norm": 1.2218215324879884, + "learning_rate": 5.86430519471764e-07, + "loss": 0.005, + "step": 9862 + }, + { + "epoch": 7.094407480668944, + "grad_norm": 1.125837597553103, + "learning_rate": 5.861606831205467e-07, + "loss": 0.0147, + "step": 9863 + }, + { + "epoch": 7.095126775759756, + "grad_norm": 1.589978001856155, + "learning_rate": 5.858908937879661e-07, + "loss": 0.0201, + "step": 9864 + }, + { + "epoch": 7.095846070850566, + "grad_norm": 0.00699918654733812, + "learning_rate": 5.856211514879038e-07, + "loss": 0.0, + "step": 9865 + }, + { + "epoch": 7.096565365941378, + "grad_norm": 6.058307714191233, + "learning_rate": 5.853514562342385e-07, + "loss": 0.0416, + "step": 9866 + }, + { + "epoch": 7.097284661032188, + "grad_norm": 1.317467049334946, + "learning_rate": 5.850818080408457e-07, + "loss": 0.0079, + "step": 9867 + }, + { + "epoch": 7.098003956123, + "grad_norm": 0.6740684170314014, + "learning_rate": 5.848122069216008e-07, + "loss": 0.0047, + "step": 9868 + }, + { + "epoch": 7.09872325121381, + "grad_norm": 2.8732617004777548, + "learning_rate": 5.84542652890374e-07, + "loss": 0.0243, + "step": 9869 + }, + { + "epoch": 7.099442546304622, + "grad_norm": 3.8476649806689784, + "learning_rate": 5.842731459610351e-07, + "loss": 0.0386, + "step": 9870 + }, + { + "epoch": 7.100161841395432, + "grad_norm": 0.44191396855846404, + "learning_rate": 5.840036861474501e-07, + "loss": 0.0012, + "step": 9871 + }, + { + "epoch": 7.100881136486244, + "grad_norm": 2.6623643911797705, + "learning_rate": 5.837342734634831e-07, + "loss": 0.0287, + "step": 9872 + }, + { + "epoch": 7.1016004315770545, + "grad_norm": 2.6703711759007627, + "learning_rate": 5.834649079229966e-07, + "loss": 0.0278, + "step": 9873 + }, + { + "epoch": 7.102319726667866, + "grad_norm": 0.06431784335013278, + "learning_rate": 5.831955895398494e-07, + "loss": 0.0002, + "step": 9874 + }, + { + "epoch": 7.1030390217586765, + "grad_norm": 0.6564844511706763, + "learning_rate": 5.829263183278988e-07, + "loss": 0.0022, + "step": 9875 + }, + { + "epoch": 7.103758316849487, + "grad_norm": 2.2092230871978575, + "learning_rate": 5.826570943009988e-07, + "loss": 0.0236, + "step": 9876 + }, + { + "epoch": 7.104477611940299, + "grad_norm": 1.1094825125791934, + "learning_rate": 5.823879174730018e-07, + "loss": 0.0014, + "step": 9877 + }, + { + "epoch": 7.105196907031109, + "grad_norm": 0.2632214819369386, + "learning_rate": 5.821187878577571e-07, + "loss": 0.002, + "step": 9878 + }, + { + "epoch": 7.105916202121921, + "grad_norm": 0.895085384412105, + "learning_rate": 5.81849705469112e-07, + "loss": 0.0041, + "step": 9879 + }, + { + "epoch": 7.106635497212731, + "grad_norm": 2.8206316487968084, + "learning_rate": 5.815806703209114e-07, + "loss": 0.0274, + "step": 9880 + }, + { + "epoch": 7.107354792303543, + "grad_norm": 4.48596527112001, + "learning_rate": 5.813116824269975e-07, + "loss": 0.0654, + "step": 9881 + }, + { + "epoch": 7.108074087394353, + "grad_norm": 2.7655494039803794, + "learning_rate": 5.810427418012098e-07, + "loss": 0.0069, + "step": 9882 + }, + { + "epoch": 7.108793382485165, + "grad_norm": 3.0687239115935574, + "learning_rate": 5.807738484573867e-07, + "loss": 0.0345, + "step": 9883 + }, + { + "epoch": 7.109512677575975, + "grad_norm": 0.77038962381266, + "learning_rate": 5.805050024093628e-07, + "loss": 0.007, + "step": 9884 + }, + { + "epoch": 7.110231972666787, + "grad_norm": 4.1782243456572035, + "learning_rate": 5.802362036709707e-07, + "loss": 0.033, + "step": 9885 + }, + { + "epoch": 7.1109512677575974, + "grad_norm": 2.2181490563521065, + "learning_rate": 5.799674522560404e-07, + "loss": 0.0081, + "step": 9886 + }, + { + "epoch": 7.111670562848409, + "grad_norm": 3.500478312991828, + "learning_rate": 5.796987481783997e-07, + "loss": 0.0578, + "step": 9887 + }, + { + "epoch": 7.1123898579392195, + "grad_norm": 8.262525941641494, + "learning_rate": 5.794300914518736e-07, + "loss": 0.0893, + "step": 9888 + }, + { + "epoch": 7.113109153030031, + "grad_norm": 0.31156003064435134, + "learning_rate": 5.791614820902857e-07, + "loss": 0.0006, + "step": 9889 + }, + { + "epoch": 7.113828448120842, + "grad_norm": 1.6660953195661516, + "learning_rate": 5.78892920107456e-07, + "loss": 0.0115, + "step": 9890 + }, + { + "epoch": 7.114547743211652, + "grad_norm": 1.1427124160444087, + "learning_rate": 5.786244055172026e-07, + "loss": 0.0111, + "step": 9891 + }, + { + "epoch": 7.115267038302464, + "grad_norm": 1.5877902278333171, + "learning_rate": 5.78355938333341e-07, + "loss": 0.0132, + "step": 9892 + }, + { + "epoch": 7.115986333393274, + "grad_norm": 1.2750480594755154, + "learning_rate": 5.780875185696841e-07, + "loss": 0.0078, + "step": 9893 + }, + { + "epoch": 7.116705628484086, + "grad_norm": 3.668560496135561, + "learning_rate": 5.778191462400429e-07, + "loss": 0.0131, + "step": 9894 + }, + { + "epoch": 7.117424923574896, + "grad_norm": 0.017130186375859904, + "learning_rate": 5.775508213582253e-07, + "loss": 0.0001, + "step": 9895 + }, + { + "epoch": 7.118144218665708, + "grad_norm": 3.7466633864958543, + "learning_rate": 5.772825439380375e-07, + "loss": 0.0519, + "step": 9896 + }, + { + "epoch": 7.118863513756518, + "grad_norm": 2.8869045114912586, + "learning_rate": 5.770143139932824e-07, + "loss": 0.0226, + "step": 9897 + }, + { + "epoch": 7.11958280884733, + "grad_norm": 2.9797867963752998, + "learning_rate": 5.767461315377605e-07, + "loss": 0.0305, + "step": 9898 + }, + { + "epoch": 7.1203021039381404, + "grad_norm": 1.31443578061454, + "learning_rate": 5.764779965852715e-07, + "loss": 0.025, + "step": 9899 + }, + { + "epoch": 7.121021399028952, + "grad_norm": 0.0852463906212562, + "learning_rate": 5.762099091496108e-07, + "loss": 0.0005, + "step": 9900 + } + ], + "logging_steps": 1, + "max_steps": 13900, + "num_input_tokens_seen": 0, + "num_train_epochs": 10, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 4550865940480000.0, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}