diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,18815 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.0, + "eval_steps": 335, + "global_step": 2674, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0007479431563201197, + "grad_norm": 13.45175720426919, + "learning_rate": 7.518796992481203e-08, + "loss": 1.7684, + "step": 1 + }, + { + "epoch": 0.0007479431563201197, + "eval_loss": 1.7927727699279785, + "eval_runtime": 67.3331, + "eval_samples_per_second": 60.49, + "eval_steps_per_second": 3.787, + "step": 1 + }, + { + "epoch": 0.0014958863126402393, + "grad_norm": 13.104103629055608, + "learning_rate": 1.5037593984962406e-07, + "loss": 1.7164, + "step": 2 + }, + { + "epoch": 0.002243829468960359, + "grad_norm": 9.837386183370636, + "learning_rate": 2.255639097744361e-07, + "loss": 1.5948, + "step": 3 + }, + { + "epoch": 0.0029917726252804786, + "grad_norm": 14.409859357072028, + "learning_rate": 3.007518796992481e-07, + "loss": 1.6663, + "step": 4 + }, + { + "epoch": 0.0037397157816005983, + "grad_norm": 36.068149398521584, + "learning_rate": 3.7593984962406015e-07, + "loss": 1.5093, + "step": 5 + }, + { + "epoch": 0.004487658937920718, + "grad_norm": 57.27525875909395, + "learning_rate": 4.511278195488722e-07, + "loss": 1.6747, + "step": 6 + }, + { + "epoch": 0.005235602094240838, + "grad_norm": 28.933938424609323, + "learning_rate": 5.263157894736843e-07, + "loss": 1.9481, + "step": 7 + }, + { + "epoch": 0.005983545250560957, + "grad_norm": 17.779613633776844, + "learning_rate": 6.015037593984962e-07, + "loss": 1.6621, + "step": 8 + }, + { + "epoch": 0.006731488406881077, + "grad_norm": 6.971360043336999, + "learning_rate": 6.766917293233083e-07, + "loss": 1.3833, + "step": 9 + }, + { + "epoch": 0.0074794315632011965, + "grad_norm": 3.707436774664596, + "learning_rate": 7.518796992481203e-07, + "loss": 1.9195, + "step": 10 + }, + { + "epoch": 0.008227374719521317, + "grad_norm": 5.480107071953668, + "learning_rate": 8.270676691729324e-07, + "loss": 1.7367, + "step": 11 + }, + { + "epoch": 0.008975317875841436, + "grad_norm": 9.927131666439568, + "learning_rate": 9.022556390977444e-07, + "loss": 1.7745, + "step": 12 + }, + { + "epoch": 0.009723261032161555, + "grad_norm": 6.214832937658226, + "learning_rate": 9.774436090225563e-07, + "loss": 1.8908, + "step": 13 + }, + { + "epoch": 0.010471204188481676, + "grad_norm": 5.068480808536163, + "learning_rate": 1.0526315789473685e-06, + "loss": 1.8964, + "step": 14 + }, + { + "epoch": 0.011219147344801795, + "grad_norm": 4.008121477108565, + "learning_rate": 1.1278195488721805e-06, + "loss": 1.6892, + "step": 15 + }, + { + "epoch": 0.011967090501121914, + "grad_norm": 26.73956015124995, + "learning_rate": 1.2030075187969925e-06, + "loss": 1.7249, + "step": 16 + }, + { + "epoch": 0.012715033657442034, + "grad_norm": 19.29695209633763, + "learning_rate": 1.2781954887218045e-06, + "loss": 1.5952, + "step": 17 + }, + { + "epoch": 0.013462976813762155, + "grad_norm": 9.966756133567973, + "learning_rate": 1.3533834586466167e-06, + "loss": 1.3901, + "step": 18 + }, + { + "epoch": 0.014210919970082274, + "grad_norm": 9.890846672020515, + "learning_rate": 1.4285714285714286e-06, + "loss": 1.5159, + "step": 19 + }, + { + "epoch": 0.014958863126402393, + "grad_norm": 4.522574797960334, + "learning_rate": 1.5037593984962406e-06, + "loss": 1.7169, + "step": 20 + }, + { + "epoch": 0.015706806282722512, + "grad_norm": 5.230539824700393, + "learning_rate": 1.5789473684210526e-06, + "loss": 1.6806, + "step": 21 + }, + { + "epoch": 0.016454749439042633, + "grad_norm": 3.607960470380683, + "learning_rate": 1.6541353383458648e-06, + "loss": 1.6827, + "step": 22 + }, + { + "epoch": 0.01720269259536275, + "grad_norm": 2.744398913240352, + "learning_rate": 1.7293233082706767e-06, + "loss": 1.7006, + "step": 23 + }, + { + "epoch": 0.01795063575168287, + "grad_norm": 2.8717639587084407, + "learning_rate": 1.8045112781954887e-06, + "loss": 1.4661, + "step": 24 + }, + { + "epoch": 0.018698578908002993, + "grad_norm": 3.0437220493398933, + "learning_rate": 1.8796992481203007e-06, + "loss": 2.1345, + "step": 25 + }, + { + "epoch": 0.01944652206432311, + "grad_norm": 3.313548909331621, + "learning_rate": 1.9548872180451127e-06, + "loss": 1.806, + "step": 26 + }, + { + "epoch": 0.02019446522064323, + "grad_norm": 3.317925845966173, + "learning_rate": 2.030075187969925e-06, + "loss": 1.6789, + "step": 27 + }, + { + "epoch": 0.020942408376963352, + "grad_norm": 5.959992035274114, + "learning_rate": 2.105263157894737e-06, + "loss": 1.6652, + "step": 28 + }, + { + "epoch": 0.02169035153328347, + "grad_norm": 2.6289169118586155, + "learning_rate": 2.180451127819549e-06, + "loss": 1.8705, + "step": 29 + }, + { + "epoch": 0.02243829468960359, + "grad_norm": 2.651121499675689, + "learning_rate": 2.255639097744361e-06, + "loss": 1.9323, + "step": 30 + }, + { + "epoch": 0.02318623784592371, + "grad_norm": 2.6464921362707705, + "learning_rate": 2.330827067669173e-06, + "loss": 1.7269, + "step": 31 + }, + { + "epoch": 0.02393418100224383, + "grad_norm": 2.862921060917644, + "learning_rate": 2.406015037593985e-06, + "loss": 1.7082, + "step": 32 + }, + { + "epoch": 0.02468212415856395, + "grad_norm": 3.1577046890232934, + "learning_rate": 2.481203007518797e-06, + "loss": 1.7428, + "step": 33 + }, + { + "epoch": 0.025430067314884067, + "grad_norm": 3.3656722128876497, + "learning_rate": 2.556390977443609e-06, + "loss": 1.6392, + "step": 34 + }, + { + "epoch": 0.02617801047120419, + "grad_norm": 2.9934289377278174, + "learning_rate": 2.631578947368421e-06, + "loss": 1.7847, + "step": 35 + }, + { + "epoch": 0.02692595362752431, + "grad_norm": 3.6624828719740394, + "learning_rate": 2.7067669172932333e-06, + "loss": 1.5719, + "step": 36 + }, + { + "epoch": 0.027673896783844427, + "grad_norm": 3.5678591831752735, + "learning_rate": 2.781954887218045e-06, + "loss": 1.4067, + "step": 37 + }, + { + "epoch": 0.028421839940164548, + "grad_norm": 2.72801225302817, + "learning_rate": 2.8571428571428573e-06, + "loss": 1.8469, + "step": 38 + }, + { + "epoch": 0.02916978309648467, + "grad_norm": 2.791714382010126, + "learning_rate": 2.9323308270676694e-06, + "loss": 1.8024, + "step": 39 + }, + { + "epoch": 0.029917726252804786, + "grad_norm": 2.7212395316413103, + "learning_rate": 3.007518796992481e-06, + "loss": 1.6512, + "step": 40 + }, + { + "epoch": 0.030665669409124907, + "grad_norm": 2.2896088147803004, + "learning_rate": 3.0827067669172934e-06, + "loss": 1.4861, + "step": 41 + }, + { + "epoch": 0.031413612565445025, + "grad_norm": 2.6686482760935566, + "learning_rate": 3.157894736842105e-06, + "loss": 1.5503, + "step": 42 + }, + { + "epoch": 0.03216155572176515, + "grad_norm": 3.3845828146378514, + "learning_rate": 3.2330827067669174e-06, + "loss": 1.715, + "step": 43 + }, + { + "epoch": 0.032909498878085267, + "grad_norm": 3.3301002196986884, + "learning_rate": 3.3082706766917295e-06, + "loss": 1.7464, + "step": 44 + }, + { + "epoch": 0.033657442034405384, + "grad_norm": 2.90516459261753, + "learning_rate": 3.3834586466165413e-06, + "loss": 1.9425, + "step": 45 + }, + { + "epoch": 0.0344053851907255, + "grad_norm": 2.4259448682356552, + "learning_rate": 3.4586466165413535e-06, + "loss": 1.9111, + "step": 46 + }, + { + "epoch": 0.035153328347045626, + "grad_norm": 3.1673889301101843, + "learning_rate": 3.5338345864661657e-06, + "loss": 1.6246, + "step": 47 + }, + { + "epoch": 0.03590127150336574, + "grad_norm": 2.138447428552514, + "learning_rate": 3.6090225563909775e-06, + "loss": 1.673, + "step": 48 + }, + { + "epoch": 0.03664921465968586, + "grad_norm": 2.291536770273987, + "learning_rate": 3.6842105263157896e-06, + "loss": 1.5991, + "step": 49 + }, + { + "epoch": 0.037397157816005985, + "grad_norm": 2.9886113057655352, + "learning_rate": 3.7593984962406014e-06, + "loss": 1.9514, + "step": 50 + }, + { + "epoch": 0.0381451009723261, + "grad_norm": 3.0888420760285, + "learning_rate": 3.834586466165414e-06, + "loss": 1.5875, + "step": 51 + }, + { + "epoch": 0.03889304412864622, + "grad_norm": 2.7201985993847053, + "learning_rate": 3.909774436090225e-06, + "loss": 1.695, + "step": 52 + }, + { + "epoch": 0.039640987284966345, + "grad_norm": 2.785592410104916, + "learning_rate": 3.9849624060150376e-06, + "loss": 1.6262, + "step": 53 + }, + { + "epoch": 0.04038893044128646, + "grad_norm": 2.4363836565126893, + "learning_rate": 4.06015037593985e-06, + "loss": 1.4682, + "step": 54 + }, + { + "epoch": 0.04113687359760658, + "grad_norm": 2.060046081310132, + "learning_rate": 4.135338345864662e-06, + "loss": 1.7407, + "step": 55 + }, + { + "epoch": 0.041884816753926704, + "grad_norm": 3.1060904376487057, + "learning_rate": 4.210526315789474e-06, + "loss": 1.8657, + "step": 56 + }, + { + "epoch": 0.04263275991024682, + "grad_norm": 2.4557507788042034, + "learning_rate": 4.2857142857142855e-06, + "loss": 1.6157, + "step": 57 + }, + { + "epoch": 0.04338070306656694, + "grad_norm": 2.740832469095763, + "learning_rate": 4.360902255639098e-06, + "loss": 1.2267, + "step": 58 + }, + { + "epoch": 0.044128646222887064, + "grad_norm": 2.5805652110244592, + "learning_rate": 4.43609022556391e-06, + "loss": 1.5782, + "step": 59 + }, + { + "epoch": 0.04487658937920718, + "grad_norm": 2.202057127425301, + "learning_rate": 4.511278195488722e-06, + "loss": 1.5647, + "step": 60 + }, + { + "epoch": 0.0456245325355273, + "grad_norm": 2.3940849975080036, + "learning_rate": 4.586466165413534e-06, + "loss": 1.8868, + "step": 61 + }, + { + "epoch": 0.04637247569184742, + "grad_norm": 2.4089677826284475, + "learning_rate": 4.661654135338346e-06, + "loss": 1.6457, + "step": 62 + }, + { + "epoch": 0.04712041884816754, + "grad_norm": 2.955047241996721, + "learning_rate": 4.736842105263158e-06, + "loss": 1.4709, + "step": 63 + }, + { + "epoch": 0.04786836200448766, + "grad_norm": 3.6906828155503923, + "learning_rate": 4.81203007518797e-06, + "loss": 1.6943, + "step": 64 + }, + { + "epoch": 0.048616305160807775, + "grad_norm": 2.6824831771816324, + "learning_rate": 4.887218045112782e-06, + "loss": 1.9541, + "step": 65 + }, + { + "epoch": 0.0493642483171279, + "grad_norm": 2.522716859155962, + "learning_rate": 4.962406015037594e-06, + "loss": 1.6531, + "step": 66 + }, + { + "epoch": 0.05011219147344802, + "grad_norm": 5.1791083319422055, + "learning_rate": 5.0375939849624065e-06, + "loss": 1.6532, + "step": 67 + }, + { + "epoch": 0.050860134629768135, + "grad_norm": 2.531034389493077, + "learning_rate": 5.112781954887218e-06, + "loss": 1.8735, + "step": 68 + }, + { + "epoch": 0.05160807778608826, + "grad_norm": 3.393924864285412, + "learning_rate": 5.187969924812031e-06, + "loss": 1.7366, + "step": 69 + }, + { + "epoch": 0.05235602094240838, + "grad_norm": 7.957729848027239, + "learning_rate": 5.263157894736842e-06, + "loss": 1.9062, + "step": 70 + }, + { + "epoch": 0.053103964098728494, + "grad_norm": 3.0992317550508726, + "learning_rate": 5.338345864661654e-06, + "loss": 1.6943, + "step": 71 + }, + { + "epoch": 0.05385190725504862, + "grad_norm": 2.3644126699846777, + "learning_rate": 5.413533834586467e-06, + "loss": 1.719, + "step": 72 + }, + { + "epoch": 0.054599850411368736, + "grad_norm": 2.792104348472198, + "learning_rate": 5.488721804511279e-06, + "loss": 1.6664, + "step": 73 + }, + { + "epoch": 0.055347793567688854, + "grad_norm": 2.376135695889631, + "learning_rate": 5.56390977443609e-06, + "loss": 1.6474, + "step": 74 + }, + { + "epoch": 0.05609573672400898, + "grad_norm": 2.2499038020231925, + "learning_rate": 5.639097744360903e-06, + "loss": 1.5651, + "step": 75 + }, + { + "epoch": 0.056843679880329095, + "grad_norm": 2.8891015680089325, + "learning_rate": 5.7142857142857145e-06, + "loss": 1.8761, + "step": 76 + }, + { + "epoch": 0.05759162303664921, + "grad_norm": 2.5518946363902995, + "learning_rate": 5.789473684210527e-06, + "loss": 1.802, + "step": 77 + }, + { + "epoch": 0.05833956619296934, + "grad_norm": 2.996794712734532, + "learning_rate": 5.864661654135339e-06, + "loss": 1.6085, + "step": 78 + }, + { + "epoch": 0.059087509349289455, + "grad_norm": 3.2987465412828003, + "learning_rate": 5.939849624060151e-06, + "loss": 1.8745, + "step": 79 + }, + { + "epoch": 0.05983545250560957, + "grad_norm": 20.322760933421865, + "learning_rate": 6.015037593984962e-06, + "loss": 1.6848, + "step": 80 + }, + { + "epoch": 0.0605833956619297, + "grad_norm": 3.455101096656784, + "learning_rate": 6.0902255639097755e-06, + "loss": 1.3069, + "step": 81 + }, + { + "epoch": 0.061331338818249814, + "grad_norm": 2.6904428230464004, + "learning_rate": 6.165413533834587e-06, + "loss": 1.6141, + "step": 82 + }, + { + "epoch": 0.06207928197456993, + "grad_norm": 2.406092339167224, + "learning_rate": 6.240601503759399e-06, + "loss": 1.6971, + "step": 83 + }, + { + "epoch": 0.06282722513089005, + "grad_norm": 2.8015888528595094, + "learning_rate": 6.31578947368421e-06, + "loss": 1.9062, + "step": 84 + }, + { + "epoch": 0.06357516828721017, + "grad_norm": 2.3131937577673813, + "learning_rate": 6.390977443609023e-06, + "loss": 1.803, + "step": 85 + }, + { + "epoch": 0.0643231114435303, + "grad_norm": 2.3405214752706964, + "learning_rate": 6.466165413533835e-06, + "loss": 1.9218, + "step": 86 + }, + { + "epoch": 0.06507105459985041, + "grad_norm": 2.2555328497819844, + "learning_rate": 6.541353383458648e-06, + "loss": 1.5619, + "step": 87 + }, + { + "epoch": 0.06581899775617053, + "grad_norm": 2.484070644859524, + "learning_rate": 6.616541353383459e-06, + "loss": 1.5157, + "step": 88 + }, + { + "epoch": 0.06656694091249066, + "grad_norm": 2.5365480608199373, + "learning_rate": 6.691729323308271e-06, + "loss": 1.7451, + "step": 89 + }, + { + "epoch": 0.06731488406881077, + "grad_norm": 2.1642631182542287, + "learning_rate": 6.766917293233083e-06, + "loss": 1.573, + "step": 90 + }, + { + "epoch": 0.06806282722513089, + "grad_norm": 2.5348174824905505, + "learning_rate": 6.842105263157896e-06, + "loss": 1.7236, + "step": 91 + }, + { + "epoch": 0.068810770381451, + "grad_norm": 2.827227742644688, + "learning_rate": 6.917293233082707e-06, + "loss": 1.9078, + "step": 92 + }, + { + "epoch": 0.06955871353777113, + "grad_norm": 2.353325557914655, + "learning_rate": 6.992481203007519e-06, + "loss": 1.7059, + "step": 93 + }, + { + "epoch": 0.07030665669409125, + "grad_norm": 2.7293748741010937, + "learning_rate": 7.067669172932331e-06, + "loss": 1.9682, + "step": 94 + }, + { + "epoch": 0.07105459985041136, + "grad_norm": 2.319815241148681, + "learning_rate": 7.1428571428571436e-06, + "loss": 1.6428, + "step": 95 + }, + { + "epoch": 0.07180254300673149, + "grad_norm": 2.4408420958741477, + "learning_rate": 7.218045112781955e-06, + "loss": 1.4682, + "step": 96 + }, + { + "epoch": 0.07255048616305161, + "grad_norm": 4.3565116281991205, + "learning_rate": 7.293233082706768e-06, + "loss": 1.6397, + "step": 97 + }, + { + "epoch": 0.07329842931937172, + "grad_norm": 2.7855757377709036, + "learning_rate": 7.368421052631579e-06, + "loss": 1.5265, + "step": 98 + }, + { + "epoch": 0.07404637247569185, + "grad_norm": 2.489763433368306, + "learning_rate": 7.4436090225563915e-06, + "loss": 1.688, + "step": 99 + }, + { + "epoch": 0.07479431563201197, + "grad_norm": 3.400358807186724, + "learning_rate": 7.518796992481203e-06, + "loss": 1.4932, + "step": 100 + }, + { + "epoch": 0.07554225878833208, + "grad_norm": 3.116034881235642, + "learning_rate": 7.593984962406016e-06, + "loss": 1.396, + "step": 101 + }, + { + "epoch": 0.0762902019446522, + "grad_norm": 2.607821266210195, + "learning_rate": 7.669172932330828e-06, + "loss": 1.5772, + "step": 102 + }, + { + "epoch": 0.07703814510097233, + "grad_norm": 2.5279266158309706, + "learning_rate": 7.74436090225564e-06, + "loss": 1.6847, + "step": 103 + }, + { + "epoch": 0.07778608825729244, + "grad_norm": 2.5899094045706867, + "learning_rate": 7.81954887218045e-06, + "loss": 1.5095, + "step": 104 + }, + { + "epoch": 0.07853403141361257, + "grad_norm": 2.9428682598898304, + "learning_rate": 7.894736842105265e-06, + "loss": 1.6034, + "step": 105 + }, + { + "epoch": 0.07928197456993269, + "grad_norm": 2.8671074457828025, + "learning_rate": 7.969924812030075e-06, + "loss": 1.3158, + "step": 106 + }, + { + "epoch": 0.0800299177262528, + "grad_norm": 2.889761554837709, + "learning_rate": 8.045112781954887e-06, + "loss": 1.6768, + "step": 107 + }, + { + "epoch": 0.08077786088257292, + "grad_norm": 2.212584898864655, + "learning_rate": 8.1203007518797e-06, + "loss": 1.5976, + "step": 108 + }, + { + "epoch": 0.08152580403889305, + "grad_norm": 2.488100290515364, + "learning_rate": 8.195488721804512e-06, + "loss": 1.7197, + "step": 109 + }, + { + "epoch": 0.08227374719521316, + "grad_norm": 2.4694457270584476, + "learning_rate": 8.270676691729324e-06, + "loss": 1.5914, + "step": 110 + }, + { + "epoch": 0.08302169035153328, + "grad_norm": 2.3476576009609516, + "learning_rate": 8.345864661654136e-06, + "loss": 1.8082, + "step": 111 + }, + { + "epoch": 0.08376963350785341, + "grad_norm": 3.267949931687407, + "learning_rate": 8.421052631578948e-06, + "loss": 1.8045, + "step": 112 + }, + { + "epoch": 0.08451757666417352, + "grad_norm": 2.035659033698736, + "learning_rate": 8.49624060150376e-06, + "loss": 1.6215, + "step": 113 + }, + { + "epoch": 0.08526551982049364, + "grad_norm": 2.9258481211430425, + "learning_rate": 8.571428571428571e-06, + "loss": 1.4913, + "step": 114 + }, + { + "epoch": 0.08601346297681377, + "grad_norm": 2.674514335157528, + "learning_rate": 8.646616541353385e-06, + "loss": 1.7057, + "step": 115 + }, + { + "epoch": 0.08676140613313388, + "grad_norm": 2.440002424313609, + "learning_rate": 8.721804511278195e-06, + "loss": 1.4774, + "step": 116 + }, + { + "epoch": 0.087509349289454, + "grad_norm": 2.527831588764391, + "learning_rate": 8.796992481203007e-06, + "loss": 1.6129, + "step": 117 + }, + { + "epoch": 0.08825729244577413, + "grad_norm": 3.10505069929713, + "learning_rate": 8.87218045112782e-06, + "loss": 1.7757, + "step": 118 + }, + { + "epoch": 0.08900523560209424, + "grad_norm": 2.3340425460144765, + "learning_rate": 8.947368421052632e-06, + "loss": 1.9254, + "step": 119 + }, + { + "epoch": 0.08975317875841436, + "grad_norm": 3.165409254905692, + "learning_rate": 9.022556390977444e-06, + "loss": 1.6244, + "step": 120 + }, + { + "epoch": 0.09050112191473449, + "grad_norm": 3.327534101861783, + "learning_rate": 9.097744360902256e-06, + "loss": 1.6068, + "step": 121 + }, + { + "epoch": 0.0912490650710546, + "grad_norm": 2.961403744510622, + "learning_rate": 9.172932330827068e-06, + "loss": 1.5006, + "step": 122 + }, + { + "epoch": 0.09199700822737472, + "grad_norm": 2.3725193806081606, + "learning_rate": 9.24812030075188e-06, + "loss": 1.9188, + "step": 123 + }, + { + "epoch": 0.09274495138369485, + "grad_norm": 2.409600340281453, + "learning_rate": 9.323308270676693e-06, + "loss": 1.7002, + "step": 124 + }, + { + "epoch": 0.09349289454001496, + "grad_norm": 2.430776965748937, + "learning_rate": 9.398496240601505e-06, + "loss": 1.6084, + "step": 125 + }, + { + "epoch": 0.09424083769633508, + "grad_norm": 2.375340172111959, + "learning_rate": 9.473684210526315e-06, + "loss": 1.5308, + "step": 126 + }, + { + "epoch": 0.09498878085265519, + "grad_norm": 2.1830854506787816, + "learning_rate": 9.54887218045113e-06, + "loss": 1.6432, + "step": 127 + }, + { + "epoch": 0.09573672400897532, + "grad_norm": 2.2519097048971926, + "learning_rate": 9.62406015037594e-06, + "loss": 1.7231, + "step": 128 + }, + { + "epoch": 0.09648466716529544, + "grad_norm": 2.513433683059653, + "learning_rate": 9.699248120300752e-06, + "loss": 1.4952, + "step": 129 + }, + { + "epoch": 0.09723261032161555, + "grad_norm": 2.5383646650083604, + "learning_rate": 9.774436090225564e-06, + "loss": 1.6865, + "step": 130 + }, + { + "epoch": 0.09798055347793568, + "grad_norm": 2.1276076466743103, + "learning_rate": 9.849624060150376e-06, + "loss": 1.5339, + "step": 131 + }, + { + "epoch": 0.0987284966342558, + "grad_norm": 2.8788967696294483, + "learning_rate": 9.924812030075189e-06, + "loss": 1.598, + "step": 132 + }, + { + "epoch": 0.09947643979057591, + "grad_norm": 2.088159312573578, + "learning_rate": 1e-05, + "loss": 1.5574, + "step": 133 + }, + { + "epoch": 0.10022438294689603, + "grad_norm": 2.353509857150076, + "learning_rate": 9.999996178530757e-06, + "loss": 1.7287, + "step": 134 + }, + { + "epoch": 0.10097232610321616, + "grad_norm": 2.2707353977481795, + "learning_rate": 9.999984714128867e-06, + "loss": 1.7614, + "step": 135 + }, + { + "epoch": 0.10172026925953627, + "grad_norm": 2.2240877461159667, + "learning_rate": 9.999965606811854e-06, + "loss": 1.757, + "step": 136 + }, + { + "epoch": 0.1024682124158564, + "grad_norm": 2.314671683345322, + "learning_rate": 9.999938856608927e-06, + "loss": 1.7775, + "step": 137 + }, + { + "epoch": 0.10321615557217652, + "grad_norm": 3.1742095392387486, + "learning_rate": 9.999904463560974e-06, + "loss": 1.6698, + "step": 138 + }, + { + "epoch": 0.10396409872849663, + "grad_norm": 4.725926401858918, + "learning_rate": 9.99986242772057e-06, + "loss": 1.7825, + "step": 139 + }, + { + "epoch": 0.10471204188481675, + "grad_norm": 3.1703134683207783, + "learning_rate": 9.999812749151968e-06, + "loss": 1.4666, + "step": 140 + }, + { + "epoch": 0.10545998504113688, + "grad_norm": 3.4509932577146056, + "learning_rate": 9.999755427931108e-06, + "loss": 1.6277, + "step": 141 + }, + { + "epoch": 0.10620792819745699, + "grad_norm": 3.1828810316058727, + "learning_rate": 9.99969046414561e-06, + "loss": 1.782, + "step": 142 + }, + { + "epoch": 0.10695587135377711, + "grad_norm": 2.724930172769225, + "learning_rate": 9.999617857894777e-06, + "loss": 1.7643, + "step": 143 + }, + { + "epoch": 0.10770381451009724, + "grad_norm": 2.7958296876934843, + "learning_rate": 9.999537609289592e-06, + "loss": 1.8199, + "step": 144 + }, + { + "epoch": 0.10845175766641735, + "grad_norm": 2.3106400757967536, + "learning_rate": 9.999449718452724e-06, + "loss": 1.6195, + "step": 145 + }, + { + "epoch": 0.10919970082273747, + "grad_norm": 2.1185455596651637, + "learning_rate": 9.999354185518525e-06, + "loss": 1.5901, + "step": 146 + }, + { + "epoch": 0.1099476439790576, + "grad_norm": 2.0081758509414156, + "learning_rate": 9.99925101063302e-06, + "loss": 1.7392, + "step": 147 + }, + { + "epoch": 0.11069558713537771, + "grad_norm": 2.401272594547328, + "learning_rate": 9.999140193953923e-06, + "loss": 1.5437, + "step": 148 + }, + { + "epoch": 0.11144353029169783, + "grad_norm": 2.34387847838624, + "learning_rate": 9.999021735650626e-06, + "loss": 1.6985, + "step": 149 + }, + { + "epoch": 0.11219147344801796, + "grad_norm": 4.491938044648106, + "learning_rate": 9.998895635904205e-06, + "loss": 1.9198, + "step": 150 + }, + { + "epoch": 0.11293941660433807, + "grad_norm": 1.9666724830650486, + "learning_rate": 9.998761894907414e-06, + "loss": 1.8332, + "step": 151 + }, + { + "epoch": 0.11368735976065819, + "grad_norm": 2.5739790456375813, + "learning_rate": 9.998620512864686e-06, + "loss": 1.783, + "step": 152 + }, + { + "epoch": 0.11443530291697832, + "grad_norm": 2.0747234190033184, + "learning_rate": 9.998471489992138e-06, + "loss": 1.7392, + "step": 153 + }, + { + "epoch": 0.11518324607329843, + "grad_norm": 2.3492111738686337, + "learning_rate": 9.998314826517564e-06, + "loss": 1.3888, + "step": 154 + }, + { + "epoch": 0.11593118922961855, + "grad_norm": 2.135746607192034, + "learning_rate": 9.998150522680437e-06, + "loss": 1.8098, + "step": 155 + }, + { + "epoch": 0.11667913238593867, + "grad_norm": 2.1268720152576295, + "learning_rate": 9.997978578731908e-06, + "loss": 1.3913, + "step": 156 + }, + { + "epoch": 0.11742707554225879, + "grad_norm": 2.11034920563073, + "learning_rate": 9.997798994934812e-06, + "loss": 1.6586, + "step": 157 + }, + { + "epoch": 0.11817501869857891, + "grad_norm": 2.0238944853353966, + "learning_rate": 9.997611771563661e-06, + "loss": 1.4424, + "step": 158 + }, + { + "epoch": 0.11892296185489903, + "grad_norm": 2.057729999365661, + "learning_rate": 9.997416908904634e-06, + "loss": 1.4321, + "step": 159 + }, + { + "epoch": 0.11967090501121914, + "grad_norm": 2.458835410653752, + "learning_rate": 9.997214407255602e-06, + "loss": 1.5388, + "step": 160 + }, + { + "epoch": 0.12041884816753927, + "grad_norm": 2.6539844359190994, + "learning_rate": 9.997004266926105e-06, + "loss": 1.7007, + "step": 161 + }, + { + "epoch": 0.1211667913238594, + "grad_norm": 2.076912543455477, + "learning_rate": 9.99678648823736e-06, + "loss": 1.6725, + "step": 162 + }, + { + "epoch": 0.1219147344801795, + "grad_norm": 2.0082033580176546, + "learning_rate": 9.996561071522264e-06, + "loss": 1.6498, + "step": 163 + }, + { + "epoch": 0.12266267763649963, + "grad_norm": 2.640532913223682, + "learning_rate": 9.996328017125381e-06, + "loss": 1.4351, + "step": 164 + }, + { + "epoch": 0.12341062079281974, + "grad_norm": 2.574261926969864, + "learning_rate": 9.996087325402959e-06, + "loss": 1.5032, + "step": 165 + }, + { + "epoch": 0.12415856394913986, + "grad_norm": 2.3959174143523536, + "learning_rate": 9.995838996722916e-06, + "loss": 1.5534, + "step": 166 + }, + { + "epoch": 0.12490650710545999, + "grad_norm": 2.0787478905248444, + "learning_rate": 9.995583031464841e-06, + "loss": 1.6124, + "step": 167 + }, + { + "epoch": 0.1256544502617801, + "grad_norm": 3.0163617408638657, + "learning_rate": 9.995319430020004e-06, + "loss": 1.8477, + "step": 168 + }, + { + "epoch": 0.12640239341810022, + "grad_norm": 2.051187549241217, + "learning_rate": 9.995048192791341e-06, + "loss": 1.4218, + "step": 169 + }, + { + "epoch": 0.12715033657442035, + "grad_norm": 2.1950776476572758, + "learning_rate": 9.99476932019346e-06, + "loss": 1.6184, + "step": 170 + }, + { + "epoch": 0.12789827973074047, + "grad_norm": 2.2630192800692503, + "learning_rate": 9.994482812652645e-06, + "loss": 1.8702, + "step": 171 + }, + { + "epoch": 0.1286462228870606, + "grad_norm": 2.2885639200765655, + "learning_rate": 9.994188670606845e-06, + "loss": 1.7576, + "step": 172 + }, + { + "epoch": 0.1293941660433807, + "grad_norm": 2.472696404475315, + "learning_rate": 9.993886894505687e-06, + "loss": 1.6947, + "step": 173 + }, + { + "epoch": 0.13014210919970082, + "grad_norm": 2.033851279514105, + "learning_rate": 9.993577484810457e-06, + "loss": 1.3651, + "step": 174 + }, + { + "epoch": 0.13089005235602094, + "grad_norm": 2.7992567733846987, + "learning_rate": 9.993260441994116e-06, + "loss": 1.9483, + "step": 175 + }, + { + "epoch": 0.13163799551234107, + "grad_norm": 2.606840083408024, + "learning_rate": 9.992935766541295e-06, + "loss": 1.6295, + "step": 176 + }, + { + "epoch": 0.1323859386686612, + "grad_norm": 2.270802892803757, + "learning_rate": 9.992603458948282e-06, + "loss": 1.7826, + "step": 177 + }, + { + "epoch": 0.13313388182498131, + "grad_norm": 2.198077984783323, + "learning_rate": 9.992263519723046e-06, + "loss": 1.6117, + "step": 178 + }, + { + "epoch": 0.1338818249813014, + "grad_norm": 3.444680436895089, + "learning_rate": 9.99191594938521e-06, + "loss": 1.6677, + "step": 179 + }, + { + "epoch": 0.13462976813762154, + "grad_norm": 2.098179651211226, + "learning_rate": 9.991560748466068e-06, + "loss": 1.6432, + "step": 180 + }, + { + "epoch": 0.13537771129394166, + "grad_norm": 2.086548984158405, + "learning_rate": 9.99119791750857e-06, + "loss": 1.8293, + "step": 181 + }, + { + "epoch": 0.13612565445026178, + "grad_norm": 2.1177446863441505, + "learning_rate": 9.990827457067342e-06, + "loss": 1.8689, + "step": 182 + }, + { + "epoch": 0.1368735976065819, + "grad_norm": 2.036535146539503, + "learning_rate": 9.990449367708662e-06, + "loss": 1.8399, + "step": 183 + }, + { + "epoch": 0.137621540762902, + "grad_norm": 3.132874113950721, + "learning_rate": 9.990063650010473e-06, + "loss": 1.6908, + "step": 184 + }, + { + "epoch": 0.13836948391922213, + "grad_norm": 2.1871272812194955, + "learning_rate": 9.989670304562378e-06, + "loss": 1.29, + "step": 185 + }, + { + "epoch": 0.13911742707554225, + "grad_norm": 4.144341860727985, + "learning_rate": 9.98926933196564e-06, + "loss": 1.4465, + "step": 186 + }, + { + "epoch": 0.13986537023186238, + "grad_norm": 2.5515782472009683, + "learning_rate": 9.988860732833183e-06, + "loss": 1.6433, + "step": 187 + }, + { + "epoch": 0.1406133133881825, + "grad_norm": 1.9072776104244475, + "learning_rate": 9.988444507789584e-06, + "loss": 1.5092, + "step": 188 + }, + { + "epoch": 0.14136125654450263, + "grad_norm": 2.645127210743757, + "learning_rate": 9.988020657471078e-06, + "loss": 1.7851, + "step": 189 + }, + { + "epoch": 0.14210919970082272, + "grad_norm": 2.430265108535724, + "learning_rate": 9.987589182525562e-06, + "loss": 1.7082, + "step": 190 + }, + { + "epoch": 0.14285714285714285, + "grad_norm": 2.3090523081251764, + "learning_rate": 9.98715008361258e-06, + "loss": 1.4922, + "step": 191 + }, + { + "epoch": 0.14360508601346297, + "grad_norm": 2.380934239906333, + "learning_rate": 9.986703361403334e-06, + "loss": 1.7853, + "step": 192 + }, + { + "epoch": 0.1443530291697831, + "grad_norm": 2.3596313363598522, + "learning_rate": 9.98624901658068e-06, + "loss": 1.5747, + "step": 193 + }, + { + "epoch": 0.14510097232610322, + "grad_norm": 2.089220617147831, + "learning_rate": 9.98578704983912e-06, + "loss": 1.7422, + "step": 194 + }, + { + "epoch": 0.14584891548242335, + "grad_norm": 2.587615041983996, + "learning_rate": 9.985317461884815e-06, + "loss": 1.3789, + "step": 195 + }, + { + "epoch": 0.14659685863874344, + "grad_norm": 2.199893811240245, + "learning_rate": 9.984840253435569e-06, + "loss": 1.8763, + "step": 196 + }, + { + "epoch": 0.14734480179506357, + "grad_norm": 1.8576412232299557, + "learning_rate": 9.984355425220835e-06, + "loss": 1.5, + "step": 197 + }, + { + "epoch": 0.1480927449513837, + "grad_norm": 1.9223594190219857, + "learning_rate": 9.983862977981717e-06, + "loss": 1.7156, + "step": 198 + }, + { + "epoch": 0.14884068810770382, + "grad_norm": 2.1093146461207954, + "learning_rate": 9.983362912470967e-06, + "loss": 1.6894, + "step": 199 + }, + { + "epoch": 0.14958863126402394, + "grad_norm": 1.9892198831085226, + "learning_rate": 9.982855229452976e-06, + "loss": 1.726, + "step": 200 + }, + { + "epoch": 0.15033657442034407, + "grad_norm": 2.115046637628218, + "learning_rate": 9.982339929703782e-06, + "loss": 1.8922, + "step": 201 + }, + { + "epoch": 0.15108451757666416, + "grad_norm": 2.042443366937338, + "learning_rate": 9.981817014011066e-06, + "loss": 1.6781, + "step": 202 + }, + { + "epoch": 0.1518324607329843, + "grad_norm": 1.9830323614260532, + "learning_rate": 9.98128648317415e-06, + "loss": 1.6952, + "step": 203 + }, + { + "epoch": 0.1525804038893044, + "grad_norm": 2.9792357243145844, + "learning_rate": 9.980748338003999e-06, + "loss": 1.7423, + "step": 204 + }, + { + "epoch": 0.15332834704562454, + "grad_norm": 1.8454893714792557, + "learning_rate": 9.980202579323212e-06, + "loss": 1.6995, + "step": 205 + }, + { + "epoch": 0.15407629020194466, + "grad_norm": 2.434644883054974, + "learning_rate": 9.97964920796603e-06, + "loss": 1.7108, + "step": 206 + }, + { + "epoch": 0.15482423335826478, + "grad_norm": 3.675161944499911, + "learning_rate": 9.979088224778336e-06, + "loss": 1.5902, + "step": 207 + }, + { + "epoch": 0.15557217651458488, + "grad_norm": 2.112899963847331, + "learning_rate": 9.978519630617632e-06, + "loss": 1.6272, + "step": 208 + }, + { + "epoch": 0.156320119670905, + "grad_norm": 2.6654004475393296, + "learning_rate": 9.977943426353067e-06, + "loss": 1.7509, + "step": 209 + }, + { + "epoch": 0.15706806282722513, + "grad_norm": 2.3478970890596793, + "learning_rate": 9.977359612865424e-06, + "loss": 1.6382, + "step": 210 + }, + { + "epoch": 0.15781600598354525, + "grad_norm": 2.0075632426984407, + "learning_rate": 9.976768191047109e-06, + "loss": 1.9422, + "step": 211 + }, + { + "epoch": 0.15856394913986538, + "grad_norm": 2.181951676309277, + "learning_rate": 9.976169161802164e-06, + "loss": 1.9156, + "step": 212 + }, + { + "epoch": 0.1593118922961855, + "grad_norm": 1.9664963809276723, + "learning_rate": 9.975562526046257e-06, + "loss": 1.753, + "step": 213 + }, + { + "epoch": 0.1600598354525056, + "grad_norm": 2.0222760021824695, + "learning_rate": 9.974948284706684e-06, + "loss": 1.5166, + "step": 214 + }, + { + "epoch": 0.16080777860882572, + "grad_norm": 1.9403730827085452, + "learning_rate": 9.974326438722367e-06, + "loss": 1.5812, + "step": 215 + }, + { + "epoch": 0.16155572176514585, + "grad_norm": 1.9089007661880997, + "learning_rate": 9.973696989043852e-06, + "loss": 1.6375, + "step": 216 + }, + { + "epoch": 0.16230366492146597, + "grad_norm": 2.9320766052702116, + "learning_rate": 9.973059936633308e-06, + "loss": 1.3819, + "step": 217 + }, + { + "epoch": 0.1630516080777861, + "grad_norm": 18.491506882565073, + "learning_rate": 9.972415282464524e-06, + "loss": 1.5822, + "step": 218 + }, + { + "epoch": 0.1637995512341062, + "grad_norm": 7.759694284788723, + "learning_rate": 9.971763027522913e-06, + "loss": 1.8902, + "step": 219 + }, + { + "epoch": 0.16454749439042632, + "grad_norm": 2.5721575988363736, + "learning_rate": 9.971103172805504e-06, + "loss": 1.6441, + "step": 220 + }, + { + "epoch": 0.16529543754674644, + "grad_norm": 2.437691732743137, + "learning_rate": 9.97043571932094e-06, + "loss": 1.7706, + "step": 221 + }, + { + "epoch": 0.16604338070306657, + "grad_norm": 1.834286171709193, + "learning_rate": 9.969760668089485e-06, + "loss": 1.2787, + "step": 222 + }, + { + "epoch": 0.1667913238593867, + "grad_norm": 2.4496677684902854, + "learning_rate": 9.969078020143013e-06, + "loss": 1.6316, + "step": 223 + }, + { + "epoch": 0.16753926701570682, + "grad_norm": 2.328702248348561, + "learning_rate": 9.968387776525009e-06, + "loss": 1.7099, + "step": 224 + }, + { + "epoch": 0.1682872101720269, + "grad_norm": 2.1621631073692105, + "learning_rate": 9.967689938290574e-06, + "loss": 1.7157, + "step": 225 + }, + { + "epoch": 0.16903515332834704, + "grad_norm": 6.129262485274919, + "learning_rate": 9.966984506506415e-06, + "loss": 1.6395, + "step": 226 + }, + { + "epoch": 0.16978309648466716, + "grad_norm": 2.0119696542687135, + "learning_rate": 9.966271482250845e-06, + "loss": 1.7695, + "step": 227 + }, + { + "epoch": 0.1705310396409873, + "grad_norm": 2.3087904356509434, + "learning_rate": 9.965550866613782e-06, + "loss": 1.7146, + "step": 228 + }, + { + "epoch": 0.1712789827973074, + "grad_norm": 2.4709056019243016, + "learning_rate": 9.964822660696753e-06, + "loss": 1.5291, + "step": 229 + }, + { + "epoch": 0.17202692595362754, + "grad_norm": 2.6128440598542717, + "learning_rate": 9.964086865612886e-06, + "loss": 1.6186, + "step": 230 + }, + { + "epoch": 0.17277486910994763, + "grad_norm": 3.651111788773553, + "learning_rate": 9.963343482486907e-06, + "loss": 1.7901, + "step": 231 + }, + { + "epoch": 0.17352281226626776, + "grad_norm": 2.1299448534624945, + "learning_rate": 9.96259251245514e-06, + "loss": 1.6026, + "step": 232 + }, + { + "epoch": 0.17427075542258788, + "grad_norm": 2.2257757716438773, + "learning_rate": 9.96183395666551e-06, + "loss": 1.8103, + "step": 233 + }, + { + "epoch": 0.175018698578908, + "grad_norm": 1.961715078271292, + "learning_rate": 9.961067816277537e-06, + "loss": 1.5911, + "step": 234 + }, + { + "epoch": 0.17576664173522813, + "grad_norm": 2.0235936601512696, + "learning_rate": 9.960294092462332e-06, + "loss": 1.7145, + "step": 235 + }, + { + "epoch": 0.17651458489154825, + "grad_norm": 1.9942391952579746, + "learning_rate": 9.959512786402603e-06, + "loss": 1.5697, + "step": 236 + }, + { + "epoch": 0.17726252804786835, + "grad_norm": 2.051752911379157, + "learning_rate": 9.958723899292642e-06, + "loss": 1.8442, + "step": 237 + }, + { + "epoch": 0.17801047120418848, + "grad_norm": 1.925369574795835, + "learning_rate": 9.957927432338332e-06, + "loss": 1.6546, + "step": 238 + }, + { + "epoch": 0.1787584143605086, + "grad_norm": 2.207801547246269, + "learning_rate": 9.957123386757144e-06, + "loss": 1.8304, + "step": 239 + }, + { + "epoch": 0.17950635751682872, + "grad_norm": 2.5052955531563716, + "learning_rate": 9.956311763778131e-06, + "loss": 1.5441, + "step": 240 + }, + { + "epoch": 0.18025430067314885, + "grad_norm": 3.854950681500674, + "learning_rate": 9.95549256464193e-06, + "loss": 1.7898, + "step": 241 + }, + { + "epoch": 0.18100224382946897, + "grad_norm": 1.9890504948722703, + "learning_rate": 9.954665790600762e-06, + "loss": 1.817, + "step": 242 + }, + { + "epoch": 0.18175018698578907, + "grad_norm": 1.8814959603570924, + "learning_rate": 9.953831442918418e-06, + "loss": 1.7476, + "step": 243 + }, + { + "epoch": 0.1824981301421092, + "grad_norm": 2.329308352863761, + "learning_rate": 9.952989522870275e-06, + "loss": 1.7265, + "step": 244 + }, + { + "epoch": 0.18324607329842932, + "grad_norm": 2.165268450718108, + "learning_rate": 9.952140031743282e-06, + "loss": 1.6233, + "step": 245 + }, + { + "epoch": 0.18399401645474944, + "grad_norm": 1.8558404286729664, + "learning_rate": 9.951282970835959e-06, + "loss": 1.5941, + "step": 246 + }, + { + "epoch": 0.18474195961106957, + "grad_norm": 1.883740354772317, + "learning_rate": 9.950418341458398e-06, + "loss": 1.5442, + "step": 247 + }, + { + "epoch": 0.1854899027673897, + "grad_norm": 1.928654487405267, + "learning_rate": 9.949546144932265e-06, + "loss": 1.6073, + "step": 248 + }, + { + "epoch": 0.1862378459237098, + "grad_norm": 2.2715936887866315, + "learning_rate": 9.948666382590785e-06, + "loss": 1.8702, + "step": 249 + }, + { + "epoch": 0.1869857890800299, + "grad_norm": 4.017420863273423, + "learning_rate": 9.947779055778752e-06, + "loss": 1.8753, + "step": 250 + }, + { + "epoch": 0.18773373223635004, + "grad_norm": 2.7788469277150196, + "learning_rate": 9.946884165852526e-06, + "loss": 1.6032, + "step": 251 + }, + { + "epoch": 0.18848167539267016, + "grad_norm": 2.051968257645773, + "learning_rate": 9.945981714180021e-06, + "loss": 1.4651, + "step": 252 + }, + { + "epoch": 0.18922961854899029, + "grad_norm": 12.596273978356587, + "learning_rate": 9.945071702140716e-06, + "loss": 1.6049, + "step": 253 + }, + { + "epoch": 0.18997756170531038, + "grad_norm": 2.5864177546613476, + "learning_rate": 9.944154131125643e-06, + "loss": 1.8477, + "step": 254 + }, + { + "epoch": 0.1907255048616305, + "grad_norm": 2.1256175915391977, + "learning_rate": 9.943229002537391e-06, + "loss": 1.5815, + "step": 255 + }, + { + "epoch": 0.19147344801795063, + "grad_norm": 2.2881367671660513, + "learning_rate": 9.9422963177901e-06, + "loss": 1.6741, + "step": 256 + }, + { + "epoch": 0.19222139117427076, + "grad_norm": 2.6251179153199793, + "learning_rate": 9.941356078309457e-06, + "loss": 2.0234, + "step": 257 + }, + { + "epoch": 0.19296933433059088, + "grad_norm": 2.5666146990861884, + "learning_rate": 9.940408285532705e-06, + "loss": 1.5827, + "step": 258 + }, + { + "epoch": 0.193717277486911, + "grad_norm": 1.9930533027142414, + "learning_rate": 9.939452940908627e-06, + "loss": 1.607, + "step": 259 + }, + { + "epoch": 0.1944652206432311, + "grad_norm": 2.413784975204512, + "learning_rate": 9.93849004589755e-06, + "loss": 1.6537, + "step": 260 + }, + { + "epoch": 0.19521316379955123, + "grad_norm": 1.9365721002579082, + "learning_rate": 9.937519601971343e-06, + "loss": 1.6864, + "step": 261 + }, + { + "epoch": 0.19596110695587135, + "grad_norm": 2.1276403482113677, + "learning_rate": 9.936541610613417e-06, + "loss": 1.7742, + "step": 262 + }, + { + "epoch": 0.19670905011219147, + "grad_norm": 2.038329000374732, + "learning_rate": 9.935556073318718e-06, + "loss": 1.5308, + "step": 263 + }, + { + "epoch": 0.1974569932685116, + "grad_norm": 2.2501250979228784, + "learning_rate": 9.934562991593723e-06, + "loss": 1.7532, + "step": 264 + }, + { + "epoch": 0.19820493642483172, + "grad_norm": 2.0181658424883926, + "learning_rate": 9.933562366956445e-06, + "loss": 1.6226, + "step": 265 + }, + { + "epoch": 0.19895287958115182, + "grad_norm": 2.0583569374521957, + "learning_rate": 9.932554200936428e-06, + "loss": 1.6312, + "step": 266 + }, + { + "epoch": 0.19970082273747194, + "grad_norm": 5.9311324052977845, + "learning_rate": 9.931538495074744e-06, + "loss": 1.8871, + "step": 267 + }, + { + "epoch": 0.20044876589379207, + "grad_norm": 3.420639590061929, + "learning_rate": 9.930515250923985e-06, + "loss": 1.9249, + "step": 268 + }, + { + "epoch": 0.2011967090501122, + "grad_norm": 2.1093397763903177, + "learning_rate": 9.92948447004827e-06, + "loss": 1.7641, + "step": 269 + }, + { + "epoch": 0.20194465220643232, + "grad_norm": 2.7117331659007786, + "learning_rate": 9.928446154023238e-06, + "loss": 1.5329, + "step": 270 + }, + { + "epoch": 0.20269259536275244, + "grad_norm": 2.255522189051637, + "learning_rate": 9.927400304436048e-06, + "loss": 1.4368, + "step": 271 + }, + { + "epoch": 0.20344053851907254, + "grad_norm": 1.9751456058247934, + "learning_rate": 9.92634692288537e-06, + "loss": 1.6583, + "step": 272 + }, + { + "epoch": 0.20418848167539266, + "grad_norm": 2.10588215903941, + "learning_rate": 9.925286010981394e-06, + "loss": 1.6677, + "step": 273 + }, + { + "epoch": 0.2049364248317128, + "grad_norm": 2.2371772072437954, + "learning_rate": 9.924217570345814e-06, + "loss": 1.6079, + "step": 274 + }, + { + "epoch": 0.2056843679880329, + "grad_norm": 1.9575719080977834, + "learning_rate": 9.923141602611835e-06, + "loss": 1.7395, + "step": 275 + }, + { + "epoch": 0.20643231114435304, + "grad_norm": 1.8618255344208277, + "learning_rate": 9.922058109424168e-06, + "loss": 1.6011, + "step": 276 + }, + { + "epoch": 0.20718025430067316, + "grad_norm": 2.0603660845219354, + "learning_rate": 9.920967092439028e-06, + "loss": 1.7279, + "step": 277 + }, + { + "epoch": 0.20792819745699326, + "grad_norm": 2.4291120322786854, + "learning_rate": 9.919868553324131e-06, + "loss": 1.482, + "step": 278 + }, + { + "epoch": 0.20867614061331338, + "grad_norm": 1.9687953040628337, + "learning_rate": 9.918762493758688e-06, + "loss": 1.681, + "step": 279 + }, + { + "epoch": 0.2094240837696335, + "grad_norm": 5.618812999232733, + "learning_rate": 9.917648915433413e-06, + "loss": 1.7231, + "step": 280 + }, + { + "epoch": 0.21017202692595363, + "grad_norm": 4.400257374981438, + "learning_rate": 9.916527820050504e-06, + "loss": 1.7708, + "step": 281 + }, + { + "epoch": 0.21091997008227376, + "grad_norm": 3.6078630509275746, + "learning_rate": 9.915399209323653e-06, + "loss": 1.4769, + "step": 282 + }, + { + "epoch": 0.21166791323859388, + "grad_norm": 2.175500032171704, + "learning_rate": 9.914263084978041e-06, + "loss": 1.4113, + "step": 283 + }, + { + "epoch": 0.21241585639491398, + "grad_norm": 2.0584106190711378, + "learning_rate": 9.913119448750338e-06, + "loss": 1.7585, + "step": 284 + }, + { + "epoch": 0.2131637995512341, + "grad_norm": 2.5281600687208154, + "learning_rate": 9.911968302388685e-06, + "loss": 1.5339, + "step": 285 + }, + { + "epoch": 0.21391174270755423, + "grad_norm": 1.8921621454677398, + "learning_rate": 9.910809647652716e-06, + "loss": 1.8206, + "step": 286 + }, + { + "epoch": 0.21465968586387435, + "grad_norm": 2.5054859648497194, + "learning_rate": 9.909643486313533e-06, + "loss": 1.6562, + "step": 287 + }, + { + "epoch": 0.21540762902019447, + "grad_norm": 2.3947054495337845, + "learning_rate": 9.90846982015372e-06, + "loss": 1.6834, + "step": 288 + }, + { + "epoch": 0.2161555721765146, + "grad_norm": 3.1141787640640675, + "learning_rate": 9.907288650967324e-06, + "loss": 1.656, + "step": 289 + }, + { + "epoch": 0.2169035153328347, + "grad_norm": 2.1154945968526864, + "learning_rate": 9.906099980559869e-06, + "loss": 1.7306, + "step": 290 + }, + { + "epoch": 0.21765145848915482, + "grad_norm": 2.280806440658014, + "learning_rate": 9.904903810748338e-06, + "loss": 1.5458, + "step": 291 + }, + { + "epoch": 0.21839940164547494, + "grad_norm": 2.3352936260354245, + "learning_rate": 9.903700143361185e-06, + "loss": 1.4963, + "step": 292 + }, + { + "epoch": 0.21914734480179507, + "grad_norm": 1.7965045993173199, + "learning_rate": 9.902488980238322e-06, + "loss": 1.4377, + "step": 293 + }, + { + "epoch": 0.2198952879581152, + "grad_norm": 2.9718188621349397, + "learning_rate": 9.901270323231114e-06, + "loss": 1.7077, + "step": 294 + }, + { + "epoch": 0.2206432311144353, + "grad_norm": 2.3191893737824567, + "learning_rate": 9.900044174202389e-06, + "loss": 1.7496, + "step": 295 + }, + { + "epoch": 0.22139117427075541, + "grad_norm": 3.0142840711859127, + "learning_rate": 9.898810535026421e-06, + "loss": 1.5668, + "step": 296 + }, + { + "epoch": 0.22213911742707554, + "grad_norm": 1.9964003459942539, + "learning_rate": 9.897569407588936e-06, + "loss": 1.7204, + "step": 297 + }, + { + "epoch": 0.22288706058339566, + "grad_norm": 2.0649097981446585, + "learning_rate": 9.896320793787106e-06, + "loss": 1.4616, + "step": 298 + }, + { + "epoch": 0.2236350037397158, + "grad_norm": 2.077554261376278, + "learning_rate": 9.895064695529549e-06, + "loss": 1.5532, + "step": 299 + }, + { + "epoch": 0.2243829468960359, + "grad_norm": 2.30478575233273, + "learning_rate": 9.893801114736318e-06, + "loss": 1.446, + "step": 300 + }, + { + "epoch": 0.225130890052356, + "grad_norm": 1.9958798660672135, + "learning_rate": 9.892530053338909e-06, + "loss": 1.7538, + "step": 301 + }, + { + "epoch": 0.22587883320867613, + "grad_norm": 2.075399613152171, + "learning_rate": 9.891251513280249e-06, + "loss": 1.5941, + "step": 302 + }, + { + "epoch": 0.22662677636499626, + "grad_norm": 2.287656122791843, + "learning_rate": 9.889965496514702e-06, + "loss": 1.7328, + "step": 303 + }, + { + "epoch": 0.22737471952131638, + "grad_norm": 2.426104839500761, + "learning_rate": 9.888672005008054e-06, + "loss": 1.6879, + "step": 304 + }, + { + "epoch": 0.2281226626776365, + "grad_norm": 1.853431844672305, + "learning_rate": 9.887371040737523e-06, + "loss": 1.8327, + "step": 305 + }, + { + "epoch": 0.22887060583395663, + "grad_norm": 3.423850031393911, + "learning_rate": 9.886062605691744e-06, + "loss": 1.6025, + "step": 306 + }, + { + "epoch": 0.22961854899027673, + "grad_norm": 2.397972872943602, + "learning_rate": 9.884746701870778e-06, + "loss": 1.5585, + "step": 307 + }, + { + "epoch": 0.23036649214659685, + "grad_norm": 2.1016099449415044, + "learning_rate": 9.883423331286096e-06, + "loss": 1.6081, + "step": 308 + }, + { + "epoch": 0.23111443530291698, + "grad_norm": 2.217745153909582, + "learning_rate": 9.882092495960589e-06, + "loss": 1.7436, + "step": 309 + }, + { + "epoch": 0.2318623784592371, + "grad_norm": 2.1555282351859915, + "learning_rate": 9.880754197928553e-06, + "loss": 1.7367, + "step": 310 + }, + { + "epoch": 0.23261032161555723, + "grad_norm": 2.413502447622768, + "learning_rate": 9.879408439235697e-06, + "loss": 1.5165, + "step": 311 + }, + { + "epoch": 0.23335826477187735, + "grad_norm": 2.868913586764976, + "learning_rate": 9.878055221939127e-06, + "loss": 1.5056, + "step": 312 + }, + { + "epoch": 0.23410620792819745, + "grad_norm": 2.4157403770213492, + "learning_rate": 9.876694548107358e-06, + "loss": 1.6591, + "step": 313 + }, + { + "epoch": 0.23485415108451757, + "grad_norm": 2.166243582465445, + "learning_rate": 9.875326419820296e-06, + "loss": 1.7058, + "step": 314 + }, + { + "epoch": 0.2356020942408377, + "grad_norm": 9.062458145013586, + "learning_rate": 9.873950839169248e-06, + "loss": 1.6324, + "step": 315 + }, + { + "epoch": 0.23635003739715782, + "grad_norm": 1.9708875143696516, + "learning_rate": 9.87256780825691e-06, + "loss": 1.897, + "step": 316 + }, + { + "epoch": 0.23709798055347794, + "grad_norm": 2.131173655776368, + "learning_rate": 9.871177329197363e-06, + "loss": 1.432, + "step": 317 + }, + { + "epoch": 0.23784592370979807, + "grad_norm": 2.271887341161589, + "learning_rate": 9.869779404116078e-06, + "loss": 1.8253, + "step": 318 + }, + { + "epoch": 0.23859386686611816, + "grad_norm": 4.699969205450249, + "learning_rate": 9.868374035149906e-06, + "loss": 1.7724, + "step": 319 + }, + { + "epoch": 0.2393418100224383, + "grad_norm": 2.6692750366539313, + "learning_rate": 9.866961224447076e-06, + "loss": 1.8592, + "step": 320 + }, + { + "epoch": 0.2400897531787584, + "grad_norm": 2.126461134306731, + "learning_rate": 9.865540974167194e-06, + "loss": 1.5291, + "step": 321 + }, + { + "epoch": 0.24083769633507854, + "grad_norm": 1.8659703649853205, + "learning_rate": 9.864113286481237e-06, + "loss": 1.5275, + "step": 322 + }, + { + "epoch": 0.24158563949139866, + "grad_norm": 3.4260262118378617, + "learning_rate": 9.86267816357155e-06, + "loss": 1.7507, + "step": 323 + }, + { + "epoch": 0.2423335826477188, + "grad_norm": 4.4106811407700235, + "learning_rate": 9.861235607631847e-06, + "loss": 1.8477, + "step": 324 + }, + { + "epoch": 0.24308152580403888, + "grad_norm": 2.043930167861354, + "learning_rate": 9.859785620867197e-06, + "loss": 1.9827, + "step": 325 + }, + { + "epoch": 0.243829468960359, + "grad_norm": 2.136251256074407, + "learning_rate": 9.858328205494035e-06, + "loss": 1.7352, + "step": 326 + }, + { + "epoch": 0.24457741211667913, + "grad_norm": 2.1441082993404326, + "learning_rate": 9.856863363740145e-06, + "loss": 1.6893, + "step": 327 + }, + { + "epoch": 0.24532535527299926, + "grad_norm": 2.5243273607722347, + "learning_rate": 9.855391097844671e-06, + "loss": 1.8254, + "step": 328 + }, + { + "epoch": 0.24607329842931938, + "grad_norm": 2.1229625873349964, + "learning_rate": 9.853911410058097e-06, + "loss": 1.7686, + "step": 329 + }, + { + "epoch": 0.24682124158563948, + "grad_norm": 2.2031590692400957, + "learning_rate": 9.852424302642256e-06, + "loss": 1.418, + "step": 330 + }, + { + "epoch": 0.2475691847419596, + "grad_norm": 1.7920430230809576, + "learning_rate": 9.850929777870324e-06, + "loss": 1.6029, + "step": 331 + }, + { + "epoch": 0.24831712789827973, + "grad_norm": 1.9285753516081376, + "learning_rate": 9.84942783802681e-06, + "loss": 1.4453, + "step": 332 + }, + { + "epoch": 0.24906507105459985, + "grad_norm": 2.5465270070100967, + "learning_rate": 9.847918485407562e-06, + "loss": 1.6027, + "step": 333 + }, + { + "epoch": 0.24981301421091998, + "grad_norm": 3.0215442361391873, + "learning_rate": 9.846401722319759e-06, + "loss": 1.7608, + "step": 334 + }, + { + "epoch": 0.2505609573672401, + "grad_norm": 2.1231138728163126, + "learning_rate": 9.844877551081905e-06, + "loss": 1.7007, + "step": 335 + }, + { + "epoch": 0.2505609573672401, + "eval_loss": 1.6879281997680664, + "eval_runtime": 69.0776, + "eval_samples_per_second": 58.963, + "eval_steps_per_second": 3.691, + "step": 335 + }, + { + "epoch": 0.2513089005235602, + "grad_norm": 2.0957466707407555, + "learning_rate": 9.843345974023833e-06, + "loss": 1.838, + "step": 336 + }, + { + "epoch": 0.25205684367988035, + "grad_norm": 1.923684747039539, + "learning_rate": 9.841806993486686e-06, + "loss": 1.5268, + "step": 337 + }, + { + "epoch": 0.25280478683620045, + "grad_norm": 2.0829828453210038, + "learning_rate": 9.840260611822938e-06, + "loss": 1.7446, + "step": 338 + }, + { + "epoch": 0.25355272999252054, + "grad_norm": 2.4906242911108847, + "learning_rate": 9.838706831396363e-06, + "loss": 1.8179, + "step": 339 + }, + { + "epoch": 0.2543006731488407, + "grad_norm": 4.813427244228531, + "learning_rate": 9.837145654582055e-06, + "loss": 1.5972, + "step": 340 + }, + { + "epoch": 0.2550486163051608, + "grad_norm": 2.2678839198608682, + "learning_rate": 9.835577083766408e-06, + "loss": 1.716, + "step": 341 + }, + { + "epoch": 0.25579655946148094, + "grad_norm": 2.484163162489966, + "learning_rate": 9.83400112134712e-06, + "loss": 1.7748, + "step": 342 + }, + { + "epoch": 0.25654450261780104, + "grad_norm": 2.403111770698812, + "learning_rate": 9.832417769733185e-06, + "loss": 1.7655, + "step": 343 + }, + { + "epoch": 0.2572924457741212, + "grad_norm": 2.1050708638634887, + "learning_rate": 9.830827031344901e-06, + "loss": 1.7227, + "step": 344 + }, + { + "epoch": 0.2580403889304413, + "grad_norm": 2.4784806542890556, + "learning_rate": 9.829228908613846e-06, + "loss": 1.4, + "step": 345 + }, + { + "epoch": 0.2587883320867614, + "grad_norm": 3.030405128429572, + "learning_rate": 9.827623403982893e-06, + "loss": 1.4555, + "step": 346 + }, + { + "epoch": 0.25953627524308154, + "grad_norm": 2.8315249381052894, + "learning_rate": 9.826010519906195e-06, + "loss": 1.8072, + "step": 347 + }, + { + "epoch": 0.26028421839940163, + "grad_norm": 2.989267215340526, + "learning_rate": 9.824390258849187e-06, + "loss": 1.4881, + "step": 348 + }, + { + "epoch": 0.2610321615557218, + "grad_norm": 2.429572064690898, + "learning_rate": 9.822762623288584e-06, + "loss": 1.6026, + "step": 349 + }, + { + "epoch": 0.2617801047120419, + "grad_norm": 2.5540280480876607, + "learning_rate": 9.821127615712365e-06, + "loss": 1.5245, + "step": 350 + }, + { + "epoch": 0.262528047868362, + "grad_norm": 1.9074408057610583, + "learning_rate": 9.819485238619785e-06, + "loss": 1.8131, + "step": 351 + }, + { + "epoch": 0.26327599102468213, + "grad_norm": 2.581621491974913, + "learning_rate": 9.81783549452136e-06, + "loss": 1.744, + "step": 352 + }, + { + "epoch": 0.26402393418100223, + "grad_norm": 2.3137410151946676, + "learning_rate": 9.816178385938867e-06, + "loss": 1.6885, + "step": 353 + }, + { + "epoch": 0.2647718773373224, + "grad_norm": 1.9398836185777395, + "learning_rate": 9.814513915405346e-06, + "loss": 1.6998, + "step": 354 + }, + { + "epoch": 0.2655198204936425, + "grad_norm": 5.054546115753261, + "learning_rate": 9.812842085465086e-06, + "loss": 1.721, + "step": 355 + }, + { + "epoch": 0.26626776364996263, + "grad_norm": 2.5484759941269983, + "learning_rate": 9.81116289867362e-06, + "loss": 1.5288, + "step": 356 + }, + { + "epoch": 0.2670157068062827, + "grad_norm": 2.2006489223913417, + "learning_rate": 9.809476357597738e-06, + "loss": 1.6924, + "step": 357 + }, + { + "epoch": 0.2677636499626028, + "grad_norm": 2.1267387938777853, + "learning_rate": 9.807782464815463e-06, + "loss": 1.7047, + "step": 358 + }, + { + "epoch": 0.268511593118923, + "grad_norm": 2.025525674892244, + "learning_rate": 9.80608122291606e-06, + "loss": 1.5595, + "step": 359 + }, + { + "epoch": 0.26925953627524307, + "grad_norm": 3.311494119266497, + "learning_rate": 9.804372634500025e-06, + "loss": 1.5061, + "step": 360 + }, + { + "epoch": 0.2700074794315632, + "grad_norm": 2.2448146795378223, + "learning_rate": 9.802656702179088e-06, + "loss": 1.4447, + "step": 361 + }, + { + "epoch": 0.2707554225878833, + "grad_norm": 2.0358507697279284, + "learning_rate": 9.800933428576201e-06, + "loss": 1.4559, + "step": 362 + }, + { + "epoch": 0.2715033657442034, + "grad_norm": 2.4439399917356264, + "learning_rate": 9.799202816325534e-06, + "loss": 1.653, + "step": 363 + }, + { + "epoch": 0.27225130890052357, + "grad_norm": 2.426153801741017, + "learning_rate": 9.797464868072489e-06, + "loss": 1.7939, + "step": 364 + }, + { + "epoch": 0.27299925205684367, + "grad_norm": 2.2685216769093426, + "learning_rate": 9.795719586473663e-06, + "loss": 1.8455, + "step": 365 + }, + { + "epoch": 0.2737471952131638, + "grad_norm": 2.129361282734897, + "learning_rate": 9.793966974196877e-06, + "loss": 1.6248, + "step": 366 + }, + { + "epoch": 0.2744951383694839, + "grad_norm": 2.04587361014568, + "learning_rate": 9.792207033921152e-06, + "loss": 1.6682, + "step": 367 + }, + { + "epoch": 0.275243081525804, + "grad_norm": 2.663084925954292, + "learning_rate": 9.79043976833671e-06, + "loss": 1.8581, + "step": 368 + }, + { + "epoch": 0.27599102468212416, + "grad_norm": 2.3092921992776128, + "learning_rate": 9.788665180144974e-06, + "loss": 1.7117, + "step": 369 + }, + { + "epoch": 0.27673896783844426, + "grad_norm": 1.9805321912571612, + "learning_rate": 9.786883272058554e-06, + "loss": 1.9802, + "step": 370 + }, + { + "epoch": 0.2774869109947644, + "grad_norm": 2.358136027713198, + "learning_rate": 9.785094046801256e-06, + "loss": 1.4705, + "step": 371 + }, + { + "epoch": 0.2782348541510845, + "grad_norm": 2.588784519263087, + "learning_rate": 9.783297507108065e-06, + "loss": 1.4585, + "step": 372 + }, + { + "epoch": 0.27898279730740466, + "grad_norm": 8.425429354793213, + "learning_rate": 9.781493655725149e-06, + "loss": 1.8809, + "step": 373 + }, + { + "epoch": 0.27973074046372476, + "grad_norm": 1.9246596593600833, + "learning_rate": 9.779682495409858e-06, + "loss": 1.6612, + "step": 374 + }, + { + "epoch": 0.28047868362004486, + "grad_norm": 2.083789719986479, + "learning_rate": 9.777864028930705e-06, + "loss": 1.8251, + "step": 375 + }, + { + "epoch": 0.281226626776365, + "grad_norm": 2.45993012840524, + "learning_rate": 9.776038259067375e-06, + "loss": 1.5595, + "step": 376 + }, + { + "epoch": 0.2819745699326851, + "grad_norm": 2.8753660177331795, + "learning_rate": 9.774205188610721e-06, + "loss": 1.7176, + "step": 377 + }, + { + "epoch": 0.28272251308900526, + "grad_norm": 1.9927546714445858, + "learning_rate": 9.77236482036275e-06, + "loss": 1.7856, + "step": 378 + }, + { + "epoch": 0.28347045624532535, + "grad_norm": 2.0876234686248796, + "learning_rate": 9.770517157136625e-06, + "loss": 1.8629, + "step": 379 + }, + { + "epoch": 0.28421839940164545, + "grad_norm": 2.0192972677108876, + "learning_rate": 9.768662201756663e-06, + "loss": 1.9263, + "step": 380 + }, + { + "epoch": 0.2849663425579656, + "grad_norm": 2.0396204740543467, + "learning_rate": 9.766799957058324e-06, + "loss": 1.7724, + "step": 381 + }, + { + "epoch": 0.2857142857142857, + "grad_norm": 1.9855905848858864, + "learning_rate": 9.764930425888216e-06, + "loss": 1.8053, + "step": 382 + }, + { + "epoch": 0.28646222887060585, + "grad_norm": 2.8717009047941753, + "learning_rate": 9.76305361110408e-06, + "loss": 1.6357, + "step": 383 + }, + { + "epoch": 0.28721017202692595, + "grad_norm": 2.2240999742409353, + "learning_rate": 9.761169515574787e-06, + "loss": 1.8057, + "step": 384 + }, + { + "epoch": 0.2879581151832461, + "grad_norm": 1.904201589227433, + "learning_rate": 9.759278142180348e-06, + "loss": 1.5197, + "step": 385 + }, + { + "epoch": 0.2887060583395662, + "grad_norm": 1.7015292665145985, + "learning_rate": 9.757379493811892e-06, + "loss": 1.7186, + "step": 386 + }, + { + "epoch": 0.2894540014958863, + "grad_norm": 1.8972758223921051, + "learning_rate": 9.75547357337167e-06, + "loss": 1.8253, + "step": 387 + }, + { + "epoch": 0.29020194465220644, + "grad_norm": 2.0716813553709676, + "learning_rate": 9.753560383773046e-06, + "loss": 1.7349, + "step": 388 + }, + { + "epoch": 0.29094988780852654, + "grad_norm": 1.8591590033211125, + "learning_rate": 9.751639927940501e-06, + "loss": 1.5194, + "step": 389 + }, + { + "epoch": 0.2916978309648467, + "grad_norm": 2.9051870470296555, + "learning_rate": 9.749712208809617e-06, + "loss": 1.5164, + "step": 390 + }, + { + "epoch": 0.2924457741211668, + "grad_norm": 2.136812967320522, + "learning_rate": 9.747777229327085e-06, + "loss": 1.5841, + "step": 391 + }, + { + "epoch": 0.2931937172774869, + "grad_norm": 2.2767440083385146, + "learning_rate": 9.745834992450688e-06, + "loss": 1.5202, + "step": 392 + }, + { + "epoch": 0.29394166043380704, + "grad_norm": 2.0887917646160554, + "learning_rate": 9.743885501149309e-06, + "loss": 1.6468, + "step": 393 + }, + { + "epoch": 0.29468960359012714, + "grad_norm": 2.11750086197939, + "learning_rate": 9.741928758402912e-06, + "loss": 1.6054, + "step": 394 + }, + { + "epoch": 0.2954375467464473, + "grad_norm": 2.0610253209257996, + "learning_rate": 9.739964767202552e-06, + "loss": 1.639, + "step": 395 + }, + { + "epoch": 0.2961854899027674, + "grad_norm": 2.3053657756984602, + "learning_rate": 9.737993530550362e-06, + "loss": 1.7326, + "step": 396 + }, + { + "epoch": 0.29693343305908754, + "grad_norm": 2.0872790296861345, + "learning_rate": 9.736015051459551e-06, + "loss": 1.7767, + "step": 397 + }, + { + "epoch": 0.29768137621540763, + "grad_norm": 1.7200044155209657, + "learning_rate": 9.734029332954396e-06, + "loss": 1.8209, + "step": 398 + }, + { + "epoch": 0.29842931937172773, + "grad_norm": 2.102807676960359, + "learning_rate": 9.732036378070243e-06, + "loss": 1.626, + "step": 399 + }, + { + "epoch": 0.2991772625280479, + "grad_norm": 2.803376193170796, + "learning_rate": 9.730036189853497e-06, + "loss": 1.7263, + "step": 400 + }, + { + "epoch": 0.299925205684368, + "grad_norm": 1.9205553558912025, + "learning_rate": 9.728028771361625e-06, + "loss": 1.5048, + "step": 401 + }, + { + "epoch": 0.30067314884068813, + "grad_norm": 2.402169755635119, + "learning_rate": 9.726014125663136e-06, + "loss": 1.5694, + "step": 402 + }, + { + "epoch": 0.30142109199700823, + "grad_norm": 2.054494405439873, + "learning_rate": 9.723992255837596e-06, + "loss": 1.6674, + "step": 403 + }, + { + "epoch": 0.3021690351533283, + "grad_norm": 2.9563391928312917, + "learning_rate": 9.721963164975613e-06, + "loss": 1.7602, + "step": 404 + }, + { + "epoch": 0.3029169783096485, + "grad_norm": 1.8222961799595736, + "learning_rate": 9.719926856178824e-06, + "loss": 1.679, + "step": 405 + }, + { + "epoch": 0.3036649214659686, + "grad_norm": 2.1687316859914665, + "learning_rate": 9.717883332559911e-06, + "loss": 1.6606, + "step": 406 + }, + { + "epoch": 0.3044128646222887, + "grad_norm": 2.4663947175348406, + "learning_rate": 9.715832597242576e-06, + "loss": 1.6875, + "step": 407 + }, + { + "epoch": 0.3051608077786088, + "grad_norm": 1.970430238059656, + "learning_rate": 9.71377465336155e-06, + "loss": 1.5468, + "step": 408 + }, + { + "epoch": 0.3059087509349289, + "grad_norm": 2.7947192956542106, + "learning_rate": 9.71170950406258e-06, + "loss": 1.6514, + "step": 409 + }, + { + "epoch": 0.30665669409124907, + "grad_norm": 2.080551817534268, + "learning_rate": 9.709637152502426e-06, + "loss": 1.6, + "step": 410 + }, + { + "epoch": 0.30740463724756917, + "grad_norm": 2.43767280630825, + "learning_rate": 9.707557601848863e-06, + "loss": 1.7181, + "step": 411 + }, + { + "epoch": 0.3081525804038893, + "grad_norm": 3.236532873376476, + "learning_rate": 9.705470855280661e-06, + "loss": 1.6313, + "step": 412 + }, + { + "epoch": 0.3089005235602094, + "grad_norm": 2.2581352275932214, + "learning_rate": 9.703376915987601e-06, + "loss": 1.6172, + "step": 413 + }, + { + "epoch": 0.30964846671652957, + "grad_norm": 2.426891771771971, + "learning_rate": 9.701275787170447e-06, + "loss": 1.5621, + "step": 414 + }, + { + "epoch": 0.31039640987284967, + "grad_norm": 2.5420914039456344, + "learning_rate": 9.699167472040965e-06, + "loss": 1.7628, + "step": 415 + }, + { + "epoch": 0.31114435302916976, + "grad_norm": 1.9758058747781502, + "learning_rate": 9.697051973821896e-06, + "loss": 1.4877, + "step": 416 + }, + { + "epoch": 0.3118922961854899, + "grad_norm": 2.192490266162548, + "learning_rate": 9.694929295746963e-06, + "loss": 1.6621, + "step": 417 + }, + { + "epoch": 0.31264023934181, + "grad_norm": 1.572530856144008, + "learning_rate": 9.692799441060868e-06, + "loss": 1.6579, + "step": 418 + }, + { + "epoch": 0.31338818249813016, + "grad_norm": 1.899298603352148, + "learning_rate": 9.69066241301928e-06, + "loss": 1.5177, + "step": 419 + }, + { + "epoch": 0.31413612565445026, + "grad_norm": 2.549096856557414, + "learning_rate": 9.688518214888836e-06, + "loss": 1.7632, + "step": 420 + }, + { + "epoch": 0.31488406881077036, + "grad_norm": 1.9552664599596825, + "learning_rate": 9.686366849947126e-06, + "loss": 1.524, + "step": 421 + }, + { + "epoch": 0.3156320119670905, + "grad_norm": 2.0309591734311114, + "learning_rate": 9.684208321482705e-06, + "loss": 1.5851, + "step": 422 + }, + { + "epoch": 0.3163799551234106, + "grad_norm": 2.7342442584805497, + "learning_rate": 9.682042632795069e-06, + "loss": 1.5495, + "step": 423 + }, + { + "epoch": 0.31712789827973076, + "grad_norm": 1.8935259708549255, + "learning_rate": 9.679869787194664e-06, + "loss": 1.4135, + "step": 424 + }, + { + "epoch": 0.31787584143605085, + "grad_norm": 2.4018664121719757, + "learning_rate": 9.67768978800288e-06, + "loss": 1.6268, + "step": 425 + }, + { + "epoch": 0.318623784592371, + "grad_norm": 2.6716758639268345, + "learning_rate": 9.67550263855203e-06, + "loss": 1.3894, + "step": 426 + }, + { + "epoch": 0.3193717277486911, + "grad_norm": 1.898513867964313, + "learning_rate": 9.673308342185366e-06, + "loss": 1.755, + "step": 427 + }, + { + "epoch": 0.3201196709050112, + "grad_norm": 1.881082893237173, + "learning_rate": 9.671106902257065e-06, + "loss": 1.6632, + "step": 428 + }, + { + "epoch": 0.32086761406133135, + "grad_norm": 1.7893685321593156, + "learning_rate": 9.668898322132218e-06, + "loss": 1.591, + "step": 429 + }, + { + "epoch": 0.32161555721765145, + "grad_norm": 3.838697526715233, + "learning_rate": 9.666682605186834e-06, + "loss": 1.463, + "step": 430 + }, + { + "epoch": 0.3223635003739716, + "grad_norm": 1.9156434594746852, + "learning_rate": 9.664459754807832e-06, + "loss": 1.6781, + "step": 431 + }, + { + "epoch": 0.3231114435302917, + "grad_norm": 2.2246615748291734, + "learning_rate": 9.662229774393033e-06, + "loss": 1.8479, + "step": 432 + }, + { + "epoch": 0.3238593866866118, + "grad_norm": 3.443568019583727, + "learning_rate": 9.659992667351157e-06, + "loss": 1.5594, + "step": 433 + }, + { + "epoch": 0.32460732984293195, + "grad_norm": 2.680378734681492, + "learning_rate": 9.657748437101819e-06, + "loss": 1.8742, + "step": 434 + }, + { + "epoch": 0.32535527299925204, + "grad_norm": 1.7824894460839487, + "learning_rate": 9.655497087075523e-06, + "loss": 1.6628, + "step": 435 + }, + { + "epoch": 0.3261032161555722, + "grad_norm": 2.116098937753906, + "learning_rate": 9.653238620713652e-06, + "loss": 1.7567, + "step": 436 + }, + { + "epoch": 0.3268511593118923, + "grad_norm": 1.9252171014240627, + "learning_rate": 9.65097304146847e-06, + "loss": 1.6716, + "step": 437 + }, + { + "epoch": 0.3275991024682124, + "grad_norm": 2.0226653391923626, + "learning_rate": 9.648700352803117e-06, + "loss": 1.4683, + "step": 438 + }, + { + "epoch": 0.32834704562453254, + "grad_norm": 2.1688505259127147, + "learning_rate": 9.646420558191596e-06, + "loss": 1.6808, + "step": 439 + }, + { + "epoch": 0.32909498878085264, + "grad_norm": 3.2110974127194347, + "learning_rate": 9.64413366111877e-06, + "loss": 1.4469, + "step": 440 + }, + { + "epoch": 0.3298429319371728, + "grad_norm": 2.535582496026069, + "learning_rate": 9.641839665080363e-06, + "loss": 1.7121, + "step": 441 + }, + { + "epoch": 0.3305908750934929, + "grad_norm": 1.7348782019020128, + "learning_rate": 9.639538573582952e-06, + "loss": 1.4216, + "step": 442 + }, + { + "epoch": 0.33133881824981304, + "grad_norm": 2.2970258631871165, + "learning_rate": 9.637230390143953e-06, + "loss": 1.6386, + "step": 443 + }, + { + "epoch": 0.33208676140613314, + "grad_norm": 1.7746662572658098, + "learning_rate": 9.63491511829163e-06, + "loss": 1.5943, + "step": 444 + }, + { + "epoch": 0.33283470456245323, + "grad_norm": 2.085508918452454, + "learning_rate": 9.632592761565078e-06, + "loss": 1.4043, + "step": 445 + }, + { + "epoch": 0.3335826477187734, + "grad_norm": 1.9355641858637305, + "learning_rate": 9.630263323514221e-06, + "loss": 1.7173, + "step": 446 + }, + { + "epoch": 0.3343305908750935, + "grad_norm": 3.47177610501069, + "learning_rate": 9.627926807699813e-06, + "loss": 1.6256, + "step": 447 + }, + { + "epoch": 0.33507853403141363, + "grad_norm": 2.4980264141103996, + "learning_rate": 9.625583217693419e-06, + "loss": 1.5768, + "step": 448 + }, + { + "epoch": 0.33582647718773373, + "grad_norm": 2.474623174594108, + "learning_rate": 9.623232557077427e-06, + "loss": 1.588, + "step": 449 + }, + { + "epoch": 0.3365744203440538, + "grad_norm": 1.8995521994602185, + "learning_rate": 9.620874829445023e-06, + "loss": 1.6555, + "step": 450 + }, + { + "epoch": 0.337322363500374, + "grad_norm": 2.2324364507644607, + "learning_rate": 9.618510038400203e-06, + "loss": 1.6073, + "step": 451 + }, + { + "epoch": 0.3380703066566941, + "grad_norm": 2.0543358321677294, + "learning_rate": 9.616138187557758e-06, + "loss": 1.782, + "step": 452 + }, + { + "epoch": 0.3388182498130142, + "grad_norm": 2.016726971248799, + "learning_rate": 9.613759280543269e-06, + "loss": 1.6511, + "step": 453 + }, + { + "epoch": 0.3395661929693343, + "grad_norm": 1.7804589296828017, + "learning_rate": 9.611373320993104e-06, + "loss": 1.4162, + "step": 454 + }, + { + "epoch": 0.3403141361256545, + "grad_norm": 2.186325824210831, + "learning_rate": 9.60898031255441e-06, + "loss": 1.6763, + "step": 455 + }, + { + "epoch": 0.3410620792819746, + "grad_norm": 2.3869063081663078, + "learning_rate": 9.606580258885115e-06, + "loss": 1.8331, + "step": 456 + }, + { + "epoch": 0.34181002243829467, + "grad_norm": 1.8896075984044782, + "learning_rate": 9.604173163653906e-06, + "loss": 1.4246, + "step": 457 + }, + { + "epoch": 0.3425579655946148, + "grad_norm": 1.83190728839978, + "learning_rate": 9.60175903054024e-06, + "loss": 1.5147, + "step": 458 + }, + { + "epoch": 0.3433059087509349, + "grad_norm": 2.5223657065593774, + "learning_rate": 9.599337863234335e-06, + "loss": 1.4314, + "step": 459 + }, + { + "epoch": 0.34405385190725507, + "grad_norm": 5.7732597718519925, + "learning_rate": 9.596909665437155e-06, + "loss": 1.6147, + "step": 460 + }, + { + "epoch": 0.34480179506357517, + "grad_norm": 2.316162956605762, + "learning_rate": 9.594474440860413e-06, + "loss": 1.5916, + "step": 461 + }, + { + "epoch": 0.34554973821989526, + "grad_norm": 2.0701662886475147, + "learning_rate": 9.592032193226564e-06, + "loss": 1.4314, + "step": 462 + }, + { + "epoch": 0.3462976813762154, + "grad_norm": 1.9593324501816367, + "learning_rate": 9.589582926268798e-06, + "loss": 1.5605, + "step": 463 + }, + { + "epoch": 0.3470456245325355, + "grad_norm": 2.3109208123074185, + "learning_rate": 9.587126643731033e-06, + "loss": 1.7181, + "step": 464 + }, + { + "epoch": 0.34779356768885566, + "grad_norm": 3.1217231175418703, + "learning_rate": 9.584663349367913e-06, + "loss": 1.4011, + "step": 465 + }, + { + "epoch": 0.34854151084517576, + "grad_norm": 1.8135681152412975, + "learning_rate": 9.582193046944799e-06, + "loss": 1.8291, + "step": 466 + }, + { + "epoch": 0.3492894540014959, + "grad_norm": 1.9947863067112748, + "learning_rate": 9.579715740237768e-06, + "loss": 1.778, + "step": 467 + }, + { + "epoch": 0.350037397157816, + "grad_norm": 1.8992404362072661, + "learning_rate": 9.577231433033595e-06, + "loss": 1.5861, + "step": 468 + }, + { + "epoch": 0.3507853403141361, + "grad_norm": 1.8526134740665148, + "learning_rate": 9.574740129129767e-06, + "loss": 1.7769, + "step": 469 + }, + { + "epoch": 0.35153328347045626, + "grad_norm": 2.444018463691342, + "learning_rate": 9.572241832334457e-06, + "loss": 1.7824, + "step": 470 + }, + { + "epoch": 0.35228122662677636, + "grad_norm": 2.7589193293054532, + "learning_rate": 9.56973654646653e-06, + "loss": 1.6946, + "step": 471 + }, + { + "epoch": 0.3530291697830965, + "grad_norm": 1.9768234599944277, + "learning_rate": 9.567224275355538e-06, + "loss": 1.6257, + "step": 472 + }, + { + "epoch": 0.3537771129394166, + "grad_norm": 2.0440150655177405, + "learning_rate": 9.564705022841708e-06, + "loss": 1.5723, + "step": 473 + }, + { + "epoch": 0.3545250560957367, + "grad_norm": 2.0864418678730545, + "learning_rate": 9.562178792775936e-06, + "loss": 1.797, + "step": 474 + }, + { + "epoch": 0.35527299925205685, + "grad_norm": 2.3629625459030867, + "learning_rate": 9.559645589019787e-06, + "loss": 1.6344, + "step": 475 + }, + { + "epoch": 0.35602094240837695, + "grad_norm": 1.998724139589099, + "learning_rate": 9.557105415445485e-06, + "loss": 1.807, + "step": 476 + }, + { + "epoch": 0.3567688855646971, + "grad_norm": 1.912622730711547, + "learning_rate": 9.554558275935907e-06, + "loss": 1.5823, + "step": 477 + }, + { + "epoch": 0.3575168287210172, + "grad_norm": 1.7518594268090364, + "learning_rate": 9.552004174384582e-06, + "loss": 1.5803, + "step": 478 + }, + { + "epoch": 0.3582647718773373, + "grad_norm": 2.075304096982823, + "learning_rate": 9.549443114695677e-06, + "loss": 1.6299, + "step": 479 + }, + { + "epoch": 0.35901271503365745, + "grad_norm": 1.6796921736657564, + "learning_rate": 9.546875100783995e-06, + "loss": 1.2458, + "step": 480 + }, + { + "epoch": 0.35976065818997754, + "grad_norm": 2.936121979906347, + "learning_rate": 9.544300136574973e-06, + "loss": 1.9492, + "step": 481 + }, + { + "epoch": 0.3605086013462977, + "grad_norm": 2.0268636768531083, + "learning_rate": 9.541718226004667e-06, + "loss": 1.729, + "step": 482 + }, + { + "epoch": 0.3612565445026178, + "grad_norm": 1.8211621506497484, + "learning_rate": 9.539129373019755e-06, + "loss": 1.55, + "step": 483 + }, + { + "epoch": 0.36200448765893795, + "grad_norm": 2.2640458401800476, + "learning_rate": 9.536533581577526e-06, + "loss": 1.4231, + "step": 484 + }, + { + "epoch": 0.36275243081525804, + "grad_norm": 2.073725051601798, + "learning_rate": 9.533930855645872e-06, + "loss": 1.5157, + "step": 485 + }, + { + "epoch": 0.36350037397157814, + "grad_norm": 2.111807641719015, + "learning_rate": 9.531321199203293e-06, + "loss": 1.7475, + "step": 486 + }, + { + "epoch": 0.3642483171278983, + "grad_norm": 1.8368210729718748, + "learning_rate": 9.528704616238875e-06, + "loss": 1.6117, + "step": 487 + }, + { + "epoch": 0.3649962602842184, + "grad_norm": 1.8649411697562672, + "learning_rate": 9.526081110752292e-06, + "loss": 1.4504, + "step": 488 + }, + { + "epoch": 0.36574420344053854, + "grad_norm": 2.1612640952758904, + "learning_rate": 9.523450686753808e-06, + "loss": 1.7414, + "step": 489 + }, + { + "epoch": 0.36649214659685864, + "grad_norm": 2.1478145104527933, + "learning_rate": 9.520813348264252e-06, + "loss": 1.7692, + "step": 490 + }, + { + "epoch": 0.36724008975317873, + "grad_norm": 2.08388665479034, + "learning_rate": 9.518169099315028e-06, + "loss": 1.4585, + "step": 491 + }, + { + "epoch": 0.3679880329094989, + "grad_norm": 1.7695534805037116, + "learning_rate": 9.515517943948104e-06, + "loss": 1.6181, + "step": 492 + }, + { + "epoch": 0.368735976065819, + "grad_norm": 2.9512283635115986, + "learning_rate": 9.512859886216003e-06, + "loss": 1.3995, + "step": 493 + }, + { + "epoch": 0.36948391922213913, + "grad_norm": 2.1888438272854716, + "learning_rate": 9.5101949301818e-06, + "loss": 1.6286, + "step": 494 + }, + { + "epoch": 0.37023186237845923, + "grad_norm": 1.9036892334337137, + "learning_rate": 9.507523079919112e-06, + "loss": 1.7524, + "step": 495 + }, + { + "epoch": 0.3709798055347794, + "grad_norm": 4.278830445452546, + "learning_rate": 9.504844339512096e-06, + "loss": 1.5784, + "step": 496 + }, + { + "epoch": 0.3717277486910995, + "grad_norm": 1.9480552449974198, + "learning_rate": 9.502158713055444e-06, + "loss": 1.6731, + "step": 497 + }, + { + "epoch": 0.3724756918474196, + "grad_norm": 2.173423162185007, + "learning_rate": 9.499466204654371e-06, + "loss": 1.3624, + "step": 498 + }, + { + "epoch": 0.37322363500373973, + "grad_norm": 2.127281008352178, + "learning_rate": 9.496766818424612e-06, + "loss": 1.7192, + "step": 499 + }, + { + "epoch": 0.3739715781600598, + "grad_norm": 2.170788054028842, + "learning_rate": 9.494060558492416e-06, + "loss": 2.0022, + "step": 500 + }, + { + "epoch": 0.37471952131638, + "grad_norm": 1.9537718580889503, + "learning_rate": 9.491347428994536e-06, + "loss": 1.7746, + "step": 501 + }, + { + "epoch": 0.3754674644727001, + "grad_norm": 2.323471706100163, + "learning_rate": 9.488627434078232e-06, + "loss": 1.5507, + "step": 502 + }, + { + "epoch": 0.37621540762902017, + "grad_norm": 1.9373633993878732, + "learning_rate": 9.485900577901252e-06, + "loss": 1.7735, + "step": 503 + }, + { + "epoch": 0.3769633507853403, + "grad_norm": 1.7717249712290426, + "learning_rate": 9.483166864631837e-06, + "loss": 1.9296, + "step": 504 + }, + { + "epoch": 0.3777112939416604, + "grad_norm": 1.9483233073332888, + "learning_rate": 9.480426298448706e-06, + "loss": 1.7331, + "step": 505 + }, + { + "epoch": 0.37845923709798057, + "grad_norm": 2.138573565740748, + "learning_rate": 9.477678883541055e-06, + "loss": 1.5852, + "step": 506 + }, + { + "epoch": 0.37920718025430067, + "grad_norm": 1.9579233203709383, + "learning_rate": 9.47492462410855e-06, + "loss": 1.4781, + "step": 507 + }, + { + "epoch": 0.37995512341062077, + "grad_norm": 2.045771450333033, + "learning_rate": 9.472163524361317e-06, + "loss": 1.7694, + "step": 508 + }, + { + "epoch": 0.3807030665669409, + "grad_norm": 2.1044615654065852, + "learning_rate": 9.469395588519938e-06, + "loss": 1.6656, + "step": 509 + }, + { + "epoch": 0.381451009723261, + "grad_norm": 2.6394749544504448, + "learning_rate": 9.466620820815446e-06, + "loss": 1.6073, + "step": 510 + }, + { + "epoch": 0.38219895287958117, + "grad_norm": 2.1794451675781534, + "learning_rate": 9.46383922548932e-06, + "loss": 1.669, + "step": 511 + }, + { + "epoch": 0.38294689603590126, + "grad_norm": 2.3679975773386586, + "learning_rate": 9.461050806793468e-06, + "loss": 1.6107, + "step": 512 + }, + { + "epoch": 0.3836948391922214, + "grad_norm": 2.664913216154818, + "learning_rate": 9.458255568990234e-06, + "loss": 1.6733, + "step": 513 + }, + { + "epoch": 0.3844427823485415, + "grad_norm": 2.2270159467758077, + "learning_rate": 9.455453516352385e-06, + "loss": 1.8602, + "step": 514 + }, + { + "epoch": 0.3851907255048616, + "grad_norm": 3.329722799800671, + "learning_rate": 9.452644653163105e-06, + "loss": 1.7609, + "step": 515 + }, + { + "epoch": 0.38593866866118176, + "grad_norm": 1.7286190515513518, + "learning_rate": 9.449828983715985e-06, + "loss": 1.6488, + "step": 516 + }, + { + "epoch": 0.38668661181750186, + "grad_norm": 1.7435209689028321, + "learning_rate": 9.447006512315025e-06, + "loss": 1.9445, + "step": 517 + }, + { + "epoch": 0.387434554973822, + "grad_norm": 2.6261242419664295, + "learning_rate": 9.444177243274619e-06, + "loss": 1.8588, + "step": 518 + }, + { + "epoch": 0.3881824981301421, + "grad_norm": 1.954203452849246, + "learning_rate": 9.441341180919552e-06, + "loss": 1.5981, + "step": 519 + }, + { + "epoch": 0.3889304412864622, + "grad_norm": 1.935441684937223, + "learning_rate": 9.438498329584995e-06, + "loss": 1.5441, + "step": 520 + }, + { + "epoch": 0.38967838444278236, + "grad_norm": 1.860253517226172, + "learning_rate": 9.435648693616496e-06, + "loss": 1.7053, + "step": 521 + }, + { + "epoch": 0.39042632759910245, + "grad_norm": 2.5102486908228765, + "learning_rate": 9.432792277369975e-06, + "loss": 1.726, + "step": 522 + }, + { + "epoch": 0.3911742707554226, + "grad_norm": 1.97577681063729, + "learning_rate": 9.429929085211711e-06, + "loss": 1.308, + "step": 523 + }, + { + "epoch": 0.3919222139117427, + "grad_norm": 1.7521838881775988, + "learning_rate": 9.427059121518346e-06, + "loss": 1.5196, + "step": 524 + }, + { + "epoch": 0.39267015706806285, + "grad_norm": 1.9671085891581934, + "learning_rate": 9.424182390676872e-06, + "loss": 1.6966, + "step": 525 + }, + { + "epoch": 0.39341810022438295, + "grad_norm": 2.9233711325300153, + "learning_rate": 9.421298897084622e-06, + "loss": 1.6328, + "step": 526 + }, + { + "epoch": 0.39416604338070305, + "grad_norm": 2.593403460144975, + "learning_rate": 9.418408645149273e-06, + "loss": 1.6276, + "step": 527 + }, + { + "epoch": 0.3949139865370232, + "grad_norm": 1.8474610278769539, + "learning_rate": 9.415511639288825e-06, + "loss": 1.7375, + "step": 528 + }, + { + "epoch": 0.3956619296933433, + "grad_norm": 2.018566099851793, + "learning_rate": 9.412607883931608e-06, + "loss": 1.8606, + "step": 529 + }, + { + "epoch": 0.39640987284966345, + "grad_norm": 2.415059771727037, + "learning_rate": 9.409697383516263e-06, + "loss": 1.8871, + "step": 530 + }, + { + "epoch": 0.39715781600598354, + "grad_norm": 2.335891761866452, + "learning_rate": 9.40678014249175e-06, + "loss": 1.7126, + "step": 531 + }, + { + "epoch": 0.39790575916230364, + "grad_norm": 2.088185604761474, + "learning_rate": 9.403856165317322e-06, + "loss": 1.5979, + "step": 532 + }, + { + "epoch": 0.3986537023186238, + "grad_norm": 2.437173781280617, + "learning_rate": 9.40092545646254e-06, + "loss": 1.5577, + "step": 533 + }, + { + "epoch": 0.3994016454749439, + "grad_norm": 2.0266222836134973, + "learning_rate": 9.397988020407246e-06, + "loss": 1.8348, + "step": 534 + }, + { + "epoch": 0.40014958863126404, + "grad_norm": 2.0722572337611602, + "learning_rate": 9.39504386164157e-06, + "loss": 1.7095, + "step": 535 + }, + { + "epoch": 0.40089753178758414, + "grad_norm": 2.0284817293719337, + "learning_rate": 9.392092984665919e-06, + "loss": 1.7312, + "step": 536 + }, + { + "epoch": 0.4016454749439043, + "grad_norm": 1.973519421393344, + "learning_rate": 9.389135393990962e-06, + "loss": 1.5351, + "step": 537 + }, + { + "epoch": 0.4023934181002244, + "grad_norm": 1.7966449250079677, + "learning_rate": 9.38617109413764e-06, + "loss": 1.9762, + "step": 538 + }, + { + "epoch": 0.4031413612565445, + "grad_norm": 1.8831555126513535, + "learning_rate": 9.383200089637143e-06, + "loss": 1.7523, + "step": 539 + }, + { + "epoch": 0.40388930441286464, + "grad_norm": 2.148151180591269, + "learning_rate": 9.380222385030916e-06, + "loss": 1.4837, + "step": 540 + }, + { + "epoch": 0.40463724756918473, + "grad_norm": 3.5320425826545083, + "learning_rate": 9.377237984870636e-06, + "loss": 1.7614, + "step": 541 + }, + { + "epoch": 0.4053851907255049, + "grad_norm": 2.3982981598898103, + "learning_rate": 9.374246893718222e-06, + "loss": 1.9088, + "step": 542 + }, + { + "epoch": 0.406133133881825, + "grad_norm": 2.259688262994906, + "learning_rate": 9.37124911614582e-06, + "loss": 1.3372, + "step": 543 + }, + { + "epoch": 0.4068810770381451, + "grad_norm": 2.014145778352322, + "learning_rate": 9.3682446567358e-06, + "loss": 1.5415, + "step": 544 + }, + { + "epoch": 0.40762902019446523, + "grad_norm": 2.089541736009368, + "learning_rate": 9.365233520080732e-06, + "loss": 1.8798, + "step": 545 + }, + { + "epoch": 0.4083769633507853, + "grad_norm": 1.866985034566015, + "learning_rate": 9.362215710783411e-06, + "loss": 1.7811, + "step": 546 + }, + { + "epoch": 0.4091249065071055, + "grad_norm": 2.079829782988966, + "learning_rate": 9.359191233456821e-06, + "loss": 1.7008, + "step": 547 + }, + { + "epoch": 0.4098728496634256, + "grad_norm": 1.7601773714331446, + "learning_rate": 9.356160092724139e-06, + "loss": 1.8718, + "step": 548 + }, + { + "epoch": 0.4106207928197457, + "grad_norm": 1.9578881509891957, + "learning_rate": 9.35312229321873e-06, + "loss": 1.7503, + "step": 549 + }, + { + "epoch": 0.4113687359760658, + "grad_norm": 2.578160352278612, + "learning_rate": 9.350077839584139e-06, + "loss": 1.5913, + "step": 550 + }, + { + "epoch": 0.4121166791323859, + "grad_norm": 1.9493525494926824, + "learning_rate": 9.347026736474077e-06, + "loss": 1.345, + "step": 551 + }, + { + "epoch": 0.4128646222887061, + "grad_norm": 1.9131426482126372, + "learning_rate": 9.343968988552427e-06, + "loss": 1.5039, + "step": 552 + }, + { + "epoch": 0.41361256544502617, + "grad_norm": 1.7652476453258712, + "learning_rate": 9.34090460049322e-06, + "loss": 1.4418, + "step": 553 + }, + { + "epoch": 0.4143605086013463, + "grad_norm": 1.8901437586953647, + "learning_rate": 9.337833576980647e-06, + "loss": 1.8592, + "step": 554 + }, + { + "epoch": 0.4151084517576664, + "grad_norm": 2.391435393685785, + "learning_rate": 9.334755922709032e-06, + "loss": 1.5797, + "step": 555 + }, + { + "epoch": 0.4158563949139865, + "grad_norm": 1.9343010136441545, + "learning_rate": 9.33167164238284e-06, + "loss": 1.3067, + "step": 556 + }, + { + "epoch": 0.41660433807030667, + "grad_norm": 1.9299606263224045, + "learning_rate": 9.328580740716668e-06, + "loss": 1.591, + "step": 557 + }, + { + "epoch": 0.41735228122662676, + "grad_norm": 1.8179505596930905, + "learning_rate": 9.325483222435225e-06, + "loss": 1.5586, + "step": 558 + }, + { + "epoch": 0.4181002243829469, + "grad_norm": 2.0873853249367973, + "learning_rate": 9.322379092273345e-06, + "loss": 1.9041, + "step": 559 + }, + { + "epoch": 0.418848167539267, + "grad_norm": 1.8252190322405553, + "learning_rate": 9.319268354975958e-06, + "loss": 1.5732, + "step": 560 + }, + { + "epoch": 0.4195961106955871, + "grad_norm": 2.1121218686149517, + "learning_rate": 9.316151015298103e-06, + "loss": 1.4726, + "step": 561 + }, + { + "epoch": 0.42034405385190726, + "grad_norm": 1.8461040157069726, + "learning_rate": 9.313027078004903e-06, + "loss": 1.7256, + "step": 562 + }, + { + "epoch": 0.42109199700822736, + "grad_norm": 2.003276132104976, + "learning_rate": 9.309896547871576e-06, + "loss": 1.773, + "step": 563 + }, + { + "epoch": 0.4218399401645475, + "grad_norm": 2.351995149132236, + "learning_rate": 9.306759429683404e-06, + "loss": 1.5638, + "step": 564 + }, + { + "epoch": 0.4225878833208676, + "grad_norm": 1.8909753929566673, + "learning_rate": 9.303615728235753e-06, + "loss": 1.4578, + "step": 565 + }, + { + "epoch": 0.42333582647718776, + "grad_norm": 2.012569973001156, + "learning_rate": 9.300465448334044e-06, + "loss": 1.4031, + "step": 566 + }, + { + "epoch": 0.42408376963350786, + "grad_norm": 2.0578828344518767, + "learning_rate": 9.297308594793757e-06, + "loss": 1.891, + "step": 567 + }, + { + "epoch": 0.42483171278982795, + "grad_norm": 1.9066610769532237, + "learning_rate": 9.29414517244042e-06, + "loss": 1.6934, + "step": 568 + }, + { + "epoch": 0.4255796559461481, + "grad_norm": 2.1035515294465066, + "learning_rate": 9.2909751861096e-06, + "loss": 1.6521, + "step": 569 + }, + { + "epoch": 0.4263275991024682, + "grad_norm": 1.932586083789388, + "learning_rate": 9.287798640646899e-06, + "loss": 1.8997, + "step": 570 + }, + { + "epoch": 0.42707554225878835, + "grad_norm": 1.9771770121431025, + "learning_rate": 9.284615540907947e-06, + "loss": 1.5117, + "step": 571 + }, + { + "epoch": 0.42782348541510845, + "grad_norm": 2.2632104803930733, + "learning_rate": 9.281425891758392e-06, + "loss": 1.6276, + "step": 572 + }, + { + "epoch": 0.42857142857142855, + "grad_norm": 1.7303888003815362, + "learning_rate": 9.278229698073889e-06, + "loss": 1.7269, + "step": 573 + }, + { + "epoch": 0.4293193717277487, + "grad_norm": 2.0113237646006046, + "learning_rate": 9.275026964740101e-06, + "loss": 1.4722, + "step": 574 + }, + { + "epoch": 0.4300673148840688, + "grad_norm": 1.8731031859006366, + "learning_rate": 9.271817696652688e-06, + "loss": 1.8367, + "step": 575 + }, + { + "epoch": 0.43081525804038895, + "grad_norm": 2.5266591264461837, + "learning_rate": 9.2686018987173e-06, + "loss": 1.6248, + "step": 576 + }, + { + "epoch": 0.43156320119670905, + "grad_norm": 1.7038626082739727, + "learning_rate": 9.265379575849561e-06, + "loss": 1.667, + "step": 577 + }, + { + "epoch": 0.4323111443530292, + "grad_norm": 2.0141209289915465, + "learning_rate": 9.262150732975078e-06, + "loss": 1.6481, + "step": 578 + }, + { + "epoch": 0.4330590875093493, + "grad_norm": 3.578017656397911, + "learning_rate": 9.258915375029419e-06, + "loss": 1.7041, + "step": 579 + }, + { + "epoch": 0.4338070306656694, + "grad_norm": 2.046975650374406, + "learning_rate": 9.255673506958114e-06, + "loss": 1.8068, + "step": 580 + }, + { + "epoch": 0.43455497382198954, + "grad_norm": 2.2315983599224536, + "learning_rate": 9.252425133716639e-06, + "loss": 1.4744, + "step": 581 + }, + { + "epoch": 0.43530291697830964, + "grad_norm": 2.4250724139020594, + "learning_rate": 9.249170260270422e-06, + "loss": 1.7192, + "step": 582 + }, + { + "epoch": 0.4360508601346298, + "grad_norm": 1.8721908647251149, + "learning_rate": 9.245908891594818e-06, + "loss": 1.5784, + "step": 583 + }, + { + "epoch": 0.4367988032909499, + "grad_norm": 2.3223641267342443, + "learning_rate": 9.242641032675118e-06, + "loss": 1.4692, + "step": 584 + }, + { + "epoch": 0.43754674644727, + "grad_norm": 1.9911919290100533, + "learning_rate": 9.23936668850653e-06, + "loss": 1.5693, + "step": 585 + }, + { + "epoch": 0.43829468960359014, + "grad_norm": 1.9666899695043338, + "learning_rate": 9.236085864094178e-06, + "loss": 1.5378, + "step": 586 + }, + { + "epoch": 0.43904263275991023, + "grad_norm": 2.2078041082919326, + "learning_rate": 9.232798564453087e-06, + "loss": 1.7456, + "step": 587 + }, + { + "epoch": 0.4397905759162304, + "grad_norm": 1.8529504690651597, + "learning_rate": 9.229504794608182e-06, + "loss": 1.5425, + "step": 588 + }, + { + "epoch": 0.4405385190725505, + "grad_norm": 2.9943470879067493, + "learning_rate": 9.226204559594284e-06, + "loss": 1.7388, + "step": 589 + }, + { + "epoch": 0.4412864622288706, + "grad_norm": 1.8306432353742277, + "learning_rate": 9.22289786445609e-06, + "loss": 1.4551, + "step": 590 + }, + { + "epoch": 0.44203440538519073, + "grad_norm": 2.0409940296640277, + "learning_rate": 9.219584714248168e-06, + "loss": 1.8065, + "step": 591 + }, + { + "epoch": 0.44278234854151083, + "grad_norm": 2.0881029719093727, + "learning_rate": 9.216265114034964e-06, + "loss": 1.6195, + "step": 592 + }, + { + "epoch": 0.443530291697831, + "grad_norm": 1.833789852606214, + "learning_rate": 9.212939068890778e-06, + "loss": 1.8474, + "step": 593 + }, + { + "epoch": 0.4442782348541511, + "grad_norm": 2.5551210659680357, + "learning_rate": 9.209606583899759e-06, + "loss": 1.4083, + "step": 594 + }, + { + "epoch": 0.44502617801047123, + "grad_norm": 2.2174973757354963, + "learning_rate": 9.206267664155906e-06, + "loss": 1.6969, + "step": 595 + }, + { + "epoch": 0.4457741211667913, + "grad_norm": 3.1316900098848475, + "learning_rate": 9.202922314763049e-06, + "loss": 1.7891, + "step": 596 + }, + { + "epoch": 0.4465220643231114, + "grad_norm": 1.8908810384432184, + "learning_rate": 9.199570540834847e-06, + "loss": 1.6869, + "step": 597 + }, + { + "epoch": 0.4472700074794316, + "grad_norm": 1.7452028573586955, + "learning_rate": 9.19621234749478e-06, + "loss": 1.6319, + "step": 598 + }, + { + "epoch": 0.44801795063575167, + "grad_norm": 2.298868938590296, + "learning_rate": 9.192847739876142e-06, + "loss": 1.5673, + "step": 599 + }, + { + "epoch": 0.4487658937920718, + "grad_norm": 2.3252289783137075, + "learning_rate": 9.189476723122034e-06, + "loss": 1.6039, + "step": 600 + }, + { + "epoch": 0.4495138369483919, + "grad_norm": 2.7160499232437387, + "learning_rate": 9.186099302385345e-06, + "loss": 1.8018, + "step": 601 + }, + { + "epoch": 0.450261780104712, + "grad_norm": 2.3346932833437704, + "learning_rate": 9.182715482828764e-06, + "loss": 1.7532, + "step": 602 + }, + { + "epoch": 0.45100972326103217, + "grad_norm": 1.8866519239999724, + "learning_rate": 9.17932526962475e-06, + "loss": 1.7261, + "step": 603 + }, + { + "epoch": 0.45175766641735227, + "grad_norm": 1.7356438759915245, + "learning_rate": 9.175928667955547e-06, + "loss": 1.6818, + "step": 604 + }, + { + "epoch": 0.4525056095736724, + "grad_norm": 2.025769424722977, + "learning_rate": 9.172525683013154e-06, + "loss": 1.7452, + "step": 605 + }, + { + "epoch": 0.4532535527299925, + "grad_norm": 2.6168045082952074, + "learning_rate": 9.169116319999336e-06, + "loss": 1.7309, + "step": 606 + }, + { + "epoch": 0.45400149588631267, + "grad_norm": 1.8655328391171715, + "learning_rate": 9.1657005841256e-06, + "loss": 1.6712, + "step": 607 + }, + { + "epoch": 0.45474943904263276, + "grad_norm": 1.7714789572355463, + "learning_rate": 9.162278480613203e-06, + "loss": 1.7047, + "step": 608 + }, + { + "epoch": 0.45549738219895286, + "grad_norm": 2.0367568260515934, + "learning_rate": 9.158850014693123e-06, + "loss": 1.8608, + "step": 609 + }, + { + "epoch": 0.456245325355273, + "grad_norm": 1.912733179744595, + "learning_rate": 9.155415191606074e-06, + "loss": 1.5081, + "step": 610 + }, + { + "epoch": 0.4569932685115931, + "grad_norm": 2.1414413328459903, + "learning_rate": 9.151974016602486e-06, + "loss": 1.6907, + "step": 611 + }, + { + "epoch": 0.45774121166791326, + "grad_norm": 2.0190636422371906, + "learning_rate": 9.148526494942496e-06, + "loss": 1.7158, + "step": 612 + }, + { + "epoch": 0.45848915482423336, + "grad_norm": 1.796592516840325, + "learning_rate": 9.145072631895942e-06, + "loss": 1.4643, + "step": 613 + }, + { + "epoch": 0.45923709798055345, + "grad_norm": 2.260412528250619, + "learning_rate": 9.141612432742357e-06, + "loss": 1.7033, + "step": 614 + }, + { + "epoch": 0.4599850411368736, + "grad_norm": 2.0703825717568223, + "learning_rate": 9.138145902770959e-06, + "loss": 1.6536, + "step": 615 + }, + { + "epoch": 0.4607329842931937, + "grad_norm": 3.43367271376812, + "learning_rate": 9.134673047280644e-06, + "loss": 1.7501, + "step": 616 + }, + { + "epoch": 0.46148092744951386, + "grad_norm": 1.8861548329102271, + "learning_rate": 9.131193871579975e-06, + "loss": 1.6005, + "step": 617 + }, + { + "epoch": 0.46222887060583395, + "grad_norm": 2.824921508314911, + "learning_rate": 9.127708380987177e-06, + "loss": 1.6078, + "step": 618 + }, + { + "epoch": 0.46297681376215405, + "grad_norm": 1.7833346406583754, + "learning_rate": 9.12421658083013e-06, + "loss": 1.5595, + "step": 619 + }, + { + "epoch": 0.4637247569184742, + "grad_norm": 2.3580100538825564, + "learning_rate": 9.120718476446354e-06, + "loss": 1.4192, + "step": 620 + }, + { + "epoch": 0.4644727000747943, + "grad_norm": 1.7679102590447409, + "learning_rate": 9.11721407318301e-06, + "loss": 1.3914, + "step": 621 + }, + { + "epoch": 0.46522064323111445, + "grad_norm": 3.6259592801974767, + "learning_rate": 9.113703376396885e-06, + "loss": 1.621, + "step": 622 + }, + { + "epoch": 0.46596858638743455, + "grad_norm": 1.985900443561016, + "learning_rate": 9.110186391454389e-06, + "loss": 1.6704, + "step": 623 + }, + { + "epoch": 0.4667165295437547, + "grad_norm": 2.0603936726818604, + "learning_rate": 9.106663123731538e-06, + "loss": 1.7007, + "step": 624 + }, + { + "epoch": 0.4674644727000748, + "grad_norm": 1.9549500751511233, + "learning_rate": 9.103133578613959e-06, + "loss": 1.5717, + "step": 625 + }, + { + "epoch": 0.4682124158563949, + "grad_norm": 2.295565327403211, + "learning_rate": 9.09959776149687e-06, + "loss": 1.5725, + "step": 626 + }, + { + "epoch": 0.46896035901271504, + "grad_norm": 2.145444964638831, + "learning_rate": 9.096055677785078e-06, + "loss": 1.6517, + "step": 627 + }, + { + "epoch": 0.46970830216903514, + "grad_norm": 2.132321331053628, + "learning_rate": 9.092507332892968e-06, + "loss": 1.6875, + "step": 628 + }, + { + "epoch": 0.4704562453253553, + "grad_norm": 1.8786779398102642, + "learning_rate": 9.088952732244497e-06, + "loss": 1.7675, + "step": 629 + }, + { + "epoch": 0.4712041884816754, + "grad_norm": 3.3600120067188826, + "learning_rate": 9.085391881273182e-06, + "loss": 1.7326, + "step": 630 + }, + { + "epoch": 0.4719521316379955, + "grad_norm": 2.465689414078988, + "learning_rate": 9.081824785422099e-06, + "loss": 1.5753, + "step": 631 + }, + { + "epoch": 0.47270007479431564, + "grad_norm": 1.852186642176982, + "learning_rate": 9.078251450143866e-06, + "loss": 1.8303, + "step": 632 + }, + { + "epoch": 0.47344801795063574, + "grad_norm": 2.055509884943897, + "learning_rate": 9.074671880900637e-06, + "loss": 1.7641, + "step": 633 + }, + { + "epoch": 0.4741959611069559, + "grad_norm": 1.8415356829002765, + "learning_rate": 9.071086083164099e-06, + "loss": 1.5496, + "step": 634 + }, + { + "epoch": 0.474943904263276, + "grad_norm": 2.1049815906854934, + "learning_rate": 9.06749406241546e-06, + "loss": 1.8797, + "step": 635 + }, + { + "epoch": 0.47569184741959614, + "grad_norm": 1.911307503575673, + "learning_rate": 9.063895824145435e-06, + "loss": 1.5454, + "step": 636 + }, + { + "epoch": 0.47643979057591623, + "grad_norm": 1.920489923259322, + "learning_rate": 9.060291373854252e-06, + "loss": 1.6532, + "step": 637 + }, + { + "epoch": 0.47718773373223633, + "grad_norm": 1.776829452763796, + "learning_rate": 9.056680717051623e-06, + "loss": 1.56, + "step": 638 + }, + { + "epoch": 0.4779356768885565, + "grad_norm": 1.8711468026657416, + "learning_rate": 9.05306385925676e-06, + "loss": 1.6343, + "step": 639 + }, + { + "epoch": 0.4786836200448766, + "grad_norm": 1.7633632649672883, + "learning_rate": 9.04944080599834e-06, + "loss": 1.8776, + "step": 640 + }, + { + "epoch": 0.47943156320119673, + "grad_norm": 1.9445907542928076, + "learning_rate": 9.045811562814526e-06, + "loss": 1.6103, + "step": 641 + }, + { + "epoch": 0.4801795063575168, + "grad_norm": 1.9615338954800459, + "learning_rate": 9.042176135252927e-06, + "loss": 1.7251, + "step": 642 + }, + { + "epoch": 0.4809274495138369, + "grad_norm": 2.326742338003008, + "learning_rate": 9.038534528870618e-06, + "loss": 1.5343, + "step": 643 + }, + { + "epoch": 0.4816753926701571, + "grad_norm": 2.226933314701637, + "learning_rate": 9.034886749234112e-06, + "loss": 1.5534, + "step": 644 + }, + { + "epoch": 0.4824233358264772, + "grad_norm": 1.7763813724877247, + "learning_rate": 9.03123280191936e-06, + "loss": 1.6936, + "step": 645 + }, + { + "epoch": 0.4831712789827973, + "grad_norm": 1.777477777973393, + "learning_rate": 9.027572692511739e-06, + "loss": 1.8508, + "step": 646 + }, + { + "epoch": 0.4839192221391174, + "grad_norm": 2.077534184380355, + "learning_rate": 9.02390642660605e-06, + "loss": 1.6757, + "step": 647 + }, + { + "epoch": 0.4846671652954376, + "grad_norm": 2.3629003345077524, + "learning_rate": 9.020234009806504e-06, + "loss": 1.8243, + "step": 648 + }, + { + "epoch": 0.48541510845175767, + "grad_norm": 2.042409869378677, + "learning_rate": 9.016555447726704e-06, + "loss": 1.6914, + "step": 649 + }, + { + "epoch": 0.48616305160807777, + "grad_norm": 1.8352046670672768, + "learning_rate": 9.012870745989663e-06, + "loss": 1.7572, + "step": 650 + }, + { + "epoch": 0.4869109947643979, + "grad_norm": 1.734096266429357, + "learning_rate": 9.009179910227767e-06, + "loss": 1.6938, + "step": 651 + }, + { + "epoch": 0.487658937920718, + "grad_norm": 2.627207317676917, + "learning_rate": 9.005482946082784e-06, + "loss": 1.3674, + "step": 652 + }, + { + "epoch": 0.48840688107703817, + "grad_norm": 3.8226384432386418, + "learning_rate": 9.001779859205846e-06, + "loss": 1.456, + "step": 653 + }, + { + "epoch": 0.48915482423335827, + "grad_norm": 1.8609019807405238, + "learning_rate": 8.998070655257447e-06, + "loss": 1.6319, + "step": 654 + }, + { + "epoch": 0.48990276738967836, + "grad_norm": 1.851512677034953, + "learning_rate": 8.994355339907429e-06, + "loss": 1.7763, + "step": 655 + }, + { + "epoch": 0.4906507105459985, + "grad_norm": 2.218296320526967, + "learning_rate": 8.990633918834978e-06, + "loss": 1.6602, + "step": 656 + }, + { + "epoch": 0.4913986537023186, + "grad_norm": 1.6645255951658595, + "learning_rate": 8.986906397728617e-06, + "loss": 1.5054, + "step": 657 + }, + { + "epoch": 0.49214659685863876, + "grad_norm": 1.7378691796138988, + "learning_rate": 8.98317278228618e-06, + "loss": 1.5755, + "step": 658 + }, + { + "epoch": 0.49289454001495886, + "grad_norm": 1.8196827047730064, + "learning_rate": 8.979433078214834e-06, + "loss": 1.7404, + "step": 659 + }, + { + "epoch": 0.49364248317127896, + "grad_norm": 2.3681622872922814, + "learning_rate": 8.975687291231041e-06, + "loss": 1.7975, + "step": 660 + }, + { + "epoch": 0.4943904263275991, + "grad_norm": 1.9611888642261228, + "learning_rate": 8.971935427060563e-06, + "loss": 1.4111, + "step": 661 + }, + { + "epoch": 0.4951383694839192, + "grad_norm": 1.8949192270102413, + "learning_rate": 8.968177491438457e-06, + "loss": 1.6398, + "step": 662 + }, + { + "epoch": 0.49588631264023936, + "grad_norm": 1.9706164608797014, + "learning_rate": 8.964413490109056e-06, + "loss": 1.682, + "step": 663 + }, + { + "epoch": 0.49663425579655945, + "grad_norm": 2.1406110347147855, + "learning_rate": 8.960643428825965e-06, + "loss": 1.6385, + "step": 664 + }, + { + "epoch": 0.4973821989528796, + "grad_norm": 1.9609193669231069, + "learning_rate": 8.956867313352055e-06, + "loss": 1.6044, + "step": 665 + }, + { + "epoch": 0.4981301421091997, + "grad_norm": 1.7580508797381957, + "learning_rate": 8.95308514945945e-06, + "loss": 1.4487, + "step": 666 + }, + { + "epoch": 0.4988780852655198, + "grad_norm": 2.072316480243733, + "learning_rate": 8.949296942929515e-06, + "loss": 1.3687, + "step": 667 + }, + { + "epoch": 0.49962602842183995, + "grad_norm": 2.8538637599122842, + "learning_rate": 8.94550269955286e-06, + "loss": 1.6923, + "step": 668 + }, + { + "epoch": 0.5003739715781601, + "grad_norm": 1.9299543862957045, + "learning_rate": 8.94170242512932e-06, + "loss": 1.6789, + "step": 669 + }, + { + "epoch": 0.5011219147344802, + "grad_norm": 2.308237117843891, + "learning_rate": 8.937896125467941e-06, + "loss": 1.5648, + "step": 670 + }, + { + "epoch": 0.5011219147344802, + "eval_loss": 1.6755473613739014, + "eval_runtime": 68.8171, + "eval_samples_per_second": 59.186, + "eval_steps_per_second": 3.705, + "step": 670 + }, + { + "epoch": 0.5018698578908003, + "grad_norm": 2.821075068214382, + "learning_rate": 8.934083806386995e-06, + "loss": 1.4394, + "step": 671 + }, + { + "epoch": 0.5026178010471204, + "grad_norm": 2.192199396793432, + "learning_rate": 8.930265473713939e-06, + "loss": 1.6489, + "step": 672 + }, + { + "epoch": 0.5033657442034405, + "grad_norm": 2.1222626491096936, + "learning_rate": 8.92644113328543e-06, + "loss": 1.7616, + "step": 673 + }, + { + "epoch": 0.5041136873597607, + "grad_norm": 1.9553128334559169, + "learning_rate": 8.922610790947312e-06, + "loss": 1.7778, + "step": 674 + }, + { + "epoch": 0.5048616305160808, + "grad_norm": 2.3260922696740796, + "learning_rate": 8.918774452554595e-06, + "loss": 1.7897, + "step": 675 + }, + { + "epoch": 0.5056095736724009, + "grad_norm": 1.9660213322268338, + "learning_rate": 8.91493212397146e-06, + "loss": 1.5902, + "step": 676 + }, + { + "epoch": 0.506357516828721, + "grad_norm": 1.6902489136798828, + "learning_rate": 8.911083811071244e-06, + "loss": 1.6719, + "step": 677 + }, + { + "epoch": 0.5071054599850411, + "grad_norm": 1.8875175367443193, + "learning_rate": 8.90722951973643e-06, + "loss": 1.4526, + "step": 678 + }, + { + "epoch": 0.5078534031413613, + "grad_norm": 2.5047072160896047, + "learning_rate": 8.90336925585864e-06, + "loss": 1.7603, + "step": 679 + }, + { + "epoch": 0.5086013462976814, + "grad_norm": 2.4433227414982555, + "learning_rate": 8.899503025338628e-06, + "loss": 1.4028, + "step": 680 + }, + { + "epoch": 0.5093492894540015, + "grad_norm": 2.640618917428092, + "learning_rate": 8.895630834086264e-06, + "loss": 1.6409, + "step": 681 + }, + { + "epoch": 0.5100972326103216, + "grad_norm": 1.8553151588024333, + "learning_rate": 8.891752688020532e-06, + "loss": 1.5755, + "step": 682 + }, + { + "epoch": 0.5108451757666418, + "grad_norm": 2.7607451010340758, + "learning_rate": 8.88786859306952e-06, + "loss": 1.24, + "step": 683 + }, + { + "epoch": 0.5115931189229619, + "grad_norm": 2.3393666997500393, + "learning_rate": 8.883978555170405e-06, + "loss": 1.8735, + "step": 684 + }, + { + "epoch": 0.512341062079282, + "grad_norm": 1.8549855873120307, + "learning_rate": 8.880082580269454e-06, + "loss": 1.6601, + "step": 685 + }, + { + "epoch": 0.5130890052356021, + "grad_norm": 1.9922061581586628, + "learning_rate": 8.876180674322006e-06, + "loss": 1.7347, + "step": 686 + }, + { + "epoch": 0.5138369483919222, + "grad_norm": 2.226028234596247, + "learning_rate": 8.872272843292465e-06, + "loss": 1.9811, + "step": 687 + }, + { + "epoch": 0.5145848915482424, + "grad_norm": 2.1797086709609195, + "learning_rate": 8.868359093154292e-06, + "loss": 1.595, + "step": 688 + }, + { + "epoch": 0.5153328347045625, + "grad_norm": 1.81759870198025, + "learning_rate": 8.864439429890001e-06, + "loss": 1.6904, + "step": 689 + }, + { + "epoch": 0.5160807778608826, + "grad_norm": 1.745711058566467, + "learning_rate": 8.86051385949114e-06, + "loss": 1.872, + "step": 690 + }, + { + "epoch": 0.5168287210172027, + "grad_norm": 2.048402651792811, + "learning_rate": 8.856582387958286e-06, + "loss": 1.4921, + "step": 691 + }, + { + "epoch": 0.5175766641735228, + "grad_norm": 2.0249627373923866, + "learning_rate": 8.852645021301041e-06, + "loss": 1.5702, + "step": 692 + }, + { + "epoch": 0.518324607329843, + "grad_norm": 1.93809092044547, + "learning_rate": 8.84870176553801e-06, + "loss": 1.5585, + "step": 693 + }, + { + "epoch": 0.5190725504861631, + "grad_norm": 1.844088464867954, + "learning_rate": 8.84475262669681e-06, + "loss": 1.6438, + "step": 694 + }, + { + "epoch": 0.5198204936424832, + "grad_norm": 1.995895787999922, + "learning_rate": 8.840797610814046e-06, + "loss": 1.5595, + "step": 695 + }, + { + "epoch": 0.5205684367988033, + "grad_norm": 2.590154684151374, + "learning_rate": 8.836836723935304e-06, + "loss": 1.6736, + "step": 696 + }, + { + "epoch": 0.5213163799551234, + "grad_norm": 2.2374559425238294, + "learning_rate": 8.832869972115148e-06, + "loss": 1.6354, + "step": 697 + }, + { + "epoch": 0.5220643231114436, + "grad_norm": 2.2054392101657045, + "learning_rate": 8.828897361417106e-06, + "loss": 1.5851, + "step": 698 + }, + { + "epoch": 0.5228122662677637, + "grad_norm": 2.0970589857765543, + "learning_rate": 8.82491889791366e-06, + "loss": 1.7063, + "step": 699 + }, + { + "epoch": 0.5235602094240838, + "grad_norm": 2.1673522393585953, + "learning_rate": 8.820934587686247e-06, + "loss": 1.4253, + "step": 700 + }, + { + "epoch": 0.5243081525804039, + "grad_norm": 1.8846548363752142, + "learning_rate": 8.816944436825227e-06, + "loss": 1.7571, + "step": 701 + }, + { + "epoch": 0.525056095736724, + "grad_norm": 1.9096416446478852, + "learning_rate": 8.812948451429898e-06, + "loss": 1.5534, + "step": 702 + }, + { + "epoch": 0.5258040388930442, + "grad_norm": 1.786819560031535, + "learning_rate": 8.808946637608476e-06, + "loss": 1.6694, + "step": 703 + }, + { + "epoch": 0.5265519820493643, + "grad_norm": 2.5721083863592424, + "learning_rate": 8.804939001478084e-06, + "loss": 1.7734, + "step": 704 + }, + { + "epoch": 0.5272999252056844, + "grad_norm": 1.949773659367809, + "learning_rate": 8.800925549164742e-06, + "loss": 1.7352, + "step": 705 + }, + { + "epoch": 0.5280478683620045, + "grad_norm": 2.1906435965530244, + "learning_rate": 8.796906286803366e-06, + "loss": 1.7231, + "step": 706 + }, + { + "epoch": 0.5287958115183246, + "grad_norm": 1.9104960726349647, + "learning_rate": 8.792881220537752e-06, + "loss": 1.5363, + "step": 707 + }, + { + "epoch": 0.5295437546746448, + "grad_norm": 1.8525833104894254, + "learning_rate": 8.788850356520567e-06, + "loss": 1.7018, + "step": 708 + }, + { + "epoch": 0.5302916978309649, + "grad_norm": 1.9326990832720403, + "learning_rate": 8.784813700913336e-06, + "loss": 1.5239, + "step": 709 + }, + { + "epoch": 0.531039640987285, + "grad_norm": 1.7739839837620432, + "learning_rate": 8.780771259886447e-06, + "loss": 1.6808, + "step": 710 + }, + { + "epoch": 0.531787584143605, + "grad_norm": 1.8535205512797215, + "learning_rate": 8.776723039619122e-06, + "loss": 1.5891, + "step": 711 + }, + { + "epoch": 0.5325355272999253, + "grad_norm": 2.091380265881628, + "learning_rate": 8.77266904629942e-06, + "loss": 1.8294, + "step": 712 + }, + { + "epoch": 0.5332834704562454, + "grad_norm": 1.9976817006051297, + "learning_rate": 8.768609286124229e-06, + "loss": 1.8486, + "step": 713 + }, + { + "epoch": 0.5340314136125655, + "grad_norm": 2.302210972237306, + "learning_rate": 8.764543765299245e-06, + "loss": 1.7164, + "step": 714 + }, + { + "epoch": 0.5347793567688855, + "grad_norm": 2.0910120488778223, + "learning_rate": 8.760472490038974e-06, + "loss": 1.7055, + "step": 715 + }, + { + "epoch": 0.5355272999252056, + "grad_norm": 2.076350170502579, + "learning_rate": 8.756395466566718e-06, + "loss": 1.5461, + "step": 716 + }, + { + "epoch": 0.5362752430815259, + "grad_norm": 2.797413772925583, + "learning_rate": 8.752312701114565e-06, + "loss": 1.4737, + "step": 717 + }, + { + "epoch": 0.537023186237846, + "grad_norm": 3.48796446492829, + "learning_rate": 8.748224199923378e-06, + "loss": 1.7769, + "step": 718 + }, + { + "epoch": 0.537771129394166, + "grad_norm": 1.857753054991564, + "learning_rate": 8.744129969242792e-06, + "loss": 1.7409, + "step": 719 + }, + { + "epoch": 0.5385190725504861, + "grad_norm": 1.919989507027575, + "learning_rate": 8.740030015331198e-06, + "loss": 1.5692, + "step": 720 + }, + { + "epoch": 0.5392670157068062, + "grad_norm": 2.0776965811085564, + "learning_rate": 8.735924344455732e-06, + "loss": 1.5002, + "step": 721 + }, + { + "epoch": 0.5400149588631264, + "grad_norm": 1.8047951315122759, + "learning_rate": 8.731812962892277e-06, + "loss": 1.6772, + "step": 722 + }, + { + "epoch": 0.5407629020194465, + "grad_norm": 1.750906919293806, + "learning_rate": 8.727695876925434e-06, + "loss": 1.5749, + "step": 723 + }, + { + "epoch": 0.5415108451757666, + "grad_norm": 2.1087327906672826, + "learning_rate": 8.723573092848535e-06, + "loss": 1.5238, + "step": 724 + }, + { + "epoch": 0.5422587883320867, + "grad_norm": 2.0095694609913286, + "learning_rate": 8.719444616963613e-06, + "loss": 1.7796, + "step": 725 + }, + { + "epoch": 0.5430067314884068, + "grad_norm": 1.949871826387318, + "learning_rate": 8.715310455581409e-06, + "loss": 1.8845, + "step": 726 + }, + { + "epoch": 0.543754674644727, + "grad_norm": 1.8968622210241604, + "learning_rate": 8.71117061502135e-06, + "loss": 1.9688, + "step": 727 + }, + { + "epoch": 0.5445026178010471, + "grad_norm": 2.1696691716151975, + "learning_rate": 8.707025101611546e-06, + "loss": 1.6048, + "step": 728 + }, + { + "epoch": 0.5452505609573672, + "grad_norm": 1.8990609858772642, + "learning_rate": 8.702873921688777e-06, + "loss": 1.7793, + "step": 729 + }, + { + "epoch": 0.5459985041136873, + "grad_norm": 1.7866132075971457, + "learning_rate": 8.698717081598484e-06, + "loss": 1.2253, + "step": 730 + }, + { + "epoch": 0.5467464472700074, + "grad_norm": 3.3074045242611674, + "learning_rate": 8.694554587694763e-06, + "loss": 1.6395, + "step": 731 + }, + { + "epoch": 0.5474943904263276, + "grad_norm": 1.6910740829454018, + "learning_rate": 8.690386446340353e-06, + "loss": 1.509, + "step": 732 + }, + { + "epoch": 0.5482423335826477, + "grad_norm": 2.116515489773148, + "learning_rate": 8.68621266390662e-06, + "loss": 1.7842, + "step": 733 + }, + { + "epoch": 0.5489902767389678, + "grad_norm": 2.1827819983643213, + "learning_rate": 8.68203324677356e-06, + "loss": 1.4332, + "step": 734 + }, + { + "epoch": 0.5497382198952879, + "grad_norm": 2.217083535739028, + "learning_rate": 8.677848201329775e-06, + "loss": 1.6965, + "step": 735 + }, + { + "epoch": 0.550486163051608, + "grad_norm": 2.0571813354082056, + "learning_rate": 8.673657533972477e-06, + "loss": 1.8138, + "step": 736 + }, + { + "epoch": 0.5512341062079282, + "grad_norm": 2.096509475960953, + "learning_rate": 8.669461251107466e-06, + "loss": 1.892, + "step": 737 + }, + { + "epoch": 0.5519820493642483, + "grad_norm": 2.4383753481507897, + "learning_rate": 8.665259359149132e-06, + "loss": 1.531, + "step": 738 + }, + { + "epoch": 0.5527299925205684, + "grad_norm": 1.937188374468386, + "learning_rate": 8.661051864520434e-06, + "loss": 1.536, + "step": 739 + }, + { + "epoch": 0.5534779356768885, + "grad_norm": 2.028672590782611, + "learning_rate": 8.656838773652891e-06, + "loss": 1.7953, + "step": 740 + }, + { + "epoch": 0.5542258788332087, + "grad_norm": 2.0625589895344, + "learning_rate": 8.65262009298659e-06, + "loss": 1.6337, + "step": 741 + }, + { + "epoch": 0.5549738219895288, + "grad_norm": 1.866528743625653, + "learning_rate": 8.64839582897015e-06, + "loss": 1.5362, + "step": 742 + }, + { + "epoch": 0.5557217651458489, + "grad_norm": 2.0182068534898385, + "learning_rate": 8.644165988060731e-06, + "loss": 1.5715, + "step": 743 + }, + { + "epoch": 0.556469708302169, + "grad_norm": 2.2526858090991375, + "learning_rate": 8.639930576724014e-06, + "loss": 1.3129, + "step": 744 + }, + { + "epoch": 0.5572176514584891, + "grad_norm": 1.8690826363321942, + "learning_rate": 8.635689601434197e-06, + "loss": 1.6623, + "step": 745 + }, + { + "epoch": 0.5579655946148093, + "grad_norm": 2.0811658737904244, + "learning_rate": 8.631443068673983e-06, + "loss": 1.4744, + "step": 746 + }, + { + "epoch": 0.5587135377711294, + "grad_norm": 1.8456349679061261, + "learning_rate": 8.627190984934568e-06, + "loss": 1.8498, + "step": 747 + }, + { + "epoch": 0.5594614809274495, + "grad_norm": 1.8371479338524248, + "learning_rate": 8.622933356715639e-06, + "loss": 1.4627, + "step": 748 + }, + { + "epoch": 0.5602094240837696, + "grad_norm": 1.8581533700526902, + "learning_rate": 8.61867019052535e-06, + "loss": 1.6945, + "step": 749 + }, + { + "epoch": 0.5609573672400897, + "grad_norm": 2.128289929396799, + "learning_rate": 8.614401492880328e-06, + "loss": 1.5856, + "step": 750 + }, + { + "epoch": 0.5617053103964099, + "grad_norm": 2.0476829899947315, + "learning_rate": 8.61012727030565e-06, + "loss": 1.6722, + "step": 751 + }, + { + "epoch": 0.56245325355273, + "grad_norm": 2.0844074458395836, + "learning_rate": 8.60584752933484e-06, + "loss": 1.6891, + "step": 752 + }, + { + "epoch": 0.5632011967090501, + "grad_norm": 1.8617386323646932, + "learning_rate": 8.601562276509855e-06, + "loss": 1.528, + "step": 753 + }, + { + "epoch": 0.5639491398653702, + "grad_norm": 1.9297028922209276, + "learning_rate": 8.597271518381083e-06, + "loss": 1.5188, + "step": 754 + }, + { + "epoch": 0.5646970830216903, + "grad_norm": 2.081786398841716, + "learning_rate": 8.592975261507323e-06, + "loss": 1.4146, + "step": 755 + }, + { + "epoch": 0.5654450261780105, + "grad_norm": 2.379402561220194, + "learning_rate": 8.588673512455781e-06, + "loss": 1.5025, + "step": 756 + }, + { + "epoch": 0.5661929693343306, + "grad_norm": 1.8498609738058995, + "learning_rate": 8.584366277802057e-06, + "loss": 1.5846, + "step": 757 + }, + { + "epoch": 0.5669409124906507, + "grad_norm": 2.218365611317838, + "learning_rate": 8.580053564130137e-06, + "loss": 1.5033, + "step": 758 + }, + { + "epoch": 0.5676888556469708, + "grad_norm": 1.8880823366945874, + "learning_rate": 8.575735378032382e-06, + "loss": 1.5963, + "step": 759 + }, + { + "epoch": 0.5684367988032909, + "grad_norm": 1.8537611946838513, + "learning_rate": 8.571411726109518e-06, + "loss": 1.7122, + "step": 760 + }, + { + "epoch": 0.5691847419596111, + "grad_norm": 2.213897222234766, + "learning_rate": 8.567082614970628e-06, + "loss": 1.5277, + "step": 761 + }, + { + "epoch": 0.5699326851159312, + "grad_norm": 1.754831170606896, + "learning_rate": 8.562748051233135e-06, + "loss": 1.6232, + "step": 762 + }, + { + "epoch": 0.5706806282722513, + "grad_norm": 1.8448368280845526, + "learning_rate": 8.558408041522801e-06, + "loss": 1.648, + "step": 763 + }, + { + "epoch": 0.5714285714285714, + "grad_norm": 2.3479660889046827, + "learning_rate": 8.554062592473713e-06, + "loss": 1.2107, + "step": 764 + }, + { + "epoch": 0.5721765145848915, + "grad_norm": 1.864556478465663, + "learning_rate": 8.549711710728269e-06, + "loss": 1.6114, + "step": 765 + }, + { + "epoch": 0.5729244577412117, + "grad_norm": 1.7429849793466181, + "learning_rate": 8.545355402937174e-06, + "loss": 1.5617, + "step": 766 + }, + { + "epoch": 0.5736724008975318, + "grad_norm": 2.032845735687417, + "learning_rate": 8.540993675759427e-06, + "loss": 1.5048, + "step": 767 + }, + { + "epoch": 0.5744203440538519, + "grad_norm": 2.118903653431299, + "learning_rate": 8.536626535862309e-06, + "loss": 1.6579, + "step": 768 + }, + { + "epoch": 0.575168287210172, + "grad_norm": 2.1084492927106497, + "learning_rate": 8.532253989921377e-06, + "loss": 1.5867, + "step": 769 + }, + { + "epoch": 0.5759162303664922, + "grad_norm": 2.027305209484783, + "learning_rate": 8.527876044620453e-06, + "loss": 1.4629, + "step": 770 + }, + { + "epoch": 0.5766641735228123, + "grad_norm": 2.2193945748533723, + "learning_rate": 8.523492706651607e-06, + "loss": 1.6628, + "step": 771 + }, + { + "epoch": 0.5774121166791324, + "grad_norm": 2.0991260149123097, + "learning_rate": 8.519103982715158e-06, + "loss": 1.6191, + "step": 772 + }, + { + "epoch": 0.5781600598354525, + "grad_norm": 2.2917776548971687, + "learning_rate": 8.514709879519654e-06, + "loss": 1.6421, + "step": 773 + }, + { + "epoch": 0.5789080029917726, + "grad_norm": 1.719088437243242, + "learning_rate": 8.510310403781868e-06, + "loss": 1.7219, + "step": 774 + }, + { + "epoch": 0.5796559461480928, + "grad_norm": 2.021272120588075, + "learning_rate": 8.505905562226784e-06, + "loss": 1.682, + "step": 775 + }, + { + "epoch": 0.5804038893044129, + "grad_norm": 1.9796917823825564, + "learning_rate": 8.501495361587589e-06, + "loss": 1.6345, + "step": 776 + }, + { + "epoch": 0.581151832460733, + "grad_norm": 1.8964921440573947, + "learning_rate": 8.497079808605659e-06, + "loss": 1.5452, + "step": 777 + }, + { + "epoch": 0.5818997756170531, + "grad_norm": 1.7688868474814945, + "learning_rate": 8.492658910030558e-06, + "loss": 1.6776, + "step": 778 + }, + { + "epoch": 0.5826477187733732, + "grad_norm": 2.2353902199367655, + "learning_rate": 8.488232672620016e-06, + "loss": 1.5749, + "step": 779 + }, + { + "epoch": 0.5833956619296934, + "grad_norm": 1.788325086014173, + "learning_rate": 8.483801103139923e-06, + "loss": 1.5166, + "step": 780 + }, + { + "epoch": 0.5841436050860135, + "grad_norm": 2.22813801110124, + "learning_rate": 8.479364208364322e-06, + "loss": 1.475, + "step": 781 + }, + { + "epoch": 0.5848915482423336, + "grad_norm": 1.9316469831324286, + "learning_rate": 8.474921995075399e-06, + "loss": 1.7447, + "step": 782 + }, + { + "epoch": 0.5856394913986537, + "grad_norm": 1.990661937296831, + "learning_rate": 8.470474470063462e-06, + "loss": 1.4568, + "step": 783 + }, + { + "epoch": 0.5863874345549738, + "grad_norm": 2.0168652928170983, + "learning_rate": 8.466021640126946e-06, + "loss": 1.7524, + "step": 784 + }, + { + "epoch": 0.587135377711294, + "grad_norm": 1.8427126922608594, + "learning_rate": 8.461563512072392e-06, + "loss": 1.4956, + "step": 785 + }, + { + "epoch": 0.5878833208676141, + "grad_norm": 3.802048212125386, + "learning_rate": 8.457100092714438e-06, + "loss": 1.5505, + "step": 786 + }, + { + "epoch": 0.5886312640239342, + "grad_norm": 1.7850049778096393, + "learning_rate": 8.452631388875814e-06, + "loss": 1.4924, + "step": 787 + }, + { + "epoch": 0.5893792071802543, + "grad_norm": 2.7968390166073367, + "learning_rate": 8.448157407387324e-06, + "loss": 1.6283, + "step": 788 + }, + { + "epoch": 0.5901271503365744, + "grad_norm": 2.166394359350806, + "learning_rate": 8.443678155087841e-06, + "loss": 1.566, + "step": 789 + }, + { + "epoch": 0.5908750934928946, + "grad_norm": 2.1047178373679367, + "learning_rate": 8.439193638824296e-06, + "loss": 1.6072, + "step": 790 + }, + { + "epoch": 0.5916230366492147, + "grad_norm": 2.1360640065004524, + "learning_rate": 8.434703865451666e-06, + "loss": 1.8176, + "step": 791 + }, + { + "epoch": 0.5923709798055348, + "grad_norm": 1.9780121040732936, + "learning_rate": 8.43020884183296e-06, + "loss": 1.9969, + "step": 792 + }, + { + "epoch": 0.5931189229618549, + "grad_norm": 2.935137942540914, + "learning_rate": 8.425708574839221e-06, + "loss": 1.7749, + "step": 793 + }, + { + "epoch": 0.5938668661181751, + "grad_norm": 1.9055905984594197, + "learning_rate": 8.421203071349498e-06, + "loss": 1.6888, + "step": 794 + }, + { + "epoch": 0.5946148092744952, + "grad_norm": 2.4204066840383467, + "learning_rate": 8.416692338250849e-06, + "loss": 1.927, + "step": 795 + }, + { + "epoch": 0.5953627524308153, + "grad_norm": 1.749287349959457, + "learning_rate": 8.412176382438327e-06, + "loss": 1.3717, + "step": 796 + }, + { + "epoch": 0.5961106955871354, + "grad_norm": 1.6781852157722852, + "learning_rate": 8.407655210814962e-06, + "loss": 1.3979, + "step": 797 + }, + { + "epoch": 0.5968586387434555, + "grad_norm": 2.278344673531173, + "learning_rate": 8.403128830291767e-06, + "loss": 1.815, + "step": 798 + }, + { + "epoch": 0.5976065818997757, + "grad_norm": 2.9612651359198976, + "learning_rate": 8.398597247787706e-06, + "loss": 1.7198, + "step": 799 + }, + { + "epoch": 0.5983545250560958, + "grad_norm": 1.8024715351319223, + "learning_rate": 8.394060470229704e-06, + "loss": 1.3771, + "step": 800 + }, + { + "epoch": 0.5991024682124159, + "grad_norm": 2.981757079462412, + "learning_rate": 8.389518504552623e-06, + "loss": 1.7364, + "step": 801 + }, + { + "epoch": 0.599850411368736, + "grad_norm": 3.4861250447075185, + "learning_rate": 8.384971357699255e-06, + "loss": 1.4686, + "step": 802 + }, + { + "epoch": 0.600598354525056, + "grad_norm": 2.2135317530286662, + "learning_rate": 8.380419036620311e-06, + "loss": 1.4984, + "step": 803 + }, + { + "epoch": 0.6013462976813763, + "grad_norm": 1.7771958190266843, + "learning_rate": 8.375861548274417e-06, + "loss": 1.5627, + "step": 804 + }, + { + "epoch": 0.6020942408376964, + "grad_norm": 1.902061348966851, + "learning_rate": 8.371298899628091e-06, + "loss": 1.6618, + "step": 805 + }, + { + "epoch": 0.6028421839940165, + "grad_norm": 1.8115560065459073, + "learning_rate": 8.366731097655742e-06, + "loss": 1.7509, + "step": 806 + }, + { + "epoch": 0.6035901271503366, + "grad_norm": 1.95625480841084, + "learning_rate": 8.362158149339657e-06, + "loss": 1.7303, + "step": 807 + }, + { + "epoch": 0.6043380703066566, + "grad_norm": 1.9176421969003437, + "learning_rate": 8.357580061669985e-06, + "loss": 1.4583, + "step": 808 + }, + { + "epoch": 0.6050860134629769, + "grad_norm": 1.970809904359506, + "learning_rate": 8.35299684164474e-06, + "loss": 1.8425, + "step": 809 + }, + { + "epoch": 0.605833956619297, + "grad_norm": 2.2497458988625705, + "learning_rate": 8.348408496269771e-06, + "loss": 1.5738, + "step": 810 + }, + { + "epoch": 0.606581899775617, + "grad_norm": 1.9137955874379167, + "learning_rate": 8.343815032558768e-06, + "loss": 1.2433, + "step": 811 + }, + { + "epoch": 0.6073298429319371, + "grad_norm": 2.3377859238283003, + "learning_rate": 8.339216457533244e-06, + "loss": 1.556, + "step": 812 + }, + { + "epoch": 0.6080777860882572, + "grad_norm": 2.3932689135511973, + "learning_rate": 8.334612778222524e-06, + "loss": 1.6939, + "step": 813 + }, + { + "epoch": 0.6088257292445775, + "grad_norm": 1.8675277443821252, + "learning_rate": 8.330004001663734e-06, + "loss": 1.5524, + "step": 814 + }, + { + "epoch": 0.6095736724008975, + "grad_norm": 1.8781305986604364, + "learning_rate": 8.325390134901794e-06, + "loss": 1.6869, + "step": 815 + }, + { + "epoch": 0.6103216155572176, + "grad_norm": 1.732789430754373, + "learning_rate": 8.320771184989404e-06, + "loss": 1.8359, + "step": 816 + }, + { + "epoch": 0.6110695587135377, + "grad_norm": 1.6868273720289357, + "learning_rate": 8.316147158987036e-06, + "loss": 1.567, + "step": 817 + }, + { + "epoch": 0.6118175018698578, + "grad_norm": 1.9862230533036425, + "learning_rate": 8.311518063962915e-06, + "loss": 1.5295, + "step": 818 + }, + { + "epoch": 0.612565445026178, + "grad_norm": 1.7524061096215595, + "learning_rate": 8.306883906993022e-06, + "loss": 1.6754, + "step": 819 + }, + { + "epoch": 0.6133133881824981, + "grad_norm": 2.1525431164140962, + "learning_rate": 8.30224469516107e-06, + "loss": 1.8232, + "step": 820 + }, + { + "epoch": 0.6140613313388182, + "grad_norm": 2.117652599114552, + "learning_rate": 8.297600435558506e-06, + "loss": 1.6133, + "step": 821 + }, + { + "epoch": 0.6148092744951383, + "grad_norm": 1.8139042877837401, + "learning_rate": 8.292951135284481e-06, + "loss": 1.7214, + "step": 822 + }, + { + "epoch": 0.6155572176514585, + "grad_norm": 2.07899992620621, + "learning_rate": 8.288296801445863e-06, + "loss": 1.7301, + "step": 823 + }, + { + "epoch": 0.6163051608077786, + "grad_norm": 1.7406715622129731, + "learning_rate": 8.28363744115721e-06, + "loss": 1.7856, + "step": 824 + }, + { + "epoch": 0.6170531039640987, + "grad_norm": 1.7002479825287526, + "learning_rate": 8.278973061540758e-06, + "loss": 1.5853, + "step": 825 + }, + { + "epoch": 0.6178010471204188, + "grad_norm": 2.9855403063465245, + "learning_rate": 8.274303669726427e-06, + "loss": 1.421, + "step": 826 + }, + { + "epoch": 0.6185489902767389, + "grad_norm": 2.093994510230207, + "learning_rate": 8.269629272851786e-06, + "loss": 1.5806, + "step": 827 + }, + { + "epoch": 0.6192969334330591, + "grad_norm": 3.0457064296027743, + "learning_rate": 8.264949878062063e-06, + "loss": 1.7388, + "step": 828 + }, + { + "epoch": 0.6200448765893792, + "grad_norm": 1.8831746899686421, + "learning_rate": 8.260265492510121e-06, + "loss": 1.637, + "step": 829 + }, + { + "epoch": 0.6207928197456993, + "grad_norm": 1.9150436373366575, + "learning_rate": 8.25557612335646e-06, + "loss": 1.5209, + "step": 830 + }, + { + "epoch": 0.6215407629020194, + "grad_norm": 2.0561929793173537, + "learning_rate": 8.250881777769187e-06, + "loss": 1.3389, + "step": 831 + }, + { + "epoch": 0.6222887060583395, + "grad_norm": 2.0714056383345376, + "learning_rate": 8.246182462924022e-06, + "loss": 1.6778, + "step": 832 + }, + { + "epoch": 0.6230366492146597, + "grad_norm": 1.7317663997672965, + "learning_rate": 8.24147818600428e-06, + "loss": 1.8065, + "step": 833 + }, + { + "epoch": 0.6237845923709798, + "grad_norm": 2.0019085986285607, + "learning_rate": 8.236768954200861e-06, + "loss": 1.7962, + "step": 834 + }, + { + "epoch": 0.6245325355272999, + "grad_norm": 2.135780207725876, + "learning_rate": 8.232054774712239e-06, + "loss": 1.7492, + "step": 835 + }, + { + "epoch": 0.62528047868362, + "grad_norm": 2.684401716705146, + "learning_rate": 8.227335654744451e-06, + "loss": 1.7266, + "step": 836 + }, + { + "epoch": 0.6260284218399401, + "grad_norm": 2.0671787750200066, + "learning_rate": 8.222611601511084e-06, + "loss": 1.7625, + "step": 837 + }, + { + "epoch": 0.6267763649962603, + "grad_norm": 1.782776663677292, + "learning_rate": 8.217882622233268e-06, + "loss": 1.5065, + "step": 838 + }, + { + "epoch": 0.6275243081525804, + "grad_norm": 2.198046140637333, + "learning_rate": 8.213148724139665e-06, + "loss": 1.8854, + "step": 839 + }, + { + "epoch": 0.6282722513089005, + "grad_norm": 2.5474047544723533, + "learning_rate": 8.20840991446645e-06, + "loss": 1.7207, + "step": 840 + }, + { + "epoch": 0.6290201944652206, + "grad_norm": 1.908227797356393, + "learning_rate": 8.203666200457312e-06, + "loss": 1.4849, + "step": 841 + }, + { + "epoch": 0.6297681376215407, + "grad_norm": 1.8611577958726355, + "learning_rate": 8.198917589363433e-06, + "loss": 1.7685, + "step": 842 + }, + { + "epoch": 0.6305160807778609, + "grad_norm": 2.87111732234712, + "learning_rate": 8.19416408844348e-06, + "loss": 1.4293, + "step": 843 + }, + { + "epoch": 0.631264023934181, + "grad_norm": 2.4313732989418018, + "learning_rate": 8.1894057049636e-06, + "loss": 1.7262, + "step": 844 + }, + { + "epoch": 0.6320119670905011, + "grad_norm": 1.921723461819831, + "learning_rate": 8.184642446197396e-06, + "loss": 1.7784, + "step": 845 + }, + { + "epoch": 0.6327599102468212, + "grad_norm": 1.7887431998890861, + "learning_rate": 8.179874319425926e-06, + "loss": 1.7871, + "step": 846 + }, + { + "epoch": 0.6335078534031413, + "grad_norm": 2.2329218573473733, + "learning_rate": 8.175101331937692e-06, + "loss": 1.732, + "step": 847 + }, + { + "epoch": 0.6342557965594615, + "grad_norm": 2.8226857665894975, + "learning_rate": 8.170323491028625e-06, + "loss": 1.6684, + "step": 848 + }, + { + "epoch": 0.6350037397157816, + "grad_norm": 2.0190939761947897, + "learning_rate": 8.16554080400207e-06, + "loss": 1.5551, + "step": 849 + }, + { + "epoch": 0.6357516828721017, + "grad_norm": 1.8351324090760177, + "learning_rate": 8.160753278168787e-06, + "loss": 1.7339, + "step": 850 + }, + { + "epoch": 0.6364996260284218, + "grad_norm": 1.6846177556602107, + "learning_rate": 8.155960920846925e-06, + "loss": 1.8692, + "step": 851 + }, + { + "epoch": 0.637247569184742, + "grad_norm": 1.7553774370602073, + "learning_rate": 8.151163739362029e-06, + "loss": 1.794, + "step": 852 + }, + { + "epoch": 0.6379955123410621, + "grad_norm": 1.9357128643207209, + "learning_rate": 8.146361741047007e-06, + "loss": 1.6285, + "step": 853 + }, + { + "epoch": 0.6387434554973822, + "grad_norm": 2.173870685646154, + "learning_rate": 8.141554933242135e-06, + "loss": 1.6914, + "step": 854 + }, + { + "epoch": 0.6394913986537023, + "grad_norm": 1.9215307841392584, + "learning_rate": 8.136743323295041e-06, + "loss": 1.7594, + "step": 855 + }, + { + "epoch": 0.6402393418100224, + "grad_norm": 1.9748190673974586, + "learning_rate": 8.131926918560692e-06, + "loss": 1.6582, + "step": 856 + }, + { + "epoch": 0.6409872849663426, + "grad_norm": 1.8608914285331364, + "learning_rate": 8.127105726401387e-06, + "loss": 1.7869, + "step": 857 + }, + { + "epoch": 0.6417352281226627, + "grad_norm": 2.024502874310778, + "learning_rate": 8.122279754186738e-06, + "loss": 1.7814, + "step": 858 + }, + { + "epoch": 0.6424831712789828, + "grad_norm": 2.0496855488521306, + "learning_rate": 8.117449009293668e-06, + "loss": 1.6463, + "step": 859 + }, + { + "epoch": 0.6432311144353029, + "grad_norm": 1.7343455164609947, + "learning_rate": 8.112613499106397e-06, + "loss": 1.702, + "step": 860 + }, + { + "epoch": 0.643979057591623, + "grad_norm": 1.8648196333794869, + "learning_rate": 8.10777323101642e-06, + "loss": 1.5164, + "step": 861 + }, + { + "epoch": 0.6447270007479432, + "grad_norm": 1.8116914649424336, + "learning_rate": 8.102928212422519e-06, + "loss": 1.8677, + "step": 862 + }, + { + "epoch": 0.6454749439042633, + "grad_norm": 1.745150868308597, + "learning_rate": 8.098078450730723e-06, + "loss": 1.6904, + "step": 863 + }, + { + "epoch": 0.6462228870605834, + "grad_norm": 1.991495777034257, + "learning_rate": 8.093223953354324e-06, + "loss": 1.6376, + "step": 864 + }, + { + "epoch": 0.6469708302169035, + "grad_norm": 1.7381975140725703, + "learning_rate": 8.08836472771384e-06, + "loss": 1.7691, + "step": 865 + }, + { + "epoch": 0.6477187733732236, + "grad_norm": 1.9358178592211837, + "learning_rate": 8.08350078123703e-06, + "loss": 1.639, + "step": 866 + }, + { + "epoch": 0.6484667165295438, + "grad_norm": 2.234519226622757, + "learning_rate": 8.07863212135886e-06, + "loss": 1.6954, + "step": 867 + }, + { + "epoch": 0.6492146596858639, + "grad_norm": 2.001213181799438, + "learning_rate": 8.073758755521506e-06, + "loss": 1.6985, + "step": 868 + }, + { + "epoch": 0.649962602842184, + "grad_norm": 2.0132756401723437, + "learning_rate": 8.06888069117433e-06, + "loss": 1.7579, + "step": 869 + }, + { + "epoch": 0.6507105459985041, + "grad_norm": 1.876669776126021, + "learning_rate": 8.063997935773885e-06, + "loss": 1.7158, + "step": 870 + }, + { + "epoch": 0.6514584891548242, + "grad_norm": 1.8208361043163923, + "learning_rate": 8.059110496783891e-06, + "loss": 1.4098, + "step": 871 + }, + { + "epoch": 0.6522064323111444, + "grad_norm": 1.7065727177126822, + "learning_rate": 8.054218381675225e-06, + "loss": 1.6405, + "step": 872 + }, + { + "epoch": 0.6529543754674645, + "grad_norm": 2.1191089059682584, + "learning_rate": 8.049321597925914e-06, + "loss": 1.5056, + "step": 873 + }, + { + "epoch": 0.6537023186237846, + "grad_norm": 1.8598551312893787, + "learning_rate": 8.044420153021123e-06, + "loss": 1.6684, + "step": 874 + }, + { + "epoch": 0.6544502617801047, + "grad_norm": 2.3708135172219964, + "learning_rate": 8.03951405445314e-06, + "loss": 1.4392, + "step": 875 + }, + { + "epoch": 0.6551982049364248, + "grad_norm": 3.2122721653799613, + "learning_rate": 8.034603309721369e-06, + "loss": 1.853, + "step": 876 + }, + { + "epoch": 0.655946148092745, + "grad_norm": 2.7228179659944813, + "learning_rate": 8.02968792633231e-06, + "loss": 1.4994, + "step": 877 + }, + { + "epoch": 0.6566940912490651, + "grad_norm": 2.3433467185554, + "learning_rate": 8.024767911799558e-06, + "loss": 1.7986, + "step": 878 + }, + { + "epoch": 0.6574420344053852, + "grad_norm": 2.23831486401973, + "learning_rate": 8.019843273643789e-06, + "loss": 1.8238, + "step": 879 + }, + { + "epoch": 0.6581899775617053, + "grad_norm": 2.0522163996528184, + "learning_rate": 8.014914019392743e-06, + "loss": 1.6491, + "step": 880 + }, + { + "epoch": 0.6589379207180255, + "grad_norm": 3.018232636933135, + "learning_rate": 8.009980156581218e-06, + "loss": 1.8223, + "step": 881 + }, + { + "epoch": 0.6596858638743456, + "grad_norm": 2.116285416444218, + "learning_rate": 8.005041692751055e-06, + "loss": 1.8775, + "step": 882 + }, + { + "epoch": 0.6604338070306657, + "grad_norm": 2.2653218591113258, + "learning_rate": 8.00009863545113e-06, + "loss": 1.8445, + "step": 883 + }, + { + "epoch": 0.6611817501869858, + "grad_norm": 1.824709034327195, + "learning_rate": 7.99515099223734e-06, + "loss": 1.7362, + "step": 884 + }, + { + "epoch": 0.6619296933433059, + "grad_norm": 2.179583400352679, + "learning_rate": 7.99019877067259e-06, + "loss": 1.5692, + "step": 885 + }, + { + "epoch": 0.6626776364996261, + "grad_norm": 1.7787111949141756, + "learning_rate": 7.985241978326786e-06, + "loss": 1.7328, + "step": 886 + }, + { + "epoch": 0.6634255796559462, + "grad_norm": 1.7997585749390628, + "learning_rate": 7.98028062277682e-06, + "loss": 1.6338, + "step": 887 + }, + { + "epoch": 0.6641735228122663, + "grad_norm": 1.7497597173311592, + "learning_rate": 7.975314711606558e-06, + "loss": 1.7798, + "step": 888 + }, + { + "epoch": 0.6649214659685864, + "grad_norm": 1.8004398278753801, + "learning_rate": 7.970344252406832e-06, + "loss": 1.5986, + "step": 889 + }, + { + "epoch": 0.6656694091249065, + "grad_norm": 2.1370831897226967, + "learning_rate": 7.965369252775423e-06, + "loss": 1.3088, + "step": 890 + }, + { + "epoch": 0.6664173522812267, + "grad_norm": 1.8219770450212083, + "learning_rate": 7.960389720317056e-06, + "loss": 1.4647, + "step": 891 + }, + { + "epoch": 0.6671652954375468, + "grad_norm": 1.9819382904621397, + "learning_rate": 7.955405662643384e-06, + "loss": 1.6167, + "step": 892 + }, + { + "epoch": 0.6679132385938669, + "grad_norm": 2.0155436010902887, + "learning_rate": 7.950417087372972e-06, + "loss": 1.8539, + "step": 893 + }, + { + "epoch": 0.668661181750187, + "grad_norm": 1.9262740755859964, + "learning_rate": 7.945424002131298e-06, + "loss": 1.4404, + "step": 894 + }, + { + "epoch": 0.6694091249065071, + "grad_norm": 1.8168208048397663, + "learning_rate": 7.940426414550732e-06, + "loss": 1.557, + "step": 895 + }, + { + "epoch": 0.6701570680628273, + "grad_norm": 3.589479726338036, + "learning_rate": 7.935424332270523e-06, + "loss": 1.6879, + "step": 896 + }, + { + "epoch": 0.6709050112191474, + "grad_norm": 1.682003409298543, + "learning_rate": 7.93041776293679e-06, + "loss": 1.5847, + "step": 897 + }, + { + "epoch": 0.6716529543754675, + "grad_norm": 1.7844794027435198, + "learning_rate": 7.925406714202518e-06, + "loss": 1.5267, + "step": 898 + }, + { + "epoch": 0.6724008975317876, + "grad_norm": 1.7555831489735445, + "learning_rate": 7.920391193727533e-06, + "loss": 1.575, + "step": 899 + }, + { + "epoch": 0.6731488406881077, + "grad_norm": 2.3863887783399944, + "learning_rate": 7.915371209178496e-06, + "loss": 1.728, + "step": 900 + }, + { + "epoch": 0.6738967838444279, + "grad_norm": 2.0384204412412914, + "learning_rate": 7.910346768228893e-06, + "loss": 1.7193, + "step": 901 + }, + { + "epoch": 0.674644727000748, + "grad_norm": 2.094324021876296, + "learning_rate": 7.905317878559026e-06, + "loss": 1.6664, + "step": 902 + }, + { + "epoch": 0.675392670157068, + "grad_norm": 3.40618846266283, + "learning_rate": 7.900284547855992e-06, + "loss": 1.5122, + "step": 903 + }, + { + "epoch": 0.6761406133133882, + "grad_norm": 1.8908821353214156, + "learning_rate": 7.895246783813679e-06, + "loss": 1.7673, + "step": 904 + }, + { + "epoch": 0.6768885564697082, + "grad_norm": 3.8834457220286236, + "learning_rate": 7.89020459413275e-06, + "loss": 1.6275, + "step": 905 + }, + { + "epoch": 0.6776364996260285, + "grad_norm": 2.5360104193537145, + "learning_rate": 7.885157986520635e-06, + "loss": 1.7025, + "step": 906 + }, + { + "epoch": 0.6783844427823486, + "grad_norm": 2.2015517792665973, + "learning_rate": 7.880106968691516e-06, + "loss": 1.6085, + "step": 907 + }, + { + "epoch": 0.6791323859386686, + "grad_norm": 2.2479091361286554, + "learning_rate": 7.875051548366317e-06, + "loss": 1.8266, + "step": 908 + }, + { + "epoch": 0.6798803290949887, + "grad_norm": 1.8912592423237955, + "learning_rate": 7.869991733272693e-06, + "loss": 1.6646, + "step": 909 + }, + { + "epoch": 0.680628272251309, + "grad_norm": 2.389872044451628, + "learning_rate": 7.864927531145012e-06, + "loss": 1.3021, + "step": 910 + }, + { + "epoch": 0.681376215407629, + "grad_norm": 1.8472023263802353, + "learning_rate": 7.859858949724351e-06, + "loss": 1.5347, + "step": 911 + }, + { + "epoch": 0.6821241585639491, + "grad_norm": 1.8494590396675423, + "learning_rate": 7.854785996758484e-06, + "loss": 1.6259, + "step": 912 + }, + { + "epoch": 0.6828721017202692, + "grad_norm": 2.0320429202878545, + "learning_rate": 7.849708680001862e-06, + "loss": 1.6132, + "step": 913 + }, + { + "epoch": 0.6836200448765893, + "grad_norm": 2.237629990431062, + "learning_rate": 7.844627007215613e-06, + "loss": 1.8721, + "step": 914 + }, + { + "epoch": 0.6843679880329095, + "grad_norm": 2.2928383046616942, + "learning_rate": 7.839540986167514e-06, + "loss": 1.4504, + "step": 915 + }, + { + "epoch": 0.6851159311892296, + "grad_norm": 5.809065655160868, + "learning_rate": 7.834450624631998e-06, + "loss": 1.5623, + "step": 916 + }, + { + "epoch": 0.6858638743455497, + "grad_norm": 2.5872634869170916, + "learning_rate": 7.829355930390126e-06, + "loss": 1.7356, + "step": 917 + }, + { + "epoch": 0.6866118175018698, + "grad_norm": 1.8227052655923113, + "learning_rate": 7.824256911229588e-06, + "loss": 1.6377, + "step": 918 + }, + { + "epoch": 0.6873597606581899, + "grad_norm": 1.9812676384691086, + "learning_rate": 7.81915357494468e-06, + "loss": 1.6195, + "step": 919 + }, + { + "epoch": 0.6881077038145101, + "grad_norm": 3.2907980942328026, + "learning_rate": 7.814045929336301e-06, + "loss": 1.5968, + "step": 920 + }, + { + "epoch": 0.6888556469708302, + "grad_norm": 2.3433951472512153, + "learning_rate": 7.808933982211932e-06, + "loss": 1.2837, + "step": 921 + }, + { + "epoch": 0.6896035901271503, + "grad_norm": 3.304139998790838, + "learning_rate": 7.803817741385636e-06, + "loss": 1.6614, + "step": 922 + }, + { + "epoch": 0.6903515332834704, + "grad_norm": 1.7989164423521893, + "learning_rate": 7.798697214678033e-06, + "loss": 1.7423, + "step": 923 + }, + { + "epoch": 0.6910994764397905, + "grad_norm": 2.0296243485871455, + "learning_rate": 7.7935724099163e-06, + "loss": 1.5719, + "step": 924 + }, + { + "epoch": 0.6918474195961107, + "grad_norm": 2.112465338022877, + "learning_rate": 7.788443334934148e-06, + "loss": 1.7717, + "step": 925 + }, + { + "epoch": 0.6925953627524308, + "grad_norm": 2.3625745994086036, + "learning_rate": 7.783309997571819e-06, + "loss": 1.7578, + "step": 926 + }, + { + "epoch": 0.6933433059087509, + "grad_norm": 2.289543958986776, + "learning_rate": 7.778172405676067e-06, + "loss": 1.303, + "step": 927 + }, + { + "epoch": 0.694091249065071, + "grad_norm": 1.9074544080548206, + "learning_rate": 7.773030567100158e-06, + "loss": 1.7562, + "step": 928 + }, + { + "epoch": 0.6948391922213911, + "grad_norm": 2.6260870015011593, + "learning_rate": 7.767884489703837e-06, + "loss": 1.7456, + "step": 929 + }, + { + "epoch": 0.6955871353777113, + "grad_norm": 1.7007624430841781, + "learning_rate": 7.762734181353336e-06, + "loss": 1.6894, + "step": 930 + }, + { + "epoch": 0.6963350785340314, + "grad_norm": 2.0931755116465136, + "learning_rate": 7.757579649921354e-06, + "loss": 1.4463, + "step": 931 + }, + { + "epoch": 0.6970830216903515, + "grad_norm": 2.776208971055814, + "learning_rate": 7.752420903287043e-06, + "loss": 1.485, + "step": 932 + }, + { + "epoch": 0.6978309648466716, + "grad_norm": 1.827555757931979, + "learning_rate": 7.747257949336003e-06, + "loss": 1.748, + "step": 933 + }, + { + "epoch": 0.6985789080029918, + "grad_norm": 1.833683262327015, + "learning_rate": 7.742090795960258e-06, + "loss": 1.7968, + "step": 934 + }, + { + "epoch": 0.6993268511593119, + "grad_norm": 1.8271968558903742, + "learning_rate": 7.736919451058257e-06, + "loss": 1.6387, + "step": 935 + }, + { + "epoch": 0.700074794315632, + "grad_norm": 1.8378323738524756, + "learning_rate": 7.731743922534854e-06, + "loss": 1.6497, + "step": 936 + }, + { + "epoch": 0.7008227374719521, + "grad_norm": 2.140241144558504, + "learning_rate": 7.726564218301298e-06, + "loss": 1.6454, + "step": 937 + }, + { + "epoch": 0.7015706806282722, + "grad_norm": 1.9617196840320235, + "learning_rate": 7.721380346275221e-06, + "loss": 1.5655, + "step": 938 + }, + { + "epoch": 0.7023186237845924, + "grad_norm": 2.3248058878594344, + "learning_rate": 7.716192314380626e-06, + "loss": 1.5793, + "step": 939 + }, + { + "epoch": 0.7030665669409125, + "grad_norm": 1.8776724917440581, + "learning_rate": 7.711000130547875e-06, + "loss": 1.6442, + "step": 940 + }, + { + "epoch": 0.7038145100972326, + "grad_norm": 2.236466042608474, + "learning_rate": 7.705803802713678e-06, + "loss": 1.8104, + "step": 941 + }, + { + "epoch": 0.7045624532535527, + "grad_norm": 1.9542484203649002, + "learning_rate": 7.700603338821074e-06, + "loss": 1.8458, + "step": 942 + }, + { + "epoch": 0.7053103964098728, + "grad_norm": 1.9626027374985076, + "learning_rate": 7.695398746819431e-06, + "loss": 1.6878, + "step": 943 + }, + { + "epoch": 0.706058339566193, + "grad_norm": 1.755425761041356, + "learning_rate": 7.690190034664423e-06, + "loss": 1.4422, + "step": 944 + }, + { + "epoch": 0.7068062827225131, + "grad_norm": 2.7153528445630837, + "learning_rate": 7.684977210318024e-06, + "loss": 1.8404, + "step": 945 + }, + { + "epoch": 0.7075542258788332, + "grad_norm": 2.503164887971006, + "learning_rate": 7.679760281748492e-06, + "loss": 1.9051, + "step": 946 + }, + { + "epoch": 0.7083021690351533, + "grad_norm": 1.8813548503209774, + "learning_rate": 7.674539256930364e-06, + "loss": 1.537, + "step": 947 + }, + { + "epoch": 0.7090501121914734, + "grad_norm": 1.7451341586102533, + "learning_rate": 7.669314143844429e-06, + "loss": 1.7578, + "step": 948 + }, + { + "epoch": 0.7097980553477936, + "grad_norm": 1.7183445696416246, + "learning_rate": 7.664084950477732e-06, + "loss": 1.8581, + "step": 949 + }, + { + "epoch": 0.7105459985041137, + "grad_norm": 1.9017576814875805, + "learning_rate": 7.658851684823554e-06, + "loss": 1.7428, + "step": 950 + }, + { + "epoch": 0.7112939416604338, + "grad_norm": 2.038474810301481, + "learning_rate": 7.653614354881403e-06, + "loss": 1.5386, + "step": 951 + }, + { + "epoch": 0.7120418848167539, + "grad_norm": 1.7558425037123067, + "learning_rate": 7.648372968656995e-06, + "loss": 1.6293, + "step": 952 + }, + { + "epoch": 0.712789827973074, + "grad_norm": 1.9021411809909445, + "learning_rate": 7.643127534162246e-06, + "loss": 1.5255, + "step": 953 + }, + { + "epoch": 0.7135377711293942, + "grad_norm": 1.658098066118051, + "learning_rate": 7.637878059415267e-06, + "loss": 1.4897, + "step": 954 + }, + { + "epoch": 0.7142857142857143, + "grad_norm": 2.849968230520914, + "learning_rate": 7.632624552440337e-06, + "loss": 1.5776, + "step": 955 + }, + { + "epoch": 0.7150336574420344, + "grad_norm": 2.095152875461566, + "learning_rate": 7.627367021267906e-06, + "loss": 1.4045, + "step": 956 + }, + { + "epoch": 0.7157816005983545, + "grad_norm": 3.816800260896842, + "learning_rate": 7.622105473934569e-06, + "loss": 1.4626, + "step": 957 + }, + { + "epoch": 0.7165295437546746, + "grad_norm": 1.7477375882849537, + "learning_rate": 7.616839918483061e-06, + "loss": 1.6932, + "step": 958 + }, + { + "epoch": 0.7172774869109948, + "grad_norm": 1.5875392896459393, + "learning_rate": 7.611570362962247e-06, + "loss": 1.5895, + "step": 959 + }, + { + "epoch": 0.7180254300673149, + "grad_norm": 1.8940821095223543, + "learning_rate": 7.606296815427106e-06, + "loss": 1.7314, + "step": 960 + }, + { + "epoch": 0.718773373223635, + "grad_norm": 1.9847613027698392, + "learning_rate": 7.601019283938718e-06, + "loss": 1.5857, + "step": 961 + }, + { + "epoch": 0.7195213163799551, + "grad_norm": 1.8851403027879032, + "learning_rate": 7.595737776564249e-06, + "loss": 1.8613, + "step": 962 + }, + { + "epoch": 0.7202692595362753, + "grad_norm": 1.9839092104400002, + "learning_rate": 7.590452301376949e-06, + "loss": 1.6777, + "step": 963 + }, + { + "epoch": 0.7210172026925954, + "grad_norm": 2.183760497032052, + "learning_rate": 7.58516286645613e-06, + "loss": 1.6553, + "step": 964 + }, + { + "epoch": 0.7217651458489155, + "grad_norm": 2.0283998837408173, + "learning_rate": 7.579869479887158e-06, + "loss": 1.5729, + "step": 965 + }, + { + "epoch": 0.7225130890052356, + "grad_norm": 1.7613642841889257, + "learning_rate": 7.574572149761437e-06, + "loss": 1.7472, + "step": 966 + }, + { + "epoch": 0.7232610321615557, + "grad_norm": 2.0002287160387286, + "learning_rate": 7.569270884176401e-06, + "loss": 1.5261, + "step": 967 + }, + { + "epoch": 0.7240089753178759, + "grad_norm": 2.561421304210536, + "learning_rate": 7.5639656912355e-06, + "loss": 1.7693, + "step": 968 + }, + { + "epoch": 0.724756918474196, + "grad_norm": 2.4065402328602743, + "learning_rate": 7.5586565790481855e-06, + "loss": 1.7596, + "step": 969 + }, + { + "epoch": 0.7255048616305161, + "grad_norm": 1.9339658649742397, + "learning_rate": 7.553343555729904e-06, + "loss": 1.6348, + "step": 970 + }, + { + "epoch": 0.7262528047868362, + "grad_norm": 2.1084854513195865, + "learning_rate": 7.548026629402074e-06, + "loss": 1.7807, + "step": 971 + }, + { + "epoch": 0.7270007479431563, + "grad_norm": 1.601821019726851, + "learning_rate": 7.542705808192085e-06, + "loss": 1.752, + "step": 972 + }, + { + "epoch": 0.7277486910994765, + "grad_norm": 1.8912981680450731, + "learning_rate": 7.5373811002332785e-06, + "loss": 1.5178, + "step": 973 + }, + { + "epoch": 0.7284966342557966, + "grad_norm": 1.6827297539280617, + "learning_rate": 7.532052513664939e-06, + "loss": 1.6527, + "step": 974 + }, + { + "epoch": 0.7292445774121167, + "grad_norm": 1.8997035722554934, + "learning_rate": 7.526720056632277e-06, + "loss": 1.6663, + "step": 975 + }, + { + "epoch": 0.7299925205684368, + "grad_norm": 1.9132337568608964, + "learning_rate": 7.521383737286423e-06, + "loss": 1.78, + "step": 976 + }, + { + "epoch": 0.7307404637247569, + "grad_norm": 1.7309450637698678, + "learning_rate": 7.516043563784405e-06, + "loss": 1.6231, + "step": 977 + }, + { + "epoch": 0.7314884068810771, + "grad_norm": 2.9261826536712388, + "learning_rate": 7.510699544289151e-06, + "loss": 1.6761, + "step": 978 + }, + { + "epoch": 0.7322363500373972, + "grad_norm": 2.0125976945776083, + "learning_rate": 7.505351686969458e-06, + "loss": 1.5989, + "step": 979 + }, + { + "epoch": 0.7329842931937173, + "grad_norm": 1.7365026071577048, + "learning_rate": 7.500000000000001e-06, + "loss": 1.8397, + "step": 980 + }, + { + "epoch": 0.7337322363500374, + "grad_norm": 1.8781317422618513, + "learning_rate": 7.494644491561299e-06, + "loss": 1.3492, + "step": 981 + }, + { + "epoch": 0.7344801795063575, + "grad_norm": 1.909547525746187, + "learning_rate": 7.4892851698397174e-06, + "loss": 1.5859, + "step": 982 + }, + { + "epoch": 0.7352281226626777, + "grad_norm": 2.141245939114944, + "learning_rate": 7.483922043027448e-06, + "loss": 1.8011, + "step": 983 + }, + { + "epoch": 0.7359760658189978, + "grad_norm": 1.757155031820437, + "learning_rate": 7.478555119322505e-06, + "loss": 1.7979, + "step": 984 + }, + { + "epoch": 0.7367240089753179, + "grad_norm": 2.0129264328336087, + "learning_rate": 7.4731844069286965e-06, + "loss": 1.7123, + "step": 985 + }, + { + "epoch": 0.737471952131638, + "grad_norm": 1.7873441049738588, + "learning_rate": 7.467809914055631e-06, + "loss": 1.6418, + "step": 986 + }, + { + "epoch": 0.7382198952879581, + "grad_norm": 1.9840008629707693, + "learning_rate": 7.462431648918689e-06, + "loss": 1.486, + "step": 987 + }, + { + "epoch": 0.7389678384442783, + "grad_norm": 1.7746697330817673, + "learning_rate": 7.457049619739023e-06, + "loss": 1.5967, + "step": 988 + }, + { + "epoch": 0.7397157816005984, + "grad_norm": 2.1197664542315975, + "learning_rate": 7.451663834743537e-06, + "loss": 1.5436, + "step": 989 + }, + { + "epoch": 0.7404637247569185, + "grad_norm": 2.288840215401231, + "learning_rate": 7.446274302164873e-06, + "loss": 1.614, + "step": 990 + }, + { + "epoch": 0.7412116679132386, + "grad_norm": 1.798959619310891, + "learning_rate": 7.440881030241407e-06, + "loss": 1.6238, + "step": 991 + }, + { + "epoch": 0.7419596110695588, + "grad_norm": 2.0401818312000444, + "learning_rate": 7.435484027217226e-06, + "loss": 1.5714, + "step": 992 + }, + { + "epoch": 0.7427075542258789, + "grad_norm": 2.176797285439103, + "learning_rate": 7.430083301342124e-06, + "loss": 1.8647, + "step": 993 + }, + { + "epoch": 0.743455497382199, + "grad_norm": 2.1473732831914534, + "learning_rate": 7.424678860871584e-06, + "loss": 1.6813, + "step": 994 + }, + { + "epoch": 0.7442034405385191, + "grad_norm": 2.0091918210991326, + "learning_rate": 7.419270714066765e-06, + "loss": 1.7861, + "step": 995 + }, + { + "epoch": 0.7449513836948392, + "grad_norm": 2.036022527717881, + "learning_rate": 7.413858869194497e-06, + "loss": 1.6724, + "step": 996 + }, + { + "epoch": 0.7456993268511594, + "grad_norm": 1.804133262925045, + "learning_rate": 7.408443334527257e-06, + "loss": 1.4007, + "step": 997 + }, + { + "epoch": 0.7464472700074795, + "grad_norm": 2.3763437143638733, + "learning_rate": 7.4030241183431675e-06, + "loss": 1.4952, + "step": 998 + }, + { + "epoch": 0.7471952131637996, + "grad_norm": 2.3559337505727522, + "learning_rate": 7.39760122892597e-06, + "loss": 1.6118, + "step": 999 + }, + { + "epoch": 0.7479431563201197, + "grad_norm": 2.55388488233513, + "learning_rate": 7.3921746745650315e-06, + "loss": 1.6621, + "step": 1000 + }, + { + "epoch": 0.7486910994764397, + "grad_norm": 2.2009464044940312, + "learning_rate": 7.3867444635553165e-06, + "loss": 1.7796, + "step": 1001 + }, + { + "epoch": 0.74943904263276, + "grad_norm": 2.5058675259805985, + "learning_rate": 7.381310604197375e-06, + "loss": 1.5201, + "step": 1002 + }, + { + "epoch": 0.75018698578908, + "grad_norm": 1.8693950202281306, + "learning_rate": 7.3758731047973416e-06, + "loss": 1.7906, + "step": 1003 + }, + { + "epoch": 0.7509349289454001, + "grad_norm": 1.8712180512471015, + "learning_rate": 7.370431973666909e-06, + "loss": 1.725, + "step": 1004 + }, + { + "epoch": 0.7516828721017202, + "grad_norm": 2.0409792928973314, + "learning_rate": 7.364987219123323e-06, + "loss": 1.7266, + "step": 1005 + }, + { + "epoch": 0.7516828721017202, + "eval_loss": 1.6695423126220703, + "eval_runtime": 68.7747, + "eval_samples_per_second": 59.222, + "eval_steps_per_second": 3.708, + "step": 1005 + }, + { + "epoch": 0.7524308152580403, + "grad_norm": 2.1452789734784585, + "learning_rate": 7.3595388494893684e-06, + "loss": 1.7343, + "step": 1006 + }, + { + "epoch": 0.7531787584143605, + "grad_norm": 2.273355009954836, + "learning_rate": 7.354086873093357e-06, + "loss": 1.6011, + "step": 1007 + }, + { + "epoch": 0.7539267015706806, + "grad_norm": 1.8309157607985298, + "learning_rate": 7.3486312982691134e-06, + "loss": 1.6884, + "step": 1008 + }, + { + "epoch": 0.7546746447270007, + "grad_norm": 1.9191865073540744, + "learning_rate": 7.343172133355958e-06, + "loss": 1.4941, + "step": 1009 + }, + { + "epoch": 0.7554225878833208, + "grad_norm": 1.9965399963190864, + "learning_rate": 7.337709386698709e-06, + "loss": 1.7548, + "step": 1010 + }, + { + "epoch": 0.7561705310396409, + "grad_norm": 2.2382350261674016, + "learning_rate": 7.332243066647651e-06, + "loss": 1.8325, + "step": 1011 + }, + { + "epoch": 0.7569184741959611, + "grad_norm": 1.8334707175762837, + "learning_rate": 7.326773181558532e-06, + "loss": 1.6944, + "step": 1012 + }, + { + "epoch": 0.7576664173522812, + "grad_norm": 1.7974084356045958, + "learning_rate": 7.321299739792553e-06, + "loss": 1.7448, + "step": 1013 + }, + { + "epoch": 0.7584143605086013, + "grad_norm": 2.1639273758130364, + "learning_rate": 7.31582274971635e-06, + "loss": 1.6704, + "step": 1014 + }, + { + "epoch": 0.7591623036649214, + "grad_norm": 1.8588849164433847, + "learning_rate": 7.310342219701981e-06, + "loss": 1.5352, + "step": 1015 + }, + { + "epoch": 0.7599102468212415, + "grad_norm": 2.1333508799402434, + "learning_rate": 7.304858158126917e-06, + "loss": 1.7727, + "step": 1016 + }, + { + "epoch": 0.7606581899775617, + "grad_norm": 2.194357248218602, + "learning_rate": 7.299370573374032e-06, + "loss": 1.3242, + "step": 1017 + }, + { + "epoch": 0.7614061331338818, + "grad_norm": 1.7119882143762646, + "learning_rate": 7.293879473831573e-06, + "loss": 1.632, + "step": 1018 + }, + { + "epoch": 0.7621540762902019, + "grad_norm": 2.00178382839197, + "learning_rate": 7.28838486789317e-06, + "loss": 1.5718, + "step": 1019 + }, + { + "epoch": 0.762902019446522, + "grad_norm": 1.8656094340204386, + "learning_rate": 7.282886763957812e-06, + "loss": 1.7046, + "step": 1020 + }, + { + "epoch": 0.7636499626028422, + "grad_norm": 1.8360980100198812, + "learning_rate": 7.2773851704298304e-06, + "loss": 1.6123, + "step": 1021 + }, + { + "epoch": 0.7643979057591623, + "grad_norm": 1.9459967160376235, + "learning_rate": 7.271880095718895e-06, + "loss": 1.5479, + "step": 1022 + }, + { + "epoch": 0.7651458489154824, + "grad_norm": 1.7977271954662462, + "learning_rate": 7.266371548239997e-06, + "loss": 1.6833, + "step": 1023 + }, + { + "epoch": 0.7658937920718025, + "grad_norm": 1.8724602388188325, + "learning_rate": 7.260859536413429e-06, + "loss": 1.5669, + "step": 1024 + }, + { + "epoch": 0.7666417352281226, + "grad_norm": 2.011725110331389, + "learning_rate": 7.255344068664789e-06, + "loss": 1.6768, + "step": 1025 + }, + { + "epoch": 0.7673896783844428, + "grad_norm": 2.4104854316395645, + "learning_rate": 7.249825153424952e-06, + "loss": 1.8344, + "step": 1026 + }, + { + "epoch": 0.7681376215407629, + "grad_norm": 2.0296085341124135, + "learning_rate": 7.244302799130064e-06, + "loss": 1.6496, + "step": 1027 + }, + { + "epoch": 0.768885564697083, + "grad_norm": 4.053888754643613, + "learning_rate": 7.238777014221528e-06, + "loss": 1.6814, + "step": 1028 + }, + { + "epoch": 0.7696335078534031, + "grad_norm": 2.10483101723546, + "learning_rate": 7.233247807145989e-06, + "loss": 1.6536, + "step": 1029 + }, + { + "epoch": 0.7703814510097232, + "grad_norm": 2.1622271675055886, + "learning_rate": 7.227715186355326e-06, + "loss": 1.7709, + "step": 1030 + }, + { + "epoch": 0.7711293941660434, + "grad_norm": 1.8633895348594718, + "learning_rate": 7.222179160306637e-06, + "loss": 1.6992, + "step": 1031 + }, + { + "epoch": 0.7718773373223635, + "grad_norm": 1.7659881224892438, + "learning_rate": 7.216639737462221e-06, + "loss": 1.7891, + "step": 1032 + }, + { + "epoch": 0.7726252804786836, + "grad_norm": 5.643252899257299, + "learning_rate": 7.211096926289572e-06, + "loss": 1.5525, + "step": 1033 + }, + { + "epoch": 0.7733732236350037, + "grad_norm": 2.806998388440608, + "learning_rate": 7.205550735261364e-06, + "loss": 1.6197, + "step": 1034 + }, + { + "epoch": 0.7741211667913238, + "grad_norm": 1.8869986707749684, + "learning_rate": 7.200001172855436e-06, + "loss": 1.579, + "step": 1035 + }, + { + "epoch": 0.774869109947644, + "grad_norm": 1.9527205868033437, + "learning_rate": 7.19444824755478e-06, + "loss": 1.4723, + "step": 1036 + }, + { + "epoch": 0.7756170531039641, + "grad_norm": 2.0148676692592877, + "learning_rate": 7.1888919678475285e-06, + "loss": 1.6524, + "step": 1037 + }, + { + "epoch": 0.7763649962602842, + "grad_norm": 1.9930062099471166, + "learning_rate": 7.183332342226945e-06, + "loss": 1.5441, + "step": 1038 + }, + { + "epoch": 0.7771129394166043, + "grad_norm": 2.1643609771008108, + "learning_rate": 7.1777693791914025e-06, + "loss": 1.6148, + "step": 1039 + }, + { + "epoch": 0.7778608825729244, + "grad_norm": 1.659308107331084, + "learning_rate": 7.1722030872443804e-06, + "loss": 1.6308, + "step": 1040 + }, + { + "epoch": 0.7786088257292446, + "grad_norm": 2.497290436643533, + "learning_rate": 7.1666334748944424e-06, + "loss": 1.6191, + "step": 1041 + }, + { + "epoch": 0.7793567688855647, + "grad_norm": 2.0232996394636316, + "learning_rate": 7.161060550655227e-06, + "loss": 1.551, + "step": 1042 + }, + { + "epoch": 0.7801047120418848, + "grad_norm": 2.2764065515064105, + "learning_rate": 7.155484323045442e-06, + "loss": 1.5017, + "step": 1043 + }, + { + "epoch": 0.7808526551982049, + "grad_norm": 2.694818564894751, + "learning_rate": 7.149904800588838e-06, + "loss": 1.733, + "step": 1044 + }, + { + "epoch": 0.7816005983545251, + "grad_norm": 1.9385236890652935, + "learning_rate": 7.144321991814205e-06, + "loss": 1.7492, + "step": 1045 + }, + { + "epoch": 0.7823485415108452, + "grad_norm": 1.7393397738989245, + "learning_rate": 7.138735905255355e-06, + "loss": 1.6583, + "step": 1046 + }, + { + "epoch": 0.7830964846671653, + "grad_norm": 1.9013701649942905, + "learning_rate": 7.133146549451113e-06, + "loss": 1.7176, + "step": 1047 + }, + { + "epoch": 0.7838444278234854, + "grad_norm": 1.7837380764761694, + "learning_rate": 7.1275539329452984e-06, + "loss": 1.6106, + "step": 1048 + }, + { + "epoch": 0.7845923709798055, + "grad_norm": 3.1058447694028843, + "learning_rate": 7.121958064286714e-06, + "loss": 1.4688, + "step": 1049 + }, + { + "epoch": 0.7853403141361257, + "grad_norm": 1.8274547892182356, + "learning_rate": 7.11635895202914e-06, + "loss": 1.6044, + "step": 1050 + }, + { + "epoch": 0.7860882572924458, + "grad_norm": 2.6282594885565427, + "learning_rate": 7.110756604731307e-06, + "loss": 1.5435, + "step": 1051 + }, + { + "epoch": 0.7868362004487659, + "grad_norm": 2.3085587619291736, + "learning_rate": 7.105151030956895e-06, + "loss": 1.4199, + "step": 1052 + }, + { + "epoch": 0.787584143605086, + "grad_norm": 1.8590086731566444, + "learning_rate": 7.099542239274516e-06, + "loss": 1.4763, + "step": 1053 + }, + { + "epoch": 0.7883320867614061, + "grad_norm": 1.703213284420821, + "learning_rate": 7.0939302382577e-06, + "loss": 1.613, + "step": 1054 + }, + { + "epoch": 0.7890800299177263, + "grad_norm": 1.6802850830213294, + "learning_rate": 7.088315036484882e-06, + "loss": 1.8002, + "step": 1055 + }, + { + "epoch": 0.7898279730740464, + "grad_norm": 1.7227383846409043, + "learning_rate": 7.082696642539391e-06, + "loss": 1.7775, + "step": 1056 + }, + { + "epoch": 0.7905759162303665, + "grad_norm": 1.757881975190259, + "learning_rate": 7.0770750650094335e-06, + "loss": 2.0094, + "step": 1057 + }, + { + "epoch": 0.7913238593866866, + "grad_norm": 2.3049853478314666, + "learning_rate": 7.071450312488084e-06, + "loss": 1.4135, + "step": 1058 + }, + { + "epoch": 0.7920718025430067, + "grad_norm": 2.37315719265942, + "learning_rate": 7.0658223935732715e-06, + "loss": 1.5088, + "step": 1059 + }, + { + "epoch": 0.7928197456993269, + "grad_norm": 1.978535562046256, + "learning_rate": 7.060191316867764e-06, + "loss": 1.7831, + "step": 1060 + }, + { + "epoch": 0.793567688855647, + "grad_norm": 2.024188798415973, + "learning_rate": 7.0545570909791525e-06, + "loss": 1.6477, + "step": 1061 + }, + { + "epoch": 0.7943156320119671, + "grad_norm": 1.7319860223694097, + "learning_rate": 7.048919724519848e-06, + "loss": 1.6294, + "step": 1062 + }, + { + "epoch": 0.7950635751682872, + "grad_norm": 2.000076358516272, + "learning_rate": 7.0432792261070605e-06, + "loss": 1.721, + "step": 1063 + }, + { + "epoch": 0.7958115183246073, + "grad_norm": 1.7782588919841251, + "learning_rate": 7.037635604362786e-06, + "loss": 1.5131, + "step": 1064 + }, + { + "epoch": 0.7965594614809275, + "grad_norm": 1.8044175921710146, + "learning_rate": 7.031988867913793e-06, + "loss": 1.6812, + "step": 1065 + }, + { + "epoch": 0.7973074046372476, + "grad_norm": 2.8785520609064514, + "learning_rate": 7.026339025391615e-06, + "loss": 1.6621, + "step": 1066 + }, + { + "epoch": 0.7980553477935677, + "grad_norm": 1.9253796191555557, + "learning_rate": 7.020686085432532e-06, + "loss": 1.6679, + "step": 1067 + }, + { + "epoch": 0.7988032909498878, + "grad_norm": 2.023321688560482, + "learning_rate": 7.015030056677559e-06, + "loss": 1.575, + "step": 1068 + }, + { + "epoch": 0.7995512341062079, + "grad_norm": 1.7157946379762181, + "learning_rate": 7.00937094777243e-06, + "loss": 1.6698, + "step": 1069 + }, + { + "epoch": 0.8002991772625281, + "grad_norm": 1.927786165218989, + "learning_rate": 7.00370876736759e-06, + "loss": 1.7783, + "step": 1070 + }, + { + "epoch": 0.8010471204188482, + "grad_norm": 1.9566968453771416, + "learning_rate": 6.9980435241181785e-06, + "loss": 1.6471, + "step": 1071 + }, + { + "epoch": 0.8017950635751683, + "grad_norm": 2.212203046822685, + "learning_rate": 6.992375226684016e-06, + "loss": 1.6964, + "step": 1072 + }, + { + "epoch": 0.8025430067314884, + "grad_norm": 1.9841463693819683, + "learning_rate": 6.986703883729595e-06, + "loss": 1.8109, + "step": 1073 + }, + { + "epoch": 0.8032909498878086, + "grad_norm": 2.794174763440219, + "learning_rate": 6.981029503924059e-06, + "loss": 1.7234, + "step": 1074 + }, + { + "epoch": 0.8040388930441287, + "grad_norm": 2.5904307702524547, + "learning_rate": 6.975352095941194e-06, + "loss": 1.5591, + "step": 1075 + }, + { + "epoch": 0.8047868362004488, + "grad_norm": 1.8951088815680908, + "learning_rate": 6.969671668459414e-06, + "loss": 1.5326, + "step": 1076 + }, + { + "epoch": 0.8055347793567689, + "grad_norm": 1.7663172827919795, + "learning_rate": 6.963988230161757e-06, + "loss": 1.6596, + "step": 1077 + }, + { + "epoch": 0.806282722513089, + "grad_norm": 1.7528066910341538, + "learning_rate": 6.958301789735853e-06, + "loss": 1.8226, + "step": 1078 + }, + { + "epoch": 0.8070306656694092, + "grad_norm": 3.735684369169787, + "learning_rate": 6.952612355873922e-06, + "loss": 1.6697, + "step": 1079 + }, + { + "epoch": 0.8077786088257293, + "grad_norm": 1.6983678110336682, + "learning_rate": 6.946919937272768e-06, + "loss": 1.8468, + "step": 1080 + }, + { + "epoch": 0.8085265519820494, + "grad_norm": 1.8687830478824856, + "learning_rate": 6.941224542633748e-06, + "loss": 1.7359, + "step": 1081 + }, + { + "epoch": 0.8092744951383695, + "grad_norm": 2.629846372817297, + "learning_rate": 6.935526180662776e-06, + "loss": 1.7665, + "step": 1082 + }, + { + "epoch": 0.8100224382946896, + "grad_norm": 1.738286127140632, + "learning_rate": 6.929824860070295e-06, + "loss": 1.8021, + "step": 1083 + }, + { + "epoch": 0.8107703814510098, + "grad_norm": 1.9707608813216067, + "learning_rate": 6.924120589571275e-06, + "loss": 1.7022, + "step": 1084 + }, + { + "epoch": 0.8115183246073299, + "grad_norm": 1.7401040318804621, + "learning_rate": 6.918413377885193e-06, + "loss": 1.845, + "step": 1085 + }, + { + "epoch": 0.81226626776365, + "grad_norm": 2.073362802829311, + "learning_rate": 6.91270323373602e-06, + "loss": 1.6391, + "step": 1086 + }, + { + "epoch": 0.8130142109199701, + "grad_norm": 2.1792120754766637, + "learning_rate": 6.906990165852218e-06, + "loss": 1.5638, + "step": 1087 + }, + { + "epoch": 0.8137621540762902, + "grad_norm": 1.7342173680102193, + "learning_rate": 6.901274182966709e-06, + "loss": 1.8254, + "step": 1088 + }, + { + "epoch": 0.8145100972326104, + "grad_norm": 1.810670372615815, + "learning_rate": 6.895555293816871e-06, + "loss": 1.5454, + "step": 1089 + }, + { + "epoch": 0.8152580403889305, + "grad_norm": 2.070821657667326, + "learning_rate": 6.889833507144534e-06, + "loss": 1.8195, + "step": 1090 + }, + { + "epoch": 0.8160059835452506, + "grad_norm": 1.733571952092035, + "learning_rate": 6.884108831695946e-06, + "loss": 1.9, + "step": 1091 + }, + { + "epoch": 0.8167539267015707, + "grad_norm": 2.093804471152808, + "learning_rate": 6.878381276221777e-06, + "loss": 1.3788, + "step": 1092 + }, + { + "epoch": 0.8175018698578908, + "grad_norm": 1.807747463992071, + "learning_rate": 6.872650849477098e-06, + "loss": 1.7503, + "step": 1093 + }, + { + "epoch": 0.818249813014211, + "grad_norm": 2.096838491182964, + "learning_rate": 6.866917560221369e-06, + "loss": 1.6161, + "step": 1094 + }, + { + "epoch": 0.818997756170531, + "grad_norm": 1.9376430650703855, + "learning_rate": 6.861181417218425e-06, + "loss": 1.6198, + "step": 1095 + }, + { + "epoch": 0.8197456993268512, + "grad_norm": 2.171355262975323, + "learning_rate": 6.855442429236463e-06, + "loss": 1.6066, + "step": 1096 + }, + { + "epoch": 0.8204936424831712, + "grad_norm": 2.3874099514943805, + "learning_rate": 6.849700605048031e-06, + "loss": 1.7574, + "step": 1097 + }, + { + "epoch": 0.8212415856394913, + "grad_norm": 1.9715538558005241, + "learning_rate": 6.843955953430008e-06, + "loss": 1.6433, + "step": 1098 + }, + { + "epoch": 0.8219895287958116, + "grad_norm": 1.8524201524397321, + "learning_rate": 6.838208483163601e-06, + "loss": 1.6868, + "step": 1099 + }, + { + "epoch": 0.8227374719521316, + "grad_norm": 1.8634123724486202, + "learning_rate": 6.832458203034321e-06, + "loss": 1.6924, + "step": 1100 + }, + { + "epoch": 0.8234854151084517, + "grad_norm": 2.415179329084346, + "learning_rate": 6.8267051218319766e-06, + "loss": 1.5566, + "step": 1101 + }, + { + "epoch": 0.8242333582647718, + "grad_norm": 2.181018462669017, + "learning_rate": 6.820949248350653e-06, + "loss": 1.7142, + "step": 1102 + }, + { + "epoch": 0.824981301421092, + "grad_norm": 2.8105922826404286, + "learning_rate": 6.815190591388713e-06, + "loss": 1.3655, + "step": 1103 + }, + { + "epoch": 0.8257292445774121, + "grad_norm": 1.9814296854554856, + "learning_rate": 6.809429159748766e-06, + "loss": 1.8322, + "step": 1104 + }, + { + "epoch": 0.8264771877337322, + "grad_norm": 1.7512996663741385, + "learning_rate": 6.803664962237665e-06, + "loss": 1.5874, + "step": 1105 + }, + { + "epoch": 0.8272251308900523, + "grad_norm": 1.8007994783073062, + "learning_rate": 6.797898007666493e-06, + "loss": 1.6375, + "step": 1106 + }, + { + "epoch": 0.8279730740463724, + "grad_norm": 1.9495555417010324, + "learning_rate": 6.792128304850543e-06, + "loss": 1.7608, + "step": 1107 + }, + { + "epoch": 0.8287210172026926, + "grad_norm": 2.4010756553692882, + "learning_rate": 6.786355862609315e-06, + "loss": 1.7942, + "step": 1108 + }, + { + "epoch": 0.8294689603590127, + "grad_norm": 1.8421556348444412, + "learning_rate": 6.780580689766492e-06, + "loss": 1.3507, + "step": 1109 + }, + { + "epoch": 0.8302169035153328, + "grad_norm": 2.167493162660865, + "learning_rate": 6.774802795149932e-06, + "loss": 1.6245, + "step": 1110 + }, + { + "epoch": 0.8309648466716529, + "grad_norm": 2.0249387837880923, + "learning_rate": 6.769022187591654e-06, + "loss": 1.8135, + "step": 1111 + }, + { + "epoch": 0.831712789827973, + "grad_norm": 1.8029399209262258, + "learning_rate": 6.7632388759278225e-06, + "loss": 1.5994, + "step": 1112 + }, + { + "epoch": 0.8324607329842932, + "grad_norm": 2.0923373540715375, + "learning_rate": 6.757452868998737e-06, + "loss": 1.9384, + "step": 1113 + }, + { + "epoch": 0.8332086761406133, + "grad_norm": 2.04874587793283, + "learning_rate": 6.751664175648817e-06, + "loss": 1.6796, + "step": 1114 + }, + { + "epoch": 0.8339566192969334, + "grad_norm": 1.7146443574255779, + "learning_rate": 6.745872804726588e-06, + "loss": 1.6267, + "step": 1115 + }, + { + "epoch": 0.8347045624532535, + "grad_norm": 1.8507218225855218, + "learning_rate": 6.74007876508467e-06, + "loss": 1.7471, + "step": 1116 + }, + { + "epoch": 0.8354525056095736, + "grad_norm": 2.17729909224941, + "learning_rate": 6.734282065579757e-06, + "loss": 1.6365, + "step": 1117 + }, + { + "epoch": 0.8362004487658938, + "grad_norm": 1.7210165355905482, + "learning_rate": 6.728482715072615e-06, + "loss": 1.6715, + "step": 1118 + }, + { + "epoch": 0.8369483919222139, + "grad_norm": 2.8765049585015365, + "learning_rate": 6.722680722428058e-06, + "loss": 1.7771, + "step": 1119 + }, + { + "epoch": 0.837696335078534, + "grad_norm": 1.9524980040560516, + "learning_rate": 6.716876096514944e-06, + "loss": 1.7715, + "step": 1120 + }, + { + "epoch": 0.8384442782348541, + "grad_norm": 1.9102645967507785, + "learning_rate": 6.711068846206149e-06, + "loss": 1.6123, + "step": 1121 + }, + { + "epoch": 0.8391922213911742, + "grad_norm": 2.7509143575551374, + "learning_rate": 6.705258980378565e-06, + "loss": 1.7084, + "step": 1122 + }, + { + "epoch": 0.8399401645474944, + "grad_norm": 2.0982858637215838, + "learning_rate": 6.699446507913083e-06, + "loss": 1.4011, + "step": 1123 + }, + { + "epoch": 0.8406881077038145, + "grad_norm": 1.940207992362092, + "learning_rate": 6.693631437694579e-06, + "loss": 1.8598, + "step": 1124 + }, + { + "epoch": 0.8414360508601346, + "grad_norm": 2.6746757201277354, + "learning_rate": 6.687813778611893e-06, + "loss": 1.7315, + "step": 1125 + }, + { + "epoch": 0.8421839940164547, + "grad_norm": 1.7971742827461676, + "learning_rate": 6.681993539557828e-06, + "loss": 1.6165, + "step": 1126 + }, + { + "epoch": 0.8429319371727748, + "grad_norm": 1.9459778054926844, + "learning_rate": 6.676170729429132e-06, + "loss": 1.5216, + "step": 1127 + }, + { + "epoch": 0.843679880329095, + "grad_norm": 2.0482552741742226, + "learning_rate": 6.670345357126481e-06, + "loss": 1.6303, + "step": 1128 + }, + { + "epoch": 0.8444278234854151, + "grad_norm": 1.782389085965577, + "learning_rate": 6.664517431554466e-06, + "loss": 1.3927, + "step": 1129 + }, + { + "epoch": 0.8451757666417352, + "grad_norm": 2.1342988912284295, + "learning_rate": 6.658686961621582e-06, + "loss": 1.6647, + "step": 1130 + }, + { + "epoch": 0.8459237097980553, + "grad_norm": 1.705490565154237, + "learning_rate": 6.652853956240215e-06, + "loss": 1.7295, + "step": 1131 + }, + { + "epoch": 0.8466716529543755, + "grad_norm": 3.1563399568505157, + "learning_rate": 6.6470184243266235e-06, + "loss": 1.5404, + "step": 1132 + }, + { + "epoch": 0.8474195961106956, + "grad_norm": 2.0582350124304494, + "learning_rate": 6.641180374800931e-06, + "loss": 1.56, + "step": 1133 + }, + { + "epoch": 0.8481675392670157, + "grad_norm": 2.218522640955612, + "learning_rate": 6.635339816587109e-06, + "loss": 1.579, + "step": 1134 + }, + { + "epoch": 0.8489154824233358, + "grad_norm": 1.8414403748183819, + "learning_rate": 6.6294967586129614e-06, + "loss": 1.6013, + "step": 1135 + }, + { + "epoch": 0.8496634255796559, + "grad_norm": 1.6079264937451831, + "learning_rate": 6.623651209810116e-06, + "loss": 1.4789, + "step": 1136 + }, + { + "epoch": 0.8504113687359761, + "grad_norm": 1.7557174789912269, + "learning_rate": 6.617803179114005e-06, + "loss": 1.704, + "step": 1137 + }, + { + "epoch": 0.8511593118922962, + "grad_norm": 1.8290171928006018, + "learning_rate": 6.611952675463859e-06, + "loss": 1.7809, + "step": 1138 + }, + { + "epoch": 0.8519072550486163, + "grad_norm": 1.8314468754098587, + "learning_rate": 6.606099707802682e-06, + "loss": 1.6675, + "step": 1139 + }, + { + "epoch": 0.8526551982049364, + "grad_norm": 2.0668800082750374, + "learning_rate": 6.600244285077251e-06, + "loss": 1.5501, + "step": 1140 + }, + { + "epoch": 0.8534031413612565, + "grad_norm": 2.073378435981234, + "learning_rate": 6.594386416238095e-06, + "loss": 1.6723, + "step": 1141 + }, + { + "epoch": 0.8541510845175767, + "grad_norm": 2.7688792172539936, + "learning_rate": 6.588526110239475e-06, + "loss": 1.6508, + "step": 1142 + }, + { + "epoch": 0.8548990276738968, + "grad_norm": 1.8750636698792063, + "learning_rate": 6.582663376039387e-06, + "loss": 1.5505, + "step": 1143 + }, + { + "epoch": 0.8556469708302169, + "grad_norm": 2.2620693110481316, + "learning_rate": 6.576798222599532e-06, + "loss": 1.5696, + "step": 1144 + }, + { + "epoch": 0.856394913986537, + "grad_norm": 2.200940790789423, + "learning_rate": 6.570930658885314e-06, + "loss": 1.7066, + "step": 1145 + }, + { + "epoch": 0.8571428571428571, + "grad_norm": 2.689207291576637, + "learning_rate": 6.565060693865815e-06, + "loss": 1.739, + "step": 1146 + }, + { + "epoch": 0.8578908002991773, + "grad_norm": 1.753487001815811, + "learning_rate": 6.559188336513794e-06, + "loss": 1.6248, + "step": 1147 + }, + { + "epoch": 0.8586387434554974, + "grad_norm": 1.851324904274212, + "learning_rate": 6.553313595805666e-06, + "loss": 1.4246, + "step": 1148 + }, + { + "epoch": 0.8593866866118175, + "grad_norm": 2.009567598549325, + "learning_rate": 6.547436480721482e-06, + "loss": 1.6636, + "step": 1149 + }, + { + "epoch": 0.8601346297681376, + "grad_norm": 2.202796891371022, + "learning_rate": 6.5415570002449315e-06, + "loss": 1.6649, + "step": 1150 + }, + { + "epoch": 0.8608825729244577, + "grad_norm": 2.0882429693941673, + "learning_rate": 6.535675163363315e-06, + "loss": 1.7353, + "step": 1151 + }, + { + "epoch": 0.8616305160807779, + "grad_norm": 2.156126047721483, + "learning_rate": 6.5297909790675375e-06, + "loss": 1.4541, + "step": 1152 + }, + { + "epoch": 0.862378459237098, + "grad_norm": 1.9857524697496662, + "learning_rate": 6.523904456352089e-06, + "loss": 1.6997, + "step": 1153 + }, + { + "epoch": 0.8631264023934181, + "grad_norm": 1.8173037786155135, + "learning_rate": 6.518015604215037e-06, + "loss": 1.6337, + "step": 1154 + }, + { + "epoch": 0.8638743455497382, + "grad_norm": 2.0841346584418616, + "learning_rate": 6.512124431658006e-06, + "loss": 1.3822, + "step": 1155 + }, + { + "epoch": 0.8646222887060584, + "grad_norm": 1.8704002976759626, + "learning_rate": 6.5062309476861714e-06, + "loss": 1.8382, + "step": 1156 + }, + { + "epoch": 0.8653702318623785, + "grad_norm": 1.9832545200670035, + "learning_rate": 6.500335161308242e-06, + "loss": 1.6475, + "step": 1157 + }, + { + "epoch": 0.8661181750186986, + "grad_norm": 1.9544505011631417, + "learning_rate": 6.494437081536442e-06, + "loss": 1.5341, + "step": 1158 + }, + { + "epoch": 0.8668661181750187, + "grad_norm": 2.2554645469652157, + "learning_rate": 6.488536717386504e-06, + "loss": 1.5171, + "step": 1159 + }, + { + "epoch": 0.8676140613313388, + "grad_norm": 1.978784901255462, + "learning_rate": 6.4826340778776534e-06, + "loss": 1.6465, + "step": 1160 + }, + { + "epoch": 0.868362004487659, + "grad_norm": 1.8972941980278133, + "learning_rate": 6.47672917203259e-06, + "loss": 1.3133, + "step": 1161 + }, + { + "epoch": 0.8691099476439791, + "grad_norm": 1.9881040776807108, + "learning_rate": 6.470822008877482e-06, + "loss": 1.8071, + "step": 1162 + }, + { + "epoch": 0.8698578908002992, + "grad_norm": 1.9398609025469538, + "learning_rate": 6.464912597441947e-06, + "loss": 1.5962, + "step": 1163 + }, + { + "epoch": 0.8706058339566193, + "grad_norm": 1.8127854329892805, + "learning_rate": 6.4590009467590366e-06, + "loss": 1.5932, + "step": 1164 + }, + { + "epoch": 0.8713537771129394, + "grad_norm": 3.33269270114937, + "learning_rate": 6.453087065865229e-06, + "loss": 1.8213, + "step": 1165 + }, + { + "epoch": 0.8721017202692596, + "grad_norm": 2.36879778688533, + "learning_rate": 6.447170963800408e-06, + "loss": 1.4386, + "step": 1166 + }, + { + "epoch": 0.8728496634255797, + "grad_norm": 2.0817829052656953, + "learning_rate": 6.4412526496078555e-06, + "loss": 1.7536, + "step": 1167 + }, + { + "epoch": 0.8735976065818998, + "grad_norm": 1.8191735223787528, + "learning_rate": 6.435332132334233e-06, + "loss": 1.5064, + "step": 1168 + }, + { + "epoch": 0.8743455497382199, + "grad_norm": 1.6526772839622264, + "learning_rate": 6.4294094210295725e-06, + "loss": 1.5125, + "step": 1169 + }, + { + "epoch": 0.87509349289454, + "grad_norm": 2.1791151688302577, + "learning_rate": 6.423484524747255e-06, + "loss": 1.7244, + "step": 1170 + }, + { + "epoch": 0.8758414360508602, + "grad_norm": 2.2673531372330715, + "learning_rate": 6.4175574525440065e-06, + "loss": 1.4952, + "step": 1171 + }, + { + "epoch": 0.8765893792071803, + "grad_norm": 2.3111408805409597, + "learning_rate": 6.411628213479875e-06, + "loss": 1.8462, + "step": 1172 + }, + { + "epoch": 0.8773373223635004, + "grad_norm": 1.824837863480445, + "learning_rate": 6.405696816618222e-06, + "loss": 1.6346, + "step": 1173 + }, + { + "epoch": 0.8780852655198205, + "grad_norm": 2.3392333923481186, + "learning_rate": 6.3997632710257095e-06, + "loss": 1.4062, + "step": 1174 + }, + { + "epoch": 0.8788332086761406, + "grad_norm": 1.755324474285726, + "learning_rate": 6.393827585772281e-06, + "loss": 1.5238, + "step": 1175 + }, + { + "epoch": 0.8795811518324608, + "grad_norm": 1.836654424222221, + "learning_rate": 6.3878897699311525e-06, + "loss": 1.7412, + "step": 1176 + }, + { + "epoch": 0.8803290949887809, + "grad_norm": 3.0529718087588873, + "learning_rate": 6.381949832578796e-06, + "loss": 1.5471, + "step": 1177 + }, + { + "epoch": 0.881077038145101, + "grad_norm": 1.6945179733732993, + "learning_rate": 6.376007782794926e-06, + "loss": 1.6164, + "step": 1178 + }, + { + "epoch": 0.8818249813014211, + "grad_norm": 1.8924066588076536, + "learning_rate": 6.370063629662488e-06, + "loss": 1.4301, + "step": 1179 + }, + { + "epoch": 0.8825729244577412, + "grad_norm": 1.8898668167085229, + "learning_rate": 6.364117382267642e-06, + "loss": 1.613, + "step": 1180 + }, + { + "epoch": 0.8833208676140614, + "grad_norm": 2.0262059252495987, + "learning_rate": 6.358169049699746e-06, + "loss": 1.6305, + "step": 1181 + }, + { + "epoch": 0.8840688107703815, + "grad_norm": 2.1927562253473054, + "learning_rate": 6.352218641051349e-06, + "loss": 1.5173, + "step": 1182 + }, + { + "epoch": 0.8848167539267016, + "grad_norm": 2.4181329673017897, + "learning_rate": 6.346266165418173e-06, + "loss": 1.7355, + "step": 1183 + }, + { + "epoch": 0.8855646970830217, + "grad_norm": 1.7499246769077093, + "learning_rate": 6.340311631899099e-06, + "loss": 1.7652, + "step": 1184 + }, + { + "epoch": 0.8863126402393419, + "grad_norm": 1.9421483884054282, + "learning_rate": 6.334355049596155e-06, + "loss": 1.6063, + "step": 1185 + }, + { + "epoch": 0.887060583395662, + "grad_norm": 2.1878722562530344, + "learning_rate": 6.328396427614496e-06, + "loss": 1.6497, + "step": 1186 + }, + { + "epoch": 0.8878085265519821, + "grad_norm": 1.8447330612983637, + "learning_rate": 6.3224357750624014e-06, + "loss": 1.4905, + "step": 1187 + }, + { + "epoch": 0.8885564697083022, + "grad_norm": 1.8967326567117848, + "learning_rate": 6.316473101051251e-06, + "loss": 1.6036, + "step": 1188 + }, + { + "epoch": 0.8893044128646223, + "grad_norm": 1.855060843009185, + "learning_rate": 6.310508414695511e-06, + "loss": 1.679, + "step": 1189 + }, + { + "epoch": 0.8900523560209425, + "grad_norm": 2.264010443231244, + "learning_rate": 6.304541725112734e-06, + "loss": 1.6173, + "step": 1190 + }, + { + "epoch": 0.8908002991772626, + "grad_norm": 1.8215795955490806, + "learning_rate": 6.298573041423523e-06, + "loss": 1.7204, + "step": 1191 + }, + { + "epoch": 0.8915482423335827, + "grad_norm": 1.7785077390338773, + "learning_rate": 6.292602372751536e-06, + "loss": 1.7242, + "step": 1192 + }, + { + "epoch": 0.8922961854899027, + "grad_norm": 1.807964932460702, + "learning_rate": 6.2866297282234654e-06, + "loss": 1.6197, + "step": 1193 + }, + { + "epoch": 0.8930441286462228, + "grad_norm": 1.8656484376517364, + "learning_rate": 6.280655116969022e-06, + "loss": 1.4489, + "step": 1194 + }, + { + "epoch": 0.893792071802543, + "grad_norm": 2.007987909636999, + "learning_rate": 6.274678548120921e-06, + "loss": 1.6396, + "step": 1195 + }, + { + "epoch": 0.8945400149588631, + "grad_norm": 1.6423491150785015, + "learning_rate": 6.268700030814871e-06, + "loss": 1.8216, + "step": 1196 + }, + { + "epoch": 0.8952879581151832, + "grad_norm": 1.8525813165236313, + "learning_rate": 6.262719574189564e-06, + "loss": 1.776, + "step": 1197 + }, + { + "epoch": 0.8960359012715033, + "grad_norm": 1.843186876288332, + "learning_rate": 6.25673718738665e-06, + "loss": 1.7521, + "step": 1198 + }, + { + "epoch": 0.8967838444278234, + "grad_norm": 2.230228363318977, + "learning_rate": 6.2507528795507326e-06, + "loss": 1.6952, + "step": 1199 + }, + { + "epoch": 0.8975317875841436, + "grad_norm": 1.8089814643907889, + "learning_rate": 6.244766659829351e-06, + "loss": 1.5959, + "step": 1200 + }, + { + "epoch": 0.8982797307404637, + "grad_norm": 2.465894850150939, + "learning_rate": 6.238778537372966e-06, + "loss": 1.4491, + "step": 1201 + }, + { + "epoch": 0.8990276738967838, + "grad_norm": 3.088379823826826, + "learning_rate": 6.232788521334948e-06, + "loss": 1.6138, + "step": 1202 + }, + { + "epoch": 0.8997756170531039, + "grad_norm": 2.2902482381328904, + "learning_rate": 6.226796620871563e-06, + "loss": 1.7367, + "step": 1203 + }, + { + "epoch": 0.900523560209424, + "grad_norm": 1.935849606962388, + "learning_rate": 6.2208028451419575e-06, + "loss": 1.574, + "step": 1204 + }, + { + "epoch": 0.9012715033657442, + "grad_norm": 2.112068469925092, + "learning_rate": 6.214807203308141e-06, + "loss": 1.5015, + "step": 1205 + }, + { + "epoch": 0.9020194465220643, + "grad_norm": 1.9768373995678108, + "learning_rate": 6.208809704534977e-06, + "loss": 1.7494, + "step": 1206 + }, + { + "epoch": 0.9027673896783844, + "grad_norm": 2.404271671899643, + "learning_rate": 6.2028103579901725e-06, + "loss": 1.5931, + "step": 1207 + }, + { + "epoch": 0.9035153328347045, + "grad_norm": 1.7543522347528944, + "learning_rate": 6.196809172844252e-06, + "loss": 1.7983, + "step": 1208 + }, + { + "epoch": 0.9042632759910246, + "grad_norm": 2.2101994761782904, + "learning_rate": 6.1908061582705506e-06, + "loss": 1.7281, + "step": 1209 + }, + { + "epoch": 0.9050112191473448, + "grad_norm": 1.8504922367562906, + "learning_rate": 6.184801323445208e-06, + "loss": 1.8025, + "step": 1210 + }, + { + "epoch": 0.9057591623036649, + "grad_norm": 2.023087955025797, + "learning_rate": 6.178794677547138e-06, + "loss": 1.448, + "step": 1211 + }, + { + "epoch": 0.906507105459985, + "grad_norm": 2.0106099045520183, + "learning_rate": 6.1727862297580245e-06, + "loss": 1.5648, + "step": 1212 + }, + { + "epoch": 0.9072550486163051, + "grad_norm": 1.9328982071159035, + "learning_rate": 6.166775989262309e-06, + "loss": 1.5578, + "step": 1213 + }, + { + "epoch": 0.9080029917726253, + "grad_norm": 1.8203793870571643, + "learning_rate": 6.160763965247171e-06, + "loss": 1.5989, + "step": 1214 + }, + { + "epoch": 0.9087509349289454, + "grad_norm": 2.4905428506685143, + "learning_rate": 6.154750166902516e-06, + "loss": 1.6774, + "step": 1215 + }, + { + "epoch": 0.9094988780852655, + "grad_norm": 1.973807462798449, + "learning_rate": 6.1487346034209605e-06, + "loss": 1.6241, + "step": 1216 + }, + { + "epoch": 0.9102468212415856, + "grad_norm": 2.0480159943691545, + "learning_rate": 6.142717283997823e-06, + "loss": 1.7351, + "step": 1217 + }, + { + "epoch": 0.9109947643979057, + "grad_norm": 1.9205616853721408, + "learning_rate": 6.136698217831106e-06, + "loss": 1.7068, + "step": 1218 + }, + { + "epoch": 0.9117427075542259, + "grad_norm": 1.965990134203855, + "learning_rate": 6.130677414121474e-06, + "loss": 1.982, + "step": 1219 + }, + { + "epoch": 0.912490650710546, + "grad_norm": 2.088389769004117, + "learning_rate": 6.124654882072258e-06, + "loss": 1.8733, + "step": 1220 + }, + { + "epoch": 0.9132385938668661, + "grad_norm": 1.9594862303790421, + "learning_rate": 6.118630630889426e-06, + "loss": 1.8926, + "step": 1221 + }, + { + "epoch": 0.9139865370231862, + "grad_norm": 1.9543402721653793, + "learning_rate": 6.112604669781572e-06, + "loss": 1.6923, + "step": 1222 + }, + { + "epoch": 0.9147344801795063, + "grad_norm": 1.6839045034832127, + "learning_rate": 6.10657700795991e-06, + "loss": 1.5843, + "step": 1223 + }, + { + "epoch": 0.9154824233358265, + "grad_norm": 1.8755804521578787, + "learning_rate": 6.100547654638245e-06, + "loss": 1.7624, + "step": 1224 + }, + { + "epoch": 0.9162303664921466, + "grad_norm": 2.435239034744799, + "learning_rate": 6.094516619032975e-06, + "loss": 1.6804, + "step": 1225 + }, + { + "epoch": 0.9169783096484667, + "grad_norm": 1.9331087565524072, + "learning_rate": 6.088483910363065e-06, + "loss": 1.6713, + "step": 1226 + }, + { + "epoch": 0.9177262528047868, + "grad_norm": 2.076391569827481, + "learning_rate": 6.082449537850043e-06, + "loss": 1.6451, + "step": 1227 + }, + { + "epoch": 0.9184741959611069, + "grad_norm": 1.9717578555423976, + "learning_rate": 6.076413510717973e-06, + "loss": 1.5012, + "step": 1228 + }, + { + "epoch": 0.9192221391174271, + "grad_norm": 2.0242101316297583, + "learning_rate": 6.070375838193453e-06, + "loss": 1.8146, + "step": 1229 + }, + { + "epoch": 0.9199700822737472, + "grad_norm": 2.0289219936163683, + "learning_rate": 6.064336529505594e-06, + "loss": 1.6571, + "step": 1230 + }, + { + "epoch": 0.9207180254300673, + "grad_norm": 1.94326017584445, + "learning_rate": 6.058295593886013e-06, + "loss": 1.5005, + "step": 1231 + }, + { + "epoch": 0.9214659685863874, + "grad_norm": 2.441286831156838, + "learning_rate": 6.052253040568804e-06, + "loss": 1.8966, + "step": 1232 + }, + { + "epoch": 0.9222139117427075, + "grad_norm": 1.9518965823078522, + "learning_rate": 6.046208878790543e-06, + "loss": 1.738, + "step": 1233 + }, + { + "epoch": 0.9229618548990277, + "grad_norm": 1.7994481225747099, + "learning_rate": 6.040163117790261e-06, + "loss": 1.718, + "step": 1234 + }, + { + "epoch": 0.9237097980553478, + "grad_norm": 1.9204448251064423, + "learning_rate": 6.034115766809434e-06, + "loss": 1.4272, + "step": 1235 + }, + { + "epoch": 0.9244577412116679, + "grad_norm": 1.6174821188257187, + "learning_rate": 6.028066835091968e-06, + "loss": 1.4777, + "step": 1236 + }, + { + "epoch": 0.925205684367988, + "grad_norm": 2.191215963732871, + "learning_rate": 6.022016331884185e-06, + "loss": 1.8531, + "step": 1237 + }, + { + "epoch": 0.9259536275243081, + "grad_norm": 2.22482935918896, + "learning_rate": 6.01596426643481e-06, + "loss": 1.7943, + "step": 1238 + }, + { + "epoch": 0.9267015706806283, + "grad_norm": 1.9180342001935813, + "learning_rate": 6.009910647994956e-06, + "loss": 1.6207, + "step": 1239 + }, + { + "epoch": 0.9274495138369484, + "grad_norm": 1.663718873373362, + "learning_rate": 6.003855485818111e-06, + "loss": 1.6016, + "step": 1240 + }, + { + "epoch": 0.9281974569932685, + "grad_norm": 1.7744786609044814, + "learning_rate": 5.997798789160121e-06, + "loss": 1.5481, + "step": 1241 + }, + { + "epoch": 0.9289454001495886, + "grad_norm": 1.8485869008228815, + "learning_rate": 5.991740567279174e-06, + "loss": 1.5233, + "step": 1242 + }, + { + "epoch": 0.9296933433059088, + "grad_norm": 1.657788150851475, + "learning_rate": 5.985680829435798e-06, + "loss": 1.5234, + "step": 1243 + }, + { + "epoch": 0.9304412864622289, + "grad_norm": 2.106768695395073, + "learning_rate": 5.979619584892834e-06, + "loss": 1.5991, + "step": 1244 + }, + { + "epoch": 0.931189229618549, + "grad_norm": 1.6608952912299104, + "learning_rate": 5.973556842915422e-06, + "loss": 1.8989, + "step": 1245 + }, + { + "epoch": 0.9319371727748691, + "grad_norm": 2.10739988666985, + "learning_rate": 5.967492612770999e-06, + "loss": 1.6571, + "step": 1246 + }, + { + "epoch": 0.9326851159311892, + "grad_norm": 2.3322560555128042, + "learning_rate": 5.961426903729271e-06, + "loss": 1.2691, + "step": 1247 + }, + { + "epoch": 0.9334330590875094, + "grad_norm": 1.8987145975164443, + "learning_rate": 5.955359725062203e-06, + "loss": 1.6099, + "step": 1248 + }, + { + "epoch": 0.9341810022438295, + "grad_norm": 1.823145900343624, + "learning_rate": 5.9492910860440135e-06, + "loss": 1.7379, + "step": 1249 + }, + { + "epoch": 0.9349289454001496, + "grad_norm": 1.9954316699120402, + "learning_rate": 5.94322099595115e-06, + "loss": 1.4049, + "step": 1250 + }, + { + "epoch": 0.9356768885564697, + "grad_norm": 1.8104378798989003, + "learning_rate": 5.937149464062275e-06, + "loss": 1.5567, + "step": 1251 + }, + { + "epoch": 0.9364248317127898, + "grad_norm": 1.9883287281788995, + "learning_rate": 5.931076499658258e-06, + "loss": 1.5457, + "step": 1252 + }, + { + "epoch": 0.93717277486911, + "grad_norm": 1.9575241882372252, + "learning_rate": 5.925002112022158e-06, + "loss": 1.8762, + "step": 1253 + }, + { + "epoch": 0.9379207180254301, + "grad_norm": 1.8299550604573904, + "learning_rate": 5.9189263104392104e-06, + "loss": 1.6207, + "step": 1254 + }, + { + "epoch": 0.9386686611817502, + "grad_norm": 1.6963794879306424, + "learning_rate": 5.91284910419681e-06, + "loss": 1.6916, + "step": 1255 + }, + { + "epoch": 0.9394166043380703, + "grad_norm": 1.9920235786659242, + "learning_rate": 5.906770502584497e-06, + "loss": 1.5733, + "step": 1256 + }, + { + "epoch": 0.9401645474943904, + "grad_norm": 1.7508272237065152, + "learning_rate": 5.900690514893951e-06, + "loss": 1.7904, + "step": 1257 + }, + { + "epoch": 0.9409124906507106, + "grad_norm": 1.9559441666476383, + "learning_rate": 5.894609150418964e-06, + "loss": 1.3888, + "step": 1258 + }, + { + "epoch": 0.9416604338070307, + "grad_norm": 2.8588365909544855, + "learning_rate": 5.888526418455436e-06, + "loss": 1.6425, + "step": 1259 + }, + { + "epoch": 0.9424083769633508, + "grad_norm": 1.7772980223911854, + "learning_rate": 5.882442328301356e-06, + "loss": 1.7347, + "step": 1260 + }, + { + "epoch": 0.9431563201196709, + "grad_norm": 3.359861031667959, + "learning_rate": 5.876356889256788e-06, + "loss": 1.9901, + "step": 1261 + }, + { + "epoch": 0.943904263275991, + "grad_norm": 1.9386885870231858, + "learning_rate": 5.870270110623859e-06, + "loss": 1.6087, + "step": 1262 + }, + { + "epoch": 0.9446522064323112, + "grad_norm": 2.632125697442696, + "learning_rate": 5.864182001706747e-06, + "loss": 1.9098, + "step": 1263 + }, + { + "epoch": 0.9454001495886313, + "grad_norm": 1.992592950829824, + "learning_rate": 5.858092571811661e-06, + "loss": 1.6321, + "step": 1264 + }, + { + "epoch": 0.9461480927449514, + "grad_norm": 1.9519635057990323, + "learning_rate": 5.852001830246823e-06, + "loss": 1.7042, + "step": 1265 + }, + { + "epoch": 0.9468960359012715, + "grad_norm": 1.903107743118032, + "learning_rate": 5.8459097863224705e-06, + "loss": 1.7101, + "step": 1266 + }, + { + "epoch": 0.9476439790575916, + "grad_norm": 1.9598955997859357, + "learning_rate": 5.839816449350824e-06, + "loss": 1.732, + "step": 1267 + }, + { + "epoch": 0.9483919222139118, + "grad_norm": 1.8780588994561271, + "learning_rate": 5.833721828646086e-06, + "loss": 1.759, + "step": 1268 + }, + { + "epoch": 0.9491398653702319, + "grad_norm": 1.921893708169811, + "learning_rate": 5.827625933524418e-06, + "loss": 1.6741, + "step": 1269 + }, + { + "epoch": 0.949887808526552, + "grad_norm": 2.0861341514008083, + "learning_rate": 5.821528773303929e-06, + "loss": 1.5348, + "step": 1270 + }, + { + "epoch": 0.9506357516828721, + "grad_norm": 1.994935996790141, + "learning_rate": 5.815430357304664e-06, + "loss": 1.5327, + "step": 1271 + }, + { + "epoch": 0.9513836948391923, + "grad_norm": 3.4556415924964545, + "learning_rate": 5.809330694848586e-06, + "loss": 1.6683, + "step": 1272 + }, + { + "epoch": 0.9521316379955124, + "grad_norm": 1.87904484365436, + "learning_rate": 5.803229795259566e-06, + "loss": 1.7362, + "step": 1273 + }, + { + "epoch": 0.9528795811518325, + "grad_norm": 1.8315375110356675, + "learning_rate": 5.7971276678633625e-06, + "loss": 1.7175, + "step": 1274 + }, + { + "epoch": 0.9536275243081526, + "grad_norm": 2.827649353610511, + "learning_rate": 5.791024321987611e-06, + "loss": 1.3203, + "step": 1275 + }, + { + "epoch": 0.9543754674644727, + "grad_norm": 2.006686214422114, + "learning_rate": 5.784919766961813e-06, + "loss": 1.5022, + "step": 1276 + }, + { + "epoch": 0.9551234106207929, + "grad_norm": 1.9485076927691891, + "learning_rate": 5.778814012117315e-06, + "loss": 1.5291, + "step": 1277 + }, + { + "epoch": 0.955871353777113, + "grad_norm": 2.0996293300125473, + "learning_rate": 5.772707066787302e-06, + "loss": 1.7887, + "step": 1278 + }, + { + "epoch": 0.9566192969334331, + "grad_norm": 2.1085345304400938, + "learning_rate": 5.76659894030677e-06, + "loss": 1.6085, + "step": 1279 + }, + { + "epoch": 0.9573672400897532, + "grad_norm": 1.9377704742857376, + "learning_rate": 5.76048964201253e-06, + "loss": 1.5919, + "step": 1280 + }, + { + "epoch": 0.9581151832460733, + "grad_norm": 2.0530447045583675, + "learning_rate": 5.754379181243179e-06, + "loss": 1.4277, + "step": 1281 + }, + { + "epoch": 0.9588631264023935, + "grad_norm": 2.3024161061491752, + "learning_rate": 5.748267567339093e-06, + "loss": 1.6844, + "step": 1282 + }, + { + "epoch": 0.9596110695587136, + "grad_norm": 2.1151557066737148, + "learning_rate": 5.742154809642408e-06, + "loss": 1.8091, + "step": 1283 + }, + { + "epoch": 0.9603590127150337, + "grad_norm": 1.844802219134383, + "learning_rate": 5.7360409174970126e-06, + "loss": 1.5289, + "step": 1284 + }, + { + "epoch": 0.9611069558713538, + "grad_norm": 1.8053659021985409, + "learning_rate": 5.729925900248524e-06, + "loss": 1.7785, + "step": 1285 + }, + { + "epoch": 0.9618548990276738, + "grad_norm": 2.8481015541897543, + "learning_rate": 5.723809767244284e-06, + "loss": 1.6287, + "step": 1286 + }, + { + "epoch": 0.9626028421839941, + "grad_norm": 1.9208080375068948, + "learning_rate": 5.717692527833341e-06, + "loss": 1.5271, + "step": 1287 + }, + { + "epoch": 0.9633507853403142, + "grad_norm": 1.824310869518419, + "learning_rate": 5.711574191366427e-06, + "loss": 1.6444, + "step": 1288 + }, + { + "epoch": 0.9640987284966342, + "grad_norm": 1.893865953357531, + "learning_rate": 5.705454767195958e-06, + "loss": 1.6575, + "step": 1289 + }, + { + "epoch": 0.9648466716529543, + "grad_norm": 1.8797305149886658, + "learning_rate": 5.699334264676011e-06, + "loss": 1.6913, + "step": 1290 + }, + { + "epoch": 0.9655946148092744, + "grad_norm": 1.665311116178181, + "learning_rate": 5.693212693162311e-06, + "loss": 1.7326, + "step": 1291 + }, + { + "epoch": 0.9663425579655947, + "grad_norm": 2.439395430560639, + "learning_rate": 5.6870900620122175e-06, + "loss": 1.8005, + "step": 1292 + }, + { + "epoch": 0.9670905011219147, + "grad_norm": 1.6943420431132414, + "learning_rate": 5.680966380584706e-06, + "loss": 1.6632, + "step": 1293 + }, + { + "epoch": 0.9678384442782348, + "grad_norm": 1.8077726081989385, + "learning_rate": 5.674841658240365e-06, + "loss": 1.5696, + "step": 1294 + }, + { + "epoch": 0.9685863874345549, + "grad_norm": 1.9897526854949865, + "learning_rate": 5.668715904341365e-06, + "loss": 1.86, + "step": 1295 + }, + { + "epoch": 0.9693343305908751, + "grad_norm": 2.0170003104219987, + "learning_rate": 5.66258912825146e-06, + "loss": 1.664, + "step": 1296 + }, + { + "epoch": 0.9700822737471952, + "grad_norm": 1.903594718532567, + "learning_rate": 5.656461339335968e-06, + "loss": 1.7158, + "step": 1297 + }, + { + "epoch": 0.9708302169035153, + "grad_norm": 1.7746828611935463, + "learning_rate": 5.6503325469617475e-06, + "loss": 1.8211, + "step": 1298 + }, + { + "epoch": 0.9715781600598354, + "grad_norm": 2.0176920557122533, + "learning_rate": 5.644202760497195e-06, + "loss": 1.9683, + "step": 1299 + }, + { + "epoch": 0.9723261032161555, + "grad_norm": 1.7575633795860828, + "learning_rate": 5.638071989312228e-06, + "loss": 1.6023, + "step": 1300 + }, + { + "epoch": 0.9730740463724757, + "grad_norm": 1.8304265862834599, + "learning_rate": 5.631940242778271e-06, + "loss": 1.9663, + "step": 1301 + }, + { + "epoch": 0.9738219895287958, + "grad_norm": 1.6683343753344615, + "learning_rate": 5.62580753026823e-06, + "loss": 1.7246, + "step": 1302 + }, + { + "epoch": 0.9745699326851159, + "grad_norm": 2.227013331554352, + "learning_rate": 5.619673861156498e-06, + "loss": 1.5254, + "step": 1303 + }, + { + "epoch": 0.975317875841436, + "grad_norm": 2.959221690910029, + "learning_rate": 5.613539244818926e-06, + "loss": 1.6322, + "step": 1304 + }, + { + "epoch": 0.9760658189977561, + "grad_norm": 1.785477432981801, + "learning_rate": 5.607403690632814e-06, + "loss": 1.612, + "step": 1305 + }, + { + "epoch": 0.9768137621540763, + "grad_norm": 1.9029700921180666, + "learning_rate": 5.60126720797689e-06, + "loss": 1.3536, + "step": 1306 + }, + { + "epoch": 0.9775617053103964, + "grad_norm": 2.1901801780601353, + "learning_rate": 5.595129806231311e-06, + "loss": 1.5871, + "step": 1307 + }, + { + "epoch": 0.9783096484667165, + "grad_norm": 1.9146117932942126, + "learning_rate": 5.5889914947776305e-06, + "loss": 1.3907, + "step": 1308 + }, + { + "epoch": 0.9790575916230366, + "grad_norm": 1.8298949257739836, + "learning_rate": 5.5828522829987965e-06, + "loss": 1.7026, + "step": 1309 + }, + { + "epoch": 0.9798055347793567, + "grad_norm": 1.9354577601141738, + "learning_rate": 5.576712180279134e-06, + "loss": 1.7815, + "step": 1310 + }, + { + "epoch": 0.9805534779356769, + "grad_norm": 1.7464063814584072, + "learning_rate": 5.570571196004328e-06, + "loss": 1.5435, + "step": 1311 + }, + { + "epoch": 0.981301421091997, + "grad_norm": 2.0133817051989737, + "learning_rate": 5.564429339561411e-06, + "loss": 1.4926, + "step": 1312 + }, + { + "epoch": 0.9820493642483171, + "grad_norm": 2.623518653746122, + "learning_rate": 5.55828662033875e-06, + "loss": 1.5605, + "step": 1313 + }, + { + "epoch": 0.9827973074046372, + "grad_norm": 2.570414992904729, + "learning_rate": 5.552143047726029e-06, + "loss": 1.4657, + "step": 1314 + }, + { + "epoch": 0.9835452505609573, + "grad_norm": 2.0438002906315087, + "learning_rate": 5.5459986311142365e-06, + "loss": 1.4961, + "step": 1315 + }, + { + "epoch": 0.9842931937172775, + "grad_norm": 1.8265994866214148, + "learning_rate": 5.539853379895656e-06, + "loss": 1.6501, + "step": 1316 + }, + { + "epoch": 0.9850411368735976, + "grad_norm": 2.372716327863262, + "learning_rate": 5.533707303463839e-06, + "loss": 1.5643, + "step": 1317 + }, + { + "epoch": 0.9857890800299177, + "grad_norm": 2.1844711358608984, + "learning_rate": 5.527560411213604e-06, + "loss": 1.5939, + "step": 1318 + }, + { + "epoch": 0.9865370231862378, + "grad_norm": 1.7077162050690549, + "learning_rate": 5.521412712541013e-06, + "loss": 1.6955, + "step": 1319 + }, + { + "epoch": 0.9872849663425579, + "grad_norm": 1.5713706331832171, + "learning_rate": 5.515264216843368e-06, + "loss": 1.4561, + "step": 1320 + }, + { + "epoch": 0.9880329094988781, + "grad_norm": 1.9086973855414195, + "learning_rate": 5.509114933519179e-06, + "loss": 1.4126, + "step": 1321 + }, + { + "epoch": 0.9887808526551982, + "grad_norm": 1.85335661541272, + "learning_rate": 5.502964871968163e-06, + "loss": 1.7036, + "step": 1322 + }, + { + "epoch": 0.9895287958115183, + "grad_norm": 1.9392445060850925, + "learning_rate": 5.496814041591234e-06, + "loss": 1.7156, + "step": 1323 + }, + { + "epoch": 0.9902767389678384, + "grad_norm": 1.8171661478287164, + "learning_rate": 5.490662451790472e-06, + "loss": 1.7373, + "step": 1324 + }, + { + "epoch": 0.9910246821241586, + "grad_norm": 1.71994855257065, + "learning_rate": 5.484510111969123e-06, + "loss": 1.5162, + "step": 1325 + }, + { + "epoch": 0.9917726252804787, + "grad_norm": 1.8785573649606526, + "learning_rate": 5.478357031531578e-06, + "loss": 1.6852, + "step": 1326 + }, + { + "epoch": 0.9925205684367988, + "grad_norm": 1.710813904902256, + "learning_rate": 5.4722032198833595e-06, + "loss": 1.6805, + "step": 1327 + }, + { + "epoch": 0.9932685115931189, + "grad_norm": 2.1364429526712754, + "learning_rate": 5.466048686431108e-06, + "loss": 1.7355, + "step": 1328 + }, + { + "epoch": 0.994016454749439, + "grad_norm": 1.8544711117475958, + "learning_rate": 5.459893440582568e-06, + "loss": 1.4089, + "step": 1329 + }, + { + "epoch": 0.9947643979057592, + "grad_norm": 2.1234667135906515, + "learning_rate": 5.453737491746572e-06, + "loss": 1.5221, + "step": 1330 + }, + { + "epoch": 0.9955123410620793, + "grad_norm": 2.0761258506102753, + "learning_rate": 5.447580849333028e-06, + "loss": 1.7043, + "step": 1331 + }, + { + "epoch": 0.9962602842183994, + "grad_norm": 1.8940234887114151, + "learning_rate": 5.441423522752904e-06, + "loss": 1.5768, + "step": 1332 + }, + { + "epoch": 0.9970082273747195, + "grad_norm": 2.3683653117017935, + "learning_rate": 5.435265521418215e-06, + "loss": 1.7589, + "step": 1333 + }, + { + "epoch": 0.9977561705310396, + "grad_norm": 10.881755580036867, + "learning_rate": 5.429106854742006e-06, + "loss": 1.5523, + "step": 1334 + }, + { + "epoch": 0.9985041136873598, + "grad_norm": 2.150702498004197, + "learning_rate": 5.422947532138335e-06, + "loss": 1.8581, + "step": 1335 + }, + { + "epoch": 0.9992520568436799, + "grad_norm": 2.6462525915355792, + "learning_rate": 5.41678756302227e-06, + "loss": 1.8585, + "step": 1336 + }, + { + "epoch": 1.0, + "grad_norm": 1.9371562953996546, + "learning_rate": 5.410626956809864e-06, + "loss": 1.4023, + "step": 1337 + }, + { + "epoch": 1.0007479431563202, + "grad_norm": 1.7895147203610855, + "learning_rate": 5.404465722918144e-06, + "loss": 1.4735, + "step": 1338 + }, + { + "epoch": 1.0014958863126402, + "grad_norm": 1.7444228257903336, + "learning_rate": 5.3983038707650955e-06, + "loss": 1.4423, + "step": 1339 + }, + { + "epoch": 1.0022438294689604, + "grad_norm": 1.8105141661724167, + "learning_rate": 5.39214140976965e-06, + "loss": 1.6158, + "step": 1340 + }, + { + "epoch": 1.0022438294689604, + "eval_loss": 1.6649024486541748, + "eval_runtime": 68.8725, + "eval_samples_per_second": 59.138, + "eval_steps_per_second": 3.702, + "step": 1340 + }, + { + "epoch": 1.0029917726252804, + "grad_norm": 2.2886698188073713, + "learning_rate": 5.38597834935167e-06, + "loss": 1.4238, + "step": 1341 + }, + { + "epoch": 1.0037397157816006, + "grad_norm": 1.6481892624382062, + "learning_rate": 5.379814698931935e-06, + "loss": 1.535, + "step": 1342 + }, + { + "epoch": 1.0044876589379208, + "grad_norm": 1.8279815584841839, + "learning_rate": 5.373650467932122e-06, + "loss": 1.6601, + "step": 1343 + }, + { + "epoch": 1.0052356020942408, + "grad_norm": 1.6274735792022965, + "learning_rate": 5.367485665774802e-06, + "loss": 1.6253, + "step": 1344 + }, + { + "epoch": 1.005983545250561, + "grad_norm": 1.782911902914613, + "learning_rate": 5.361320301883413e-06, + "loss": 1.4637, + "step": 1345 + }, + { + "epoch": 1.006731488406881, + "grad_norm": 2.1354663750494165, + "learning_rate": 5.355154385682255e-06, + "loss": 1.4436, + "step": 1346 + }, + { + "epoch": 1.0074794315632012, + "grad_norm": 2.0483264389336306, + "learning_rate": 5.348987926596474e-06, + "loss": 1.4556, + "step": 1347 + }, + { + "epoch": 1.0082273747195214, + "grad_norm": 1.8960978530755148, + "learning_rate": 5.342820934052042e-06, + "loss": 1.5942, + "step": 1348 + }, + { + "epoch": 1.0089753178758414, + "grad_norm": 2.766754335669627, + "learning_rate": 5.336653417475745e-06, + "loss": 1.5182, + "step": 1349 + }, + { + "epoch": 1.0097232610321616, + "grad_norm": 1.6891323353057215, + "learning_rate": 5.3304853862951785e-06, + "loss": 1.6323, + "step": 1350 + }, + { + "epoch": 1.0104712041884816, + "grad_norm": 1.7792896419711364, + "learning_rate": 5.324316849938715e-06, + "loss": 1.4492, + "step": 1351 + }, + { + "epoch": 1.0112191473448018, + "grad_norm": 2.1796536225530856, + "learning_rate": 5.318147817835506e-06, + "loss": 1.4132, + "step": 1352 + }, + { + "epoch": 1.011967090501122, + "grad_norm": 2.422231835798018, + "learning_rate": 5.3119782994154556e-06, + "loss": 1.5633, + "step": 1353 + }, + { + "epoch": 1.012715033657442, + "grad_norm": 2.0523014557377985, + "learning_rate": 5.3058083041092145e-06, + "loss": 1.3268, + "step": 1354 + }, + { + "epoch": 1.0134629768137622, + "grad_norm": 1.965676106979729, + "learning_rate": 5.299637841348162e-06, + "loss": 1.471, + "step": 1355 + }, + { + "epoch": 1.0142109199700822, + "grad_norm": 1.9969999406787273, + "learning_rate": 5.29346692056439e-06, + "loss": 1.6326, + "step": 1356 + }, + { + "epoch": 1.0149588631264024, + "grad_norm": 2.010731701736663, + "learning_rate": 5.2872955511906974e-06, + "loss": 1.5258, + "step": 1357 + }, + { + "epoch": 1.0157068062827226, + "grad_norm": 1.9154057112931784, + "learning_rate": 5.281123742660558e-06, + "loss": 1.4405, + "step": 1358 + }, + { + "epoch": 1.0164547494390426, + "grad_norm": 2.073467934932341, + "learning_rate": 5.274951504408122e-06, + "loss": 1.5646, + "step": 1359 + }, + { + "epoch": 1.0172026925953628, + "grad_norm": 1.9499195692973341, + "learning_rate": 5.268778845868201e-06, + "loss": 1.4904, + "step": 1360 + }, + { + "epoch": 1.0179506357516828, + "grad_norm": 2.0099537394291462, + "learning_rate": 5.262605776476243e-06, + "loss": 1.7251, + "step": 1361 + }, + { + "epoch": 1.018698578908003, + "grad_norm": 2.529391196491734, + "learning_rate": 5.256432305668324e-06, + "loss": 1.4836, + "step": 1362 + }, + { + "epoch": 1.0194465220643232, + "grad_norm": 2.1525487280266615, + "learning_rate": 5.250258442881139e-06, + "loss": 1.8363, + "step": 1363 + }, + { + "epoch": 1.0201944652206432, + "grad_norm": 5.5037760298959695, + "learning_rate": 5.244084197551976e-06, + "loss": 1.7748, + "step": 1364 + }, + { + "epoch": 1.0209424083769634, + "grad_norm": 2.3583544958668923, + "learning_rate": 5.237909579118713e-06, + "loss": 1.6735, + "step": 1365 + }, + { + "epoch": 1.0216903515332834, + "grad_norm": 1.9935636198536903, + "learning_rate": 5.231734597019792e-06, + "loss": 1.6294, + "step": 1366 + }, + { + "epoch": 1.0224382946896036, + "grad_norm": 1.9255665178134589, + "learning_rate": 5.2255592606942185e-06, + "loss": 1.3412, + "step": 1367 + }, + { + "epoch": 1.0231862378459238, + "grad_norm": 2.003046550922734, + "learning_rate": 5.219383579581534e-06, + "loss": 1.5061, + "step": 1368 + }, + { + "epoch": 1.0239341810022438, + "grad_norm": 1.7750161011665737, + "learning_rate": 5.2132075631218075e-06, + "loss": 1.3871, + "step": 1369 + }, + { + "epoch": 1.024682124158564, + "grad_norm": 3.016259940467309, + "learning_rate": 5.2070312207556225e-06, + "loss": 1.6657, + "step": 1370 + }, + { + "epoch": 1.0254300673148842, + "grad_norm": 2.0465164509659397, + "learning_rate": 5.200854561924063e-06, + "loss": 1.4233, + "step": 1371 + }, + { + "epoch": 1.0261780104712042, + "grad_norm": 2.179764686442697, + "learning_rate": 5.194677596068689e-06, + "loss": 1.7838, + "step": 1372 + }, + { + "epoch": 1.0269259536275244, + "grad_norm": 2.2043918017332813, + "learning_rate": 5.188500332631538e-06, + "loss": 1.4991, + "step": 1373 + }, + { + "epoch": 1.0276738967838444, + "grad_norm": 2.0680982327920017, + "learning_rate": 5.182322781055097e-06, + "loss": 1.5326, + "step": 1374 + }, + { + "epoch": 1.0284218399401646, + "grad_norm": 2.0297460519715997, + "learning_rate": 5.176144950782296e-06, + "loss": 1.4831, + "step": 1375 + }, + { + "epoch": 1.0291697830964848, + "grad_norm": 2.3417618103243862, + "learning_rate": 5.169966851256489e-06, + "loss": 1.5077, + "step": 1376 + }, + { + "epoch": 1.0299177262528048, + "grad_norm": 1.9632739370358858, + "learning_rate": 5.163788491921447e-06, + "loss": 1.6901, + "step": 1377 + }, + { + "epoch": 1.030665669409125, + "grad_norm": 2.2369436465184065, + "learning_rate": 5.1576098822213285e-06, + "loss": 1.1958, + "step": 1378 + }, + { + "epoch": 1.031413612565445, + "grad_norm": 2.3498088138008404, + "learning_rate": 5.1514310316006835e-06, + "loss": 1.3139, + "step": 1379 + }, + { + "epoch": 1.0321615557217652, + "grad_norm": 2.34516234747464, + "learning_rate": 5.1452519495044265e-06, + "loss": 1.2756, + "step": 1380 + }, + { + "epoch": 1.0329094988780854, + "grad_norm": 2.0964261296398385, + "learning_rate": 5.139072645377829e-06, + "loss": 1.5209, + "step": 1381 + }, + { + "epoch": 1.0336574420344053, + "grad_norm": 2.1876892112585007, + "learning_rate": 5.132893128666493e-06, + "loss": 1.6294, + "step": 1382 + }, + { + "epoch": 1.0344053851907256, + "grad_norm": 2.129419438018858, + "learning_rate": 5.1267134088163576e-06, + "loss": 1.4689, + "step": 1383 + }, + { + "epoch": 1.0351533283470455, + "grad_norm": 1.9412743628909102, + "learning_rate": 5.120533495273664e-06, + "loss": 1.5706, + "step": 1384 + }, + { + "epoch": 1.0359012715033658, + "grad_norm": 1.9810070061638643, + "learning_rate": 5.1143533974849526e-06, + "loss": 1.6188, + "step": 1385 + }, + { + "epoch": 1.036649214659686, + "grad_norm": 2.047536193332397, + "learning_rate": 5.1081731248970435e-06, + "loss": 1.4631, + "step": 1386 + }, + { + "epoch": 1.037397157816006, + "grad_norm": 2.167307374287563, + "learning_rate": 5.101992686957028e-06, + "loss": 1.4962, + "step": 1387 + }, + { + "epoch": 1.0381451009723262, + "grad_norm": 2.544574448148535, + "learning_rate": 5.095812093112245e-06, + "loss": 1.6901, + "step": 1388 + }, + { + "epoch": 1.0388930441286461, + "grad_norm": 1.808559756934424, + "learning_rate": 5.089631352810275e-06, + "loss": 1.3961, + "step": 1389 + }, + { + "epoch": 1.0396409872849663, + "grad_norm": 1.9662334202645155, + "learning_rate": 5.0834504754989225e-06, + "loss": 1.4775, + "step": 1390 + }, + { + "epoch": 1.0403889304412866, + "grad_norm": 1.9396531797219332, + "learning_rate": 5.0772694706262e-06, + "loss": 1.343, + "step": 1391 + }, + { + "epoch": 1.0411368735976065, + "grad_norm": 2.168218296993762, + "learning_rate": 5.0710883476403135e-06, + "loss": 1.4643, + "step": 1392 + }, + { + "epoch": 1.0418848167539267, + "grad_norm": 2.2088239154857026, + "learning_rate": 5.064907115989655e-06, + "loss": 1.6689, + "step": 1393 + }, + { + "epoch": 1.0426327599102467, + "grad_norm": 2.0468706392282656, + "learning_rate": 5.058725785122777e-06, + "loss": 1.6093, + "step": 1394 + }, + { + "epoch": 1.043380703066567, + "grad_norm": 2.099960943016367, + "learning_rate": 5.052544364488386e-06, + "loss": 1.555, + "step": 1395 + }, + { + "epoch": 1.0441286462228871, + "grad_norm": 2.3073065478970287, + "learning_rate": 5.046362863535327e-06, + "loss": 1.3827, + "step": 1396 + }, + { + "epoch": 1.0448765893792071, + "grad_norm": 2.094595398551958, + "learning_rate": 5.040181291712564e-06, + "loss": 1.5493, + "step": 1397 + }, + { + "epoch": 1.0456245325355273, + "grad_norm": 2.1709809205684376, + "learning_rate": 5.033999658469174e-06, + "loss": 1.3733, + "step": 1398 + }, + { + "epoch": 1.0463724756918473, + "grad_norm": 2.1343582168110617, + "learning_rate": 5.0278179732543244e-06, + "loss": 1.4405, + "step": 1399 + }, + { + "epoch": 1.0471204188481675, + "grad_norm": 1.9402354986830461, + "learning_rate": 5.021636245517261e-06, + "loss": 1.5371, + "step": 1400 + }, + { + "epoch": 1.0478683620044877, + "grad_norm": 1.8774044108683747, + "learning_rate": 5.0154544847073015e-06, + "loss": 1.5712, + "step": 1401 + }, + { + "epoch": 1.0486163051608077, + "grad_norm": 1.953107661456675, + "learning_rate": 5.009272700273804e-06, + "loss": 1.6615, + "step": 1402 + }, + { + "epoch": 1.049364248317128, + "grad_norm": 1.9294999929686356, + "learning_rate": 5.003090901666173e-06, + "loss": 1.26, + "step": 1403 + }, + { + "epoch": 1.050112191473448, + "grad_norm": 2.247582582679804, + "learning_rate": 4.9969090983338285e-06, + "loss": 1.457, + "step": 1404 + }, + { + "epoch": 1.0508601346297681, + "grad_norm": 2.000840273808582, + "learning_rate": 4.990727299726196e-06, + "loss": 1.6852, + "step": 1405 + }, + { + "epoch": 1.0516080777860883, + "grad_norm": 2.1282229808104516, + "learning_rate": 4.9845455152927e-06, + "loss": 1.5689, + "step": 1406 + }, + { + "epoch": 1.0523560209424083, + "grad_norm": 1.87148607155346, + "learning_rate": 4.978363754482741e-06, + "loss": 1.2106, + "step": 1407 + }, + { + "epoch": 1.0531039640987285, + "grad_norm": 7.377724394221026, + "learning_rate": 4.972182026745678e-06, + "loss": 1.6909, + "step": 1408 + }, + { + "epoch": 1.0538519072550485, + "grad_norm": 2.333474665036998, + "learning_rate": 4.966000341530827e-06, + "loss": 1.7126, + "step": 1409 + }, + { + "epoch": 1.0545998504113687, + "grad_norm": 2.127847690743402, + "learning_rate": 4.959818708287436e-06, + "loss": 1.4577, + "step": 1410 + }, + { + "epoch": 1.055347793567689, + "grad_norm": 2.1341512393109165, + "learning_rate": 4.953637136464675e-06, + "loss": 1.6661, + "step": 1411 + }, + { + "epoch": 1.056095736724009, + "grad_norm": 1.9009111529368734, + "learning_rate": 4.947455635511615e-06, + "loss": 1.5207, + "step": 1412 + }, + { + "epoch": 1.0568436798803291, + "grad_norm": 2.159110612524291, + "learning_rate": 4.941274214877223e-06, + "loss": 1.5231, + "step": 1413 + }, + { + "epoch": 1.057591623036649, + "grad_norm": 2.157969439868393, + "learning_rate": 4.935092884010347e-06, + "loss": 1.2835, + "step": 1414 + }, + { + "epoch": 1.0583395661929693, + "grad_norm": 2.0989871966604303, + "learning_rate": 4.928911652359687e-06, + "loss": 1.3975, + "step": 1415 + }, + { + "epoch": 1.0590875093492895, + "grad_norm": 2.031182291549457, + "learning_rate": 4.922730529373801e-06, + "loss": 1.4512, + "step": 1416 + }, + { + "epoch": 1.0598354525056095, + "grad_norm": 2.479292737433032, + "learning_rate": 4.91654952450108e-06, + "loss": 1.4876, + "step": 1417 + }, + { + "epoch": 1.0605833956619297, + "grad_norm": 1.9826374986161348, + "learning_rate": 4.910368647189726e-06, + "loss": 1.5574, + "step": 1418 + }, + { + "epoch": 1.06133133881825, + "grad_norm": 2.1765678987387265, + "learning_rate": 4.904187906887756e-06, + "loss": 1.1496, + "step": 1419 + }, + { + "epoch": 1.06207928197457, + "grad_norm": 1.8271775867616562, + "learning_rate": 4.898007313042975e-06, + "loss": 1.6526, + "step": 1420 + }, + { + "epoch": 1.0628272251308901, + "grad_norm": 2.0819663545887725, + "learning_rate": 4.891826875102958e-06, + "loss": 1.5681, + "step": 1421 + }, + { + "epoch": 1.06357516828721, + "grad_norm": 3.464318132348866, + "learning_rate": 4.885646602515049e-06, + "loss": 1.4775, + "step": 1422 + }, + { + "epoch": 1.0643231114435303, + "grad_norm": 2.007094511387975, + "learning_rate": 4.879466504726337e-06, + "loss": 1.6528, + "step": 1423 + }, + { + "epoch": 1.0650710545998505, + "grad_norm": 1.8638953185046747, + "learning_rate": 4.873286591183644e-06, + "loss": 1.439, + "step": 1424 + }, + { + "epoch": 1.0658189977561705, + "grad_norm": 2.5832132115694333, + "learning_rate": 4.867106871333509e-06, + "loss": 1.4786, + "step": 1425 + }, + { + "epoch": 1.0665669409124907, + "grad_norm": 1.9392902583081104, + "learning_rate": 4.860927354622173e-06, + "loss": 1.4166, + "step": 1426 + }, + { + "epoch": 1.0673148840688107, + "grad_norm": 2.2142568702376284, + "learning_rate": 4.854748050495574e-06, + "loss": 1.4133, + "step": 1427 + }, + { + "epoch": 1.068062827225131, + "grad_norm": 1.9918008801583895, + "learning_rate": 4.848568968399317e-06, + "loss": 1.5782, + "step": 1428 + }, + { + "epoch": 1.0688107703814511, + "grad_norm": 2.4377877415101152, + "learning_rate": 4.842390117778673e-06, + "loss": 1.4966, + "step": 1429 + }, + { + "epoch": 1.069558713537771, + "grad_norm": 1.9681040117398607, + "learning_rate": 4.836211508078557e-06, + "loss": 1.799, + "step": 1430 + }, + { + "epoch": 1.0703066566940913, + "grad_norm": 2.1959191095967103, + "learning_rate": 4.830033148743512e-06, + "loss": 1.4518, + "step": 1431 + }, + { + "epoch": 1.0710545998504113, + "grad_norm": 2.017981804684803, + "learning_rate": 4.8238550492177065e-06, + "loss": 1.5417, + "step": 1432 + }, + { + "epoch": 1.0718025430067315, + "grad_norm": 2.511008043986542, + "learning_rate": 4.817677218944904e-06, + "loss": 1.1791, + "step": 1433 + }, + { + "epoch": 1.0725504861630517, + "grad_norm": 2.245173314732633, + "learning_rate": 4.811499667368464e-06, + "loss": 1.4282, + "step": 1434 + }, + { + "epoch": 1.0732984293193717, + "grad_norm": 2.210376557023909, + "learning_rate": 4.805322403931312e-06, + "loss": 1.421, + "step": 1435 + }, + { + "epoch": 1.074046372475692, + "grad_norm": 1.9837353541263227, + "learning_rate": 4.7991454380759376e-06, + "loss": 1.5148, + "step": 1436 + }, + { + "epoch": 1.0747943156320119, + "grad_norm": 1.9324067524193247, + "learning_rate": 4.792968779244378e-06, + "loss": 1.4424, + "step": 1437 + }, + { + "epoch": 1.075542258788332, + "grad_norm": 2.19665741146211, + "learning_rate": 4.786792436878193e-06, + "loss": 1.1855, + "step": 1438 + }, + { + "epoch": 1.0762902019446523, + "grad_norm": 2.124713453402527, + "learning_rate": 4.780616420418468e-06, + "loss": 1.3293, + "step": 1439 + }, + { + "epoch": 1.0770381451009723, + "grad_norm": 2.56837278572991, + "learning_rate": 4.774440739305784e-06, + "loss": 1.4438, + "step": 1440 + }, + { + "epoch": 1.0777860882572925, + "grad_norm": 2.2113389130137886, + "learning_rate": 4.768265402980209e-06, + "loss": 1.393, + "step": 1441 + }, + { + "epoch": 1.0785340314136125, + "grad_norm": 1.959037431661476, + "learning_rate": 4.762090420881289e-06, + "loss": 1.7422, + "step": 1442 + }, + { + "epoch": 1.0792819745699327, + "grad_norm": 1.9089530008276454, + "learning_rate": 4.755915802448024e-06, + "loss": 1.523, + "step": 1443 + }, + { + "epoch": 1.080029917726253, + "grad_norm": 2.0153749087838086, + "learning_rate": 4.749741557118863e-06, + "loss": 1.5976, + "step": 1444 + }, + { + "epoch": 1.0807778608825729, + "grad_norm": 1.9743706305522764, + "learning_rate": 4.743567694331677e-06, + "loss": 1.5913, + "step": 1445 + }, + { + "epoch": 1.081525804038893, + "grad_norm": 1.9292280905952306, + "learning_rate": 4.737394223523758e-06, + "loss": 1.3394, + "step": 1446 + }, + { + "epoch": 1.082273747195213, + "grad_norm": 2.671523713493291, + "learning_rate": 4.7312211541318e-06, + "loss": 1.2654, + "step": 1447 + }, + { + "epoch": 1.0830216903515333, + "grad_norm": 2.21306649352015, + "learning_rate": 4.7250484955918794e-06, + "loss": 1.4091, + "step": 1448 + }, + { + "epoch": 1.0837696335078535, + "grad_norm": 1.8392186697123603, + "learning_rate": 4.718876257339444e-06, + "loss": 1.6268, + "step": 1449 + }, + { + "epoch": 1.0845175766641735, + "grad_norm": 2.050253436303179, + "learning_rate": 4.712704448809305e-06, + "loss": 1.5923, + "step": 1450 + }, + { + "epoch": 1.0852655198204937, + "grad_norm": 1.8744182903103137, + "learning_rate": 4.7065330794356096e-06, + "loss": 1.5946, + "step": 1451 + }, + { + "epoch": 1.0860134629768137, + "grad_norm": 2.769207113593209, + "learning_rate": 4.7003621586518385e-06, + "loss": 1.3816, + "step": 1452 + }, + { + "epoch": 1.0867614061331339, + "grad_norm": 2.1033938868296094, + "learning_rate": 4.694191695890788e-06, + "loss": 1.6422, + "step": 1453 + }, + { + "epoch": 1.087509349289454, + "grad_norm": 2.066236651331623, + "learning_rate": 4.688021700584546e-06, + "loss": 1.4178, + "step": 1454 + }, + { + "epoch": 1.088257292445774, + "grad_norm": 1.9102083099521219, + "learning_rate": 4.681852182164496e-06, + "loss": 1.3842, + "step": 1455 + }, + { + "epoch": 1.0890052356020943, + "grad_norm": 1.965116890698696, + "learning_rate": 4.6756831500612846e-06, + "loss": 1.7723, + "step": 1456 + }, + { + "epoch": 1.0897531787584143, + "grad_norm": 2.3804238525982866, + "learning_rate": 4.669514613704823e-06, + "loss": 1.4637, + "step": 1457 + }, + { + "epoch": 1.0905011219147345, + "grad_norm": 2.0364771812661733, + "learning_rate": 4.6633465825242555e-06, + "loss": 1.6858, + "step": 1458 + }, + { + "epoch": 1.0912490650710547, + "grad_norm": 2.2537987795043137, + "learning_rate": 4.657179065947959e-06, + "loss": 1.665, + "step": 1459 + }, + { + "epoch": 1.0919970082273747, + "grad_norm": 2.269613329435273, + "learning_rate": 4.651012073403528e-06, + "loss": 1.6534, + "step": 1460 + }, + { + "epoch": 1.0927449513836949, + "grad_norm": 2.1363203968089217, + "learning_rate": 4.644845614317745e-06, + "loss": 1.2467, + "step": 1461 + }, + { + "epoch": 1.0934928945400149, + "grad_norm": 2.1638349435495092, + "learning_rate": 4.638679698116588e-06, + "loss": 1.5005, + "step": 1462 + }, + { + "epoch": 1.094240837696335, + "grad_norm": 2.012537448899041, + "learning_rate": 4.632514334225201e-06, + "loss": 1.5948, + "step": 1463 + }, + { + "epoch": 1.0949887808526553, + "grad_norm": 2.2477658751629943, + "learning_rate": 4.626349532067879e-06, + "loss": 1.3474, + "step": 1464 + }, + { + "epoch": 1.0957367240089753, + "grad_norm": 2.0052991296335994, + "learning_rate": 4.620185301068067e-06, + "loss": 1.6051, + "step": 1465 + }, + { + "epoch": 1.0964846671652955, + "grad_norm": 2.535961086415829, + "learning_rate": 4.614021650648329e-06, + "loss": 1.2952, + "step": 1466 + }, + { + "epoch": 1.0972326103216155, + "grad_norm": 2.4256631520897765, + "learning_rate": 4.607858590230351e-06, + "loss": 1.3475, + "step": 1467 + }, + { + "epoch": 1.0979805534779357, + "grad_norm": 1.9105362042679925, + "learning_rate": 4.601696129234907e-06, + "loss": 1.2979, + "step": 1468 + }, + { + "epoch": 1.0987284966342559, + "grad_norm": 2.217809249918699, + "learning_rate": 4.5955342770818564e-06, + "loss": 1.6951, + "step": 1469 + }, + { + "epoch": 1.0994764397905759, + "grad_norm": 2.0334651579935263, + "learning_rate": 4.589373043190137e-06, + "loss": 1.3832, + "step": 1470 + }, + { + "epoch": 1.100224382946896, + "grad_norm": 2.1974916630251355, + "learning_rate": 4.583212436977731e-06, + "loss": 1.5927, + "step": 1471 + }, + { + "epoch": 1.100972326103216, + "grad_norm": 2.318199575210842, + "learning_rate": 4.5770524678616665e-06, + "loss": 1.7325, + "step": 1472 + }, + { + "epoch": 1.1017202692595363, + "grad_norm": 1.9697694785823903, + "learning_rate": 4.5708931452579975e-06, + "loss": 1.5967, + "step": 1473 + }, + { + "epoch": 1.1024682124158565, + "grad_norm": 2.0030328815657192, + "learning_rate": 4.564734478581785e-06, + "loss": 1.4144, + "step": 1474 + }, + { + "epoch": 1.1032161555721764, + "grad_norm": 2.110149380721777, + "learning_rate": 4.558576477247097e-06, + "loss": 1.5848, + "step": 1475 + }, + { + "epoch": 1.1039640987284967, + "grad_norm": 1.8096255647776327, + "learning_rate": 4.552419150666973e-06, + "loss": 1.2586, + "step": 1476 + }, + { + "epoch": 1.1047120418848166, + "grad_norm": 2.2240344519250983, + "learning_rate": 4.546262508253429e-06, + "loss": 1.3125, + "step": 1477 + }, + { + "epoch": 1.1054599850411369, + "grad_norm": 2.176211037292367, + "learning_rate": 4.5401065594174345e-06, + "loss": 1.6971, + "step": 1478 + }, + { + "epoch": 1.106207928197457, + "grad_norm": 1.9395763972868005, + "learning_rate": 4.533951313568894e-06, + "loss": 1.5907, + "step": 1479 + }, + { + "epoch": 1.106955871353777, + "grad_norm": 1.9143791160079537, + "learning_rate": 4.527796780116642e-06, + "loss": 1.447, + "step": 1480 + }, + { + "epoch": 1.1077038145100973, + "grad_norm": 2.017244790836672, + "learning_rate": 4.5216429684684245e-06, + "loss": 1.8326, + "step": 1481 + }, + { + "epoch": 1.1084517576664172, + "grad_norm": 1.8042992745519646, + "learning_rate": 4.515489888030878e-06, + "loss": 1.5779, + "step": 1482 + }, + { + "epoch": 1.1091997008227374, + "grad_norm": 2.062340574207803, + "learning_rate": 4.509337548209529e-06, + "loss": 1.5006, + "step": 1483 + }, + { + "epoch": 1.1099476439790577, + "grad_norm": 3.1691210234513663, + "learning_rate": 4.503185958408767e-06, + "loss": 1.2027, + "step": 1484 + }, + { + "epoch": 1.1106955871353776, + "grad_norm": 2.5903158199665737, + "learning_rate": 4.497035128031838e-06, + "loss": 1.6404, + "step": 1485 + }, + { + "epoch": 1.1114435302916978, + "grad_norm": 2.5889138691280293, + "learning_rate": 4.4908850664808245e-06, + "loss": 1.5647, + "step": 1486 + }, + { + "epoch": 1.112191473448018, + "grad_norm": 2.191558332219471, + "learning_rate": 4.484735783156634e-06, + "loss": 1.6952, + "step": 1487 + }, + { + "epoch": 1.112939416604338, + "grad_norm": 2.268185972910703, + "learning_rate": 4.478587287458988e-06, + "loss": 1.4529, + "step": 1488 + }, + { + "epoch": 1.1136873597606582, + "grad_norm": 2.703782126166686, + "learning_rate": 4.472439588786398e-06, + "loss": 1.5174, + "step": 1489 + }, + { + "epoch": 1.1144353029169782, + "grad_norm": 2.5314048278183394, + "learning_rate": 4.466292696536162e-06, + "loss": 1.5108, + "step": 1490 + }, + { + "epoch": 1.1151832460732984, + "grad_norm": 3.2592784881117405, + "learning_rate": 4.460146620104347e-06, + "loss": 1.5948, + "step": 1491 + }, + { + "epoch": 1.1159311892296186, + "grad_norm": 1.9956854761152982, + "learning_rate": 4.454001368885764e-06, + "loss": 1.5602, + "step": 1492 + }, + { + "epoch": 1.1166791323859386, + "grad_norm": 2.6425158607273285, + "learning_rate": 4.4478569522739736e-06, + "loss": 1.6493, + "step": 1493 + }, + { + "epoch": 1.1174270755422588, + "grad_norm": 2.2136878095084485, + "learning_rate": 4.441713379661251e-06, + "loss": 1.4258, + "step": 1494 + }, + { + "epoch": 1.1181750186985788, + "grad_norm": 2.407142995084283, + "learning_rate": 4.4355706604385905e-06, + "loss": 1.5202, + "step": 1495 + }, + { + "epoch": 1.118922961854899, + "grad_norm": 2.1880304951153233, + "learning_rate": 4.4294288039956735e-06, + "loss": 1.6106, + "step": 1496 + }, + { + "epoch": 1.1196709050112192, + "grad_norm": 2.2845841065034027, + "learning_rate": 4.423287819720866e-06, + "loss": 1.6691, + "step": 1497 + }, + { + "epoch": 1.1204188481675392, + "grad_norm": 2.763616267550361, + "learning_rate": 4.417147717001205e-06, + "loss": 1.2191, + "step": 1498 + }, + { + "epoch": 1.1211667913238594, + "grad_norm": 2.291298607453395, + "learning_rate": 4.411008505222371e-06, + "loss": 1.522, + "step": 1499 + }, + { + "epoch": 1.1219147344801794, + "grad_norm": 2.065430533705428, + "learning_rate": 4.4048701937686914e-06, + "loss": 1.3739, + "step": 1500 + }, + { + "epoch": 1.1226626776364996, + "grad_norm": 1.9898494501673747, + "learning_rate": 4.3987327920231124e-06, + "loss": 1.6727, + "step": 1501 + }, + { + "epoch": 1.1234106207928198, + "grad_norm": 2.1661625642282463, + "learning_rate": 4.392596309367189e-06, + "loss": 1.3519, + "step": 1502 + }, + { + "epoch": 1.1241585639491398, + "grad_norm": 2.232726167063726, + "learning_rate": 4.386460755181075e-06, + "loss": 1.5118, + "step": 1503 + }, + { + "epoch": 1.12490650710546, + "grad_norm": 1.9868034805316215, + "learning_rate": 4.380326138843503e-06, + "loss": 1.5585, + "step": 1504 + }, + { + "epoch": 1.12565445026178, + "grad_norm": 2.5653184123641943, + "learning_rate": 4.374192469731771e-06, + "loss": 1.3333, + "step": 1505 + }, + { + "epoch": 1.1264023934181002, + "grad_norm": 2.7898004731368777, + "learning_rate": 4.368059757221731e-06, + "loss": 1.5277, + "step": 1506 + }, + { + "epoch": 1.1271503365744204, + "grad_norm": 2.410572491964202, + "learning_rate": 4.3619280106877716e-06, + "loss": 1.6597, + "step": 1507 + }, + { + "epoch": 1.1278982797307404, + "grad_norm": 2.45864540706677, + "learning_rate": 4.355797239502807e-06, + "loss": 1.5283, + "step": 1508 + }, + { + "epoch": 1.1286462228870606, + "grad_norm": 2.539858519729722, + "learning_rate": 4.349667453038256e-06, + "loss": 1.4461, + "step": 1509 + }, + { + "epoch": 1.1293941660433806, + "grad_norm": 2.1795005895819752, + "learning_rate": 4.343538660664033e-06, + "loss": 1.4246, + "step": 1510 + }, + { + "epoch": 1.1301421091997008, + "grad_norm": 2.1405642095772044, + "learning_rate": 4.337410871748541e-06, + "loss": 1.383, + "step": 1511 + }, + { + "epoch": 1.130890052356021, + "grad_norm": 2.8952983379135717, + "learning_rate": 4.331284095658637e-06, + "loss": 1.5505, + "step": 1512 + }, + { + "epoch": 1.131637995512341, + "grad_norm": 2.084119688479437, + "learning_rate": 4.325158341759638e-06, + "loss": 1.7146, + "step": 1513 + }, + { + "epoch": 1.1323859386686612, + "grad_norm": 2.1375304885651505, + "learning_rate": 4.319033619415296e-06, + "loss": 1.4788, + "step": 1514 + }, + { + "epoch": 1.1331338818249814, + "grad_norm": 2.6522877524256288, + "learning_rate": 4.312909937987785e-06, + "loss": 1.4799, + "step": 1515 + }, + { + "epoch": 1.1338818249813014, + "grad_norm": 2.0757306860098494, + "learning_rate": 4.30678730683769e-06, + "loss": 1.4625, + "step": 1516 + }, + { + "epoch": 1.1346297681376216, + "grad_norm": 2.252045910921994, + "learning_rate": 4.3006657353239885e-06, + "loss": 1.512, + "step": 1517 + }, + { + "epoch": 1.1353777112939416, + "grad_norm": 2.229698298664332, + "learning_rate": 4.294545232804043e-06, + "loss": 1.5643, + "step": 1518 + }, + { + "epoch": 1.1361256544502618, + "grad_norm": 2.161090881196785, + "learning_rate": 4.2884258086335755e-06, + "loss": 1.4964, + "step": 1519 + }, + { + "epoch": 1.136873597606582, + "grad_norm": 2.0274260291223696, + "learning_rate": 4.282307472166661e-06, + "loss": 1.3519, + "step": 1520 + }, + { + "epoch": 1.137621540762902, + "grad_norm": 1.8743634609916981, + "learning_rate": 4.276190232755718e-06, + "loss": 1.4859, + "step": 1521 + }, + { + "epoch": 1.1383694839192222, + "grad_norm": 2.7180809185541963, + "learning_rate": 4.270074099751478e-06, + "loss": 1.4018, + "step": 1522 + }, + { + "epoch": 1.1391174270755422, + "grad_norm": 2.523167130232032, + "learning_rate": 4.26395908250299e-06, + "loss": 1.6766, + "step": 1523 + }, + { + "epoch": 1.1398653702318624, + "grad_norm": 2.0305309106669376, + "learning_rate": 4.257845190357594e-06, + "loss": 1.6201, + "step": 1524 + }, + { + "epoch": 1.1406133133881826, + "grad_norm": 2.1748610689297303, + "learning_rate": 4.251732432660909e-06, + "loss": 1.5114, + "step": 1525 + }, + { + "epoch": 1.1413612565445026, + "grad_norm": 1.9243388720723411, + "learning_rate": 4.245620818756822e-06, + "loss": 1.5394, + "step": 1526 + }, + { + "epoch": 1.1421091997008228, + "grad_norm": 1.953965374012726, + "learning_rate": 4.239510357987469e-06, + "loss": 1.5007, + "step": 1527 + }, + { + "epoch": 1.1428571428571428, + "grad_norm": 2.160060003535349, + "learning_rate": 4.233401059693231e-06, + "loss": 1.6414, + "step": 1528 + }, + { + "epoch": 1.143605086013463, + "grad_norm": 2.394534337337246, + "learning_rate": 4.2272929332127e-06, + "loss": 1.4952, + "step": 1529 + }, + { + "epoch": 1.1443530291697832, + "grad_norm": 2.19355028863127, + "learning_rate": 4.221185987882684e-06, + "loss": 1.6688, + "step": 1530 + }, + { + "epoch": 1.1451009723261032, + "grad_norm": 2.055385113781556, + "learning_rate": 4.215080233038189e-06, + "loss": 1.2875, + "step": 1531 + }, + { + "epoch": 1.1458489154824234, + "grad_norm": 2.006330785704172, + "learning_rate": 4.208975678012391e-06, + "loss": 1.34, + "step": 1532 + }, + { + "epoch": 1.1465968586387434, + "grad_norm": 2.3121452205032487, + "learning_rate": 4.202872332136639e-06, + "loss": 1.1892, + "step": 1533 + }, + { + "epoch": 1.1473448017950636, + "grad_norm": 2.06166535753739, + "learning_rate": 4.1967702047404365e-06, + "loss": 1.7356, + "step": 1534 + }, + { + "epoch": 1.1480927449513838, + "grad_norm": 2.3025456323517743, + "learning_rate": 4.190669305151414e-06, + "loss": 1.5487, + "step": 1535 + }, + { + "epoch": 1.1488406881077038, + "grad_norm": 2.1702405796044846, + "learning_rate": 4.184569642695337e-06, + "loss": 1.3831, + "step": 1536 + }, + { + "epoch": 1.149588631264024, + "grad_norm": 2.5767409003776094, + "learning_rate": 4.178471226696073e-06, + "loss": 1.5361, + "step": 1537 + }, + { + "epoch": 1.150336574420344, + "grad_norm": 1.9144493859411555, + "learning_rate": 4.172374066475583e-06, + "loss": 1.5644, + "step": 1538 + }, + { + "epoch": 1.1510845175766642, + "grad_norm": 2.1809926306260614, + "learning_rate": 4.166278171353915e-06, + "loss": 1.2065, + "step": 1539 + }, + { + "epoch": 1.1518324607329844, + "grad_norm": 1.9497107354816496, + "learning_rate": 4.160183550649176e-06, + "loss": 1.341, + "step": 1540 + }, + { + "epoch": 1.1525804038893044, + "grad_norm": 2.1518436130194067, + "learning_rate": 4.154090213677531e-06, + "loss": 1.8712, + "step": 1541 + }, + { + "epoch": 1.1533283470456246, + "grad_norm": 2.2861154717844667, + "learning_rate": 4.147998169753179e-06, + "loss": 1.4301, + "step": 1542 + }, + { + "epoch": 1.1540762902019446, + "grad_norm": 2.086009711553455, + "learning_rate": 4.141907428188341e-06, + "loss": 1.3329, + "step": 1543 + }, + { + "epoch": 1.1548242333582648, + "grad_norm": 3.0271204290362426, + "learning_rate": 4.1358179982932535e-06, + "loss": 1.7777, + "step": 1544 + }, + { + "epoch": 1.155572176514585, + "grad_norm": 2.140900079446597, + "learning_rate": 4.129729889376141e-06, + "loss": 1.5402, + "step": 1545 + }, + { + "epoch": 1.156320119670905, + "grad_norm": 2.2311808944857483, + "learning_rate": 4.123643110743214e-06, + "loss": 1.3495, + "step": 1546 + }, + { + "epoch": 1.1570680628272252, + "grad_norm": 1.9782341621439907, + "learning_rate": 4.117557671698648e-06, + "loss": 1.3967, + "step": 1547 + }, + { + "epoch": 1.1578160059835452, + "grad_norm": 2.1835309162942607, + "learning_rate": 4.111473581544566e-06, + "loss": 1.4601, + "step": 1548 + }, + { + "epoch": 1.1585639491398654, + "grad_norm": 2.1469227802214994, + "learning_rate": 4.105390849581037e-06, + "loss": 1.5723, + "step": 1549 + }, + { + "epoch": 1.1593118922961856, + "grad_norm": 2.0522406526648544, + "learning_rate": 4.09930948510605e-06, + "loss": 1.5741, + "step": 1550 + }, + { + "epoch": 1.1600598354525056, + "grad_norm": 2.2430470340509587, + "learning_rate": 4.093229497415504e-06, + "loss": 1.593, + "step": 1551 + }, + { + "epoch": 1.1608077786088258, + "grad_norm": 2.049268441362586, + "learning_rate": 4.087150895803192e-06, + "loss": 1.7058, + "step": 1552 + }, + { + "epoch": 1.1615557217651458, + "grad_norm": 2.6488530781000095, + "learning_rate": 4.08107368956079e-06, + "loss": 1.5274, + "step": 1553 + }, + { + "epoch": 1.162303664921466, + "grad_norm": 2.0364491123777713, + "learning_rate": 4.074997887977843e-06, + "loss": 1.2193, + "step": 1554 + }, + { + "epoch": 1.1630516080777862, + "grad_norm": 2.0852308379251125, + "learning_rate": 4.0689235003417425e-06, + "loss": 1.5229, + "step": 1555 + }, + { + "epoch": 1.1637995512341062, + "grad_norm": 1.9173915167729811, + "learning_rate": 4.062850535937726e-06, + "loss": 1.6183, + "step": 1556 + }, + { + "epoch": 1.1645474943904264, + "grad_norm": 2.151189531329751, + "learning_rate": 4.056779004048852e-06, + "loss": 1.5553, + "step": 1557 + }, + { + "epoch": 1.1652954375467464, + "grad_norm": 1.8760517734871536, + "learning_rate": 4.050708913955987e-06, + "loss": 1.5314, + "step": 1558 + }, + { + "epoch": 1.1660433807030666, + "grad_norm": 2.015980255145151, + "learning_rate": 4.044640274937798e-06, + "loss": 1.3807, + "step": 1559 + }, + { + "epoch": 1.1667913238593868, + "grad_norm": 1.9898444584685144, + "learning_rate": 4.0385730962707325e-06, + "loss": 1.523, + "step": 1560 + }, + { + "epoch": 1.1675392670157068, + "grad_norm": 1.9825159165083037, + "learning_rate": 4.032507387229002e-06, + "loss": 1.8422, + "step": 1561 + }, + { + "epoch": 1.168287210172027, + "grad_norm": 1.954405109816679, + "learning_rate": 4.026443157084579e-06, + "loss": 1.5926, + "step": 1562 + }, + { + "epoch": 1.169035153328347, + "grad_norm": 2.4257785232832556, + "learning_rate": 4.020380415107167e-06, + "loss": 1.7091, + "step": 1563 + }, + { + "epoch": 1.1697830964846672, + "grad_norm": 2.0653196623678354, + "learning_rate": 4.014319170564202e-06, + "loss": 1.3885, + "step": 1564 + }, + { + "epoch": 1.1705310396409874, + "grad_norm": 2.1576171841293763, + "learning_rate": 4.008259432720827e-06, + "loss": 1.3862, + "step": 1565 + }, + { + "epoch": 1.1712789827973074, + "grad_norm": 2.5550077823612853, + "learning_rate": 4.002201210839881e-06, + "loss": 1.8935, + "step": 1566 + }, + { + "epoch": 1.1720269259536276, + "grad_norm": 1.8788494737714097, + "learning_rate": 3.996144514181891e-06, + "loss": 1.4649, + "step": 1567 + }, + { + "epoch": 1.1727748691099475, + "grad_norm": 2.0625235856403714, + "learning_rate": 3.9900893520050446e-06, + "loss": 1.5912, + "step": 1568 + }, + { + "epoch": 1.1735228122662678, + "grad_norm": 2.06875855664409, + "learning_rate": 3.9840357335651914e-06, + "loss": 1.577, + "step": 1569 + }, + { + "epoch": 1.174270755422588, + "grad_norm": 1.919026945503743, + "learning_rate": 3.977983668115818e-06, + "loss": 1.7289, + "step": 1570 + }, + { + "epoch": 1.175018698578908, + "grad_norm": 1.9338547590296657, + "learning_rate": 3.971933164908034e-06, + "loss": 1.3084, + "step": 1571 + }, + { + "epoch": 1.1757666417352282, + "grad_norm": 2.6488775314296698, + "learning_rate": 3.9658842331905675e-06, + "loss": 1.5185, + "step": 1572 + }, + { + "epoch": 1.1765145848915481, + "grad_norm": 1.8955734053282778, + "learning_rate": 3.959836882209739e-06, + "loss": 1.7371, + "step": 1573 + }, + { + "epoch": 1.1772625280478684, + "grad_norm": 1.9982578782816407, + "learning_rate": 3.953791121209458e-06, + "loss": 1.5591, + "step": 1574 + }, + { + "epoch": 1.1780104712041886, + "grad_norm": 2.5939737093072317, + "learning_rate": 3.9477469594311975e-06, + "loss": 1.479, + "step": 1575 + }, + { + "epoch": 1.1787584143605085, + "grad_norm": 2.007192373604227, + "learning_rate": 3.941704406113989e-06, + "loss": 1.4887, + "step": 1576 + }, + { + "epoch": 1.1795063575168288, + "grad_norm": 1.968276770756505, + "learning_rate": 3.9356634704944065e-06, + "loss": 1.6026, + "step": 1577 + }, + { + "epoch": 1.1802543006731487, + "grad_norm": 2.1957470060610085, + "learning_rate": 3.929624161806548e-06, + "loss": 1.7089, + "step": 1578 + }, + { + "epoch": 1.181002243829469, + "grad_norm": 2.352574777761244, + "learning_rate": 3.923586489282028e-06, + "loss": 1.3553, + "step": 1579 + }, + { + "epoch": 1.1817501869857892, + "grad_norm": 2.4058615937214634, + "learning_rate": 3.917550462149959e-06, + "loss": 1.3958, + "step": 1580 + }, + { + "epoch": 1.1824981301421091, + "grad_norm": 2.3670584333742344, + "learning_rate": 3.911516089636936e-06, + "loss": 1.4296, + "step": 1581 + }, + { + "epoch": 1.1832460732984293, + "grad_norm": 2.115391678097799, + "learning_rate": 3.905483380967027e-06, + "loss": 1.518, + "step": 1582 + }, + { + "epoch": 1.1839940164547493, + "grad_norm": 2.119220388136537, + "learning_rate": 3.899452345361756e-06, + "loss": 1.4707, + "step": 1583 + }, + { + "epoch": 1.1847419596110695, + "grad_norm": 2.2057692448321595, + "learning_rate": 3.893422992040092e-06, + "loss": 1.3134, + "step": 1584 + }, + { + "epoch": 1.1854899027673897, + "grad_norm": 2.0272908165876236, + "learning_rate": 3.887395330218429e-06, + "loss": 1.8119, + "step": 1585 + }, + { + "epoch": 1.1862378459237097, + "grad_norm": 1.9253771989876713, + "learning_rate": 3.881369369110575e-06, + "loss": 1.5278, + "step": 1586 + }, + { + "epoch": 1.18698578908003, + "grad_norm": 3.1109532726826137, + "learning_rate": 3.8753451179277425e-06, + "loss": 1.5198, + "step": 1587 + }, + { + "epoch": 1.18773373223635, + "grad_norm": 2.142413978769863, + "learning_rate": 3.869322585878527e-06, + "loss": 1.6169, + "step": 1588 + }, + { + "epoch": 1.1884816753926701, + "grad_norm": 2.2749891489041834, + "learning_rate": 3.863301782168896e-06, + "loss": 1.4593, + "step": 1589 + }, + { + "epoch": 1.1892296185489903, + "grad_norm": 2.070707422242646, + "learning_rate": 3.857282716002178e-06, + "loss": 1.3807, + "step": 1590 + }, + { + "epoch": 1.1899775617053103, + "grad_norm": 2.046899188521169, + "learning_rate": 3.8512653965790395e-06, + "loss": 1.373, + "step": 1591 + }, + { + "epoch": 1.1907255048616305, + "grad_norm": 2.1068182717625654, + "learning_rate": 3.845249833097486e-06, + "loss": 1.461, + "step": 1592 + }, + { + "epoch": 1.1914734480179505, + "grad_norm": 2.0442466505173704, + "learning_rate": 3.8392360347528315e-06, + "loss": 1.6108, + "step": 1593 + }, + { + "epoch": 1.1922213911742707, + "grad_norm": 3.686457346566068, + "learning_rate": 3.8332240107376925e-06, + "loss": 1.5842, + "step": 1594 + }, + { + "epoch": 1.192969334330591, + "grad_norm": 2.111789931002416, + "learning_rate": 3.827213770241977e-06, + "loss": 1.7348, + "step": 1595 + }, + { + "epoch": 1.193717277486911, + "grad_norm": 2.0691440976369204, + "learning_rate": 3.821205322452863e-06, + "loss": 1.7118, + "step": 1596 + }, + { + "epoch": 1.1944652206432311, + "grad_norm": 2.029407796286081, + "learning_rate": 3.815198676554794e-06, + "loss": 1.755, + "step": 1597 + }, + { + "epoch": 1.1952131637995511, + "grad_norm": 2.0984132632131822, + "learning_rate": 3.8091938417294515e-06, + "loss": 1.4562, + "step": 1598 + }, + { + "epoch": 1.1959611069558713, + "grad_norm": 2.8078832161623453, + "learning_rate": 3.8031908271557504e-06, + "loss": 1.7228, + "step": 1599 + }, + { + "epoch": 1.1967090501121915, + "grad_norm": 2.156619634597294, + "learning_rate": 3.7971896420098296e-06, + "loss": 1.5143, + "step": 1600 + }, + { + "epoch": 1.1974569932685115, + "grad_norm": 1.8906104738483296, + "learning_rate": 3.7911902954650226e-06, + "loss": 1.532, + "step": 1601 + }, + { + "epoch": 1.1982049364248317, + "grad_norm": 2.086360068592666, + "learning_rate": 3.7851927966918602e-06, + "loss": 1.4256, + "step": 1602 + }, + { + "epoch": 1.1989528795811517, + "grad_norm": 2.395149094682344, + "learning_rate": 3.779197154858044e-06, + "loss": 1.317, + "step": 1603 + }, + { + "epoch": 1.199700822737472, + "grad_norm": 2.1046913677440413, + "learning_rate": 3.7732033791284367e-06, + "loss": 1.4774, + "step": 1604 + }, + { + "epoch": 1.2004487658937921, + "grad_norm": 1.9706393447251682, + "learning_rate": 3.767211478665054e-06, + "loss": 1.5447, + "step": 1605 + }, + { + "epoch": 1.201196709050112, + "grad_norm": 1.8893382634309561, + "learning_rate": 3.761221462627034e-06, + "loss": 1.6797, + "step": 1606 + }, + { + "epoch": 1.2019446522064323, + "grad_norm": 1.8820623447999296, + "learning_rate": 3.7552333401706508e-06, + "loss": 1.4426, + "step": 1607 + }, + { + "epoch": 1.2026925953627525, + "grad_norm": 2.091022526868933, + "learning_rate": 3.74924712044927e-06, + "loss": 1.5766, + "step": 1608 + }, + { + "epoch": 1.2034405385190725, + "grad_norm": 2.111434197219793, + "learning_rate": 3.7432628126133497e-06, + "loss": 1.5172, + "step": 1609 + }, + { + "epoch": 1.2041884816753927, + "grad_norm": 1.9246113778182963, + "learning_rate": 3.7372804258104367e-06, + "loss": 1.4931, + "step": 1610 + }, + { + "epoch": 1.2049364248317127, + "grad_norm": 2.254091992931464, + "learning_rate": 3.731299969185131e-06, + "loss": 1.412, + "step": 1611 + }, + { + "epoch": 1.205684367988033, + "grad_norm": 2.1306218659684704, + "learning_rate": 3.7253214518790814e-06, + "loss": 1.5683, + "step": 1612 + }, + { + "epoch": 1.2064323111443531, + "grad_norm": 2.2052352219865785, + "learning_rate": 3.7193448830309802e-06, + "loss": 1.8866, + "step": 1613 + }, + { + "epoch": 1.207180254300673, + "grad_norm": 2.2287244056516466, + "learning_rate": 3.7133702717765345e-06, + "loss": 1.4006, + "step": 1614 + }, + { + "epoch": 1.2079281974569933, + "grad_norm": 2.295719552674388, + "learning_rate": 3.7073976272484647e-06, + "loss": 1.6049, + "step": 1615 + }, + { + "epoch": 1.2086761406133133, + "grad_norm": 2.101072674489867, + "learning_rate": 3.7014269585764796e-06, + "loss": 1.6819, + "step": 1616 + }, + { + "epoch": 1.2094240837696335, + "grad_norm": 2.076571600308359, + "learning_rate": 3.695458274887268e-06, + "loss": 1.3212, + "step": 1617 + }, + { + "epoch": 1.2101720269259537, + "grad_norm": 2.74301762870769, + "learning_rate": 3.689491585304491e-06, + "loss": 1.4217, + "step": 1618 + }, + { + "epoch": 1.2109199700822737, + "grad_norm": 1.686543064355575, + "learning_rate": 3.6835268989487518e-06, + "loss": 1.3812, + "step": 1619 + }, + { + "epoch": 1.211667913238594, + "grad_norm": 1.9286805138871737, + "learning_rate": 3.6775642249376e-06, + "loss": 1.1726, + "step": 1620 + }, + { + "epoch": 1.212415856394914, + "grad_norm": 2.1142041101951685, + "learning_rate": 3.671603572385506e-06, + "loss": 1.7279, + "step": 1621 + }, + { + "epoch": 1.213163799551234, + "grad_norm": 2.055831625130678, + "learning_rate": 3.665644950403847e-06, + "loss": 1.5606, + "step": 1622 + }, + { + "epoch": 1.2139117427075543, + "grad_norm": 2.3743385175915437, + "learning_rate": 3.659688368100902e-06, + "loss": 1.5733, + "step": 1623 + }, + { + "epoch": 1.2146596858638743, + "grad_norm": 2.263426713903562, + "learning_rate": 3.6537338345818273e-06, + "loss": 1.3476, + "step": 1624 + }, + { + "epoch": 1.2154076290201945, + "grad_norm": 1.7459030946799412, + "learning_rate": 3.647781358948653e-06, + "loss": 1.6408, + "step": 1625 + }, + { + "epoch": 1.2161555721765147, + "grad_norm": 2.202064345918964, + "learning_rate": 3.641830950300257e-06, + "loss": 1.7434, + "step": 1626 + }, + { + "epoch": 1.2169035153328347, + "grad_norm": 2.8740427496601346, + "learning_rate": 3.63588261773236e-06, + "loss": 1.2815, + "step": 1627 + }, + { + "epoch": 1.217651458489155, + "grad_norm": 1.9854795920134987, + "learning_rate": 3.629936370337514e-06, + "loss": 1.6692, + "step": 1628 + }, + { + "epoch": 1.2183994016454749, + "grad_norm": 2.2197429385874483, + "learning_rate": 3.623992217205075e-06, + "loss": 1.6915, + "step": 1629 + }, + { + "epoch": 1.219147344801795, + "grad_norm": 1.8816022936506722, + "learning_rate": 3.6180501674212052e-06, + "loss": 1.6606, + "step": 1630 + }, + { + "epoch": 1.2198952879581153, + "grad_norm": 2.6756246977409095, + "learning_rate": 3.6121102300688504e-06, + "loss": 1.3809, + "step": 1631 + }, + { + "epoch": 1.2206432311144353, + "grad_norm": 2.2031443500880687, + "learning_rate": 3.606172414227721e-06, + "loss": 1.5601, + "step": 1632 + }, + { + "epoch": 1.2213911742707555, + "grad_norm": 1.9378681497830372, + "learning_rate": 3.600236728974292e-06, + "loss": 1.5971, + "step": 1633 + }, + { + "epoch": 1.2221391174270755, + "grad_norm": 2.3367339060710592, + "learning_rate": 3.5943031833817788e-06, + "loss": 1.4479, + "step": 1634 + }, + { + "epoch": 1.2228870605833957, + "grad_norm": 2.404619965527919, + "learning_rate": 3.588371786520127e-06, + "loss": 1.3655, + "step": 1635 + }, + { + "epoch": 1.223635003739716, + "grad_norm": 2.6826674700464896, + "learning_rate": 3.5824425474559956e-06, + "loss": 1.6138, + "step": 1636 + }, + { + "epoch": 1.2243829468960359, + "grad_norm": 2.740413966270276, + "learning_rate": 3.576515475252745e-06, + "loss": 1.5643, + "step": 1637 + }, + { + "epoch": 1.225130890052356, + "grad_norm": 3.079481798358016, + "learning_rate": 3.5705905789704296e-06, + "loss": 1.5559, + "step": 1638 + }, + { + "epoch": 1.225878833208676, + "grad_norm": 2.456024217146243, + "learning_rate": 3.5646678676657687e-06, + "loss": 1.5508, + "step": 1639 + }, + { + "epoch": 1.2266267763649963, + "grad_norm": 1.8870637056098845, + "learning_rate": 3.558747350392146e-06, + "loss": 1.6663, + "step": 1640 + }, + { + "epoch": 1.2273747195213165, + "grad_norm": 1.939431794053562, + "learning_rate": 3.552829036199595e-06, + "loss": 1.4707, + "step": 1641 + }, + { + "epoch": 1.2281226626776365, + "grad_norm": 2.1590051575688474, + "learning_rate": 3.546912934134773e-06, + "loss": 1.3361, + "step": 1642 + }, + { + "epoch": 1.2288706058339567, + "grad_norm": 2.348730325941636, + "learning_rate": 3.540999053240965e-06, + "loss": 1.5024, + "step": 1643 + }, + { + "epoch": 1.2296185489902767, + "grad_norm": 2.2729749048702246, + "learning_rate": 3.5350874025580555e-06, + "loss": 1.549, + "step": 1644 + }, + { + "epoch": 1.2303664921465969, + "grad_norm": 2.249363974688746, + "learning_rate": 3.529177991122519e-06, + "loss": 1.5379, + "step": 1645 + }, + { + "epoch": 1.231114435302917, + "grad_norm": 1.8758525067524716, + "learning_rate": 3.523270827967411e-06, + "loss": 1.5456, + "step": 1646 + }, + { + "epoch": 1.231862378459237, + "grad_norm": 1.9733822602046769, + "learning_rate": 3.517365922122348e-06, + "loss": 1.6445, + "step": 1647 + }, + { + "epoch": 1.2326103216155573, + "grad_norm": 2.2759285090396952, + "learning_rate": 3.5114632826134977e-06, + "loss": 1.7006, + "step": 1648 + }, + { + "epoch": 1.2333582647718773, + "grad_norm": 2.4455769844012627, + "learning_rate": 3.50556291846356e-06, + "loss": 1.7125, + "step": 1649 + }, + { + "epoch": 1.2341062079281975, + "grad_norm": 2.1884572695095272, + "learning_rate": 3.4996648386917585e-06, + "loss": 1.4852, + "step": 1650 + }, + { + "epoch": 1.2348541510845177, + "grad_norm": 2.394116778114252, + "learning_rate": 3.4937690523138302e-06, + "loss": 1.4142, + "step": 1651 + }, + { + "epoch": 1.2356020942408377, + "grad_norm": 2.2811587028495435, + "learning_rate": 3.487875568341995e-06, + "loss": 1.6287, + "step": 1652 + }, + { + "epoch": 1.2363500373971579, + "grad_norm": 2.1417287980117043, + "learning_rate": 3.4819843957849655e-06, + "loss": 1.3024, + "step": 1653 + }, + { + "epoch": 1.2370979805534779, + "grad_norm": 2.254021953439467, + "learning_rate": 3.4760955436479134e-06, + "loss": 1.1949, + "step": 1654 + }, + { + "epoch": 1.237845923709798, + "grad_norm": 1.9566745256920022, + "learning_rate": 3.4702090209324634e-06, + "loss": 1.6019, + "step": 1655 + }, + { + "epoch": 1.2385938668661183, + "grad_norm": 2.4245562215627037, + "learning_rate": 3.4643248366366853e-06, + "loss": 1.6498, + "step": 1656 + }, + { + "epoch": 1.2393418100224383, + "grad_norm": 2.3055077224865808, + "learning_rate": 3.4584429997550685e-06, + "loss": 1.5815, + "step": 1657 + }, + { + "epoch": 1.2400897531787585, + "grad_norm": 2.0268345824896783, + "learning_rate": 3.45256351927852e-06, + "loss": 1.362, + "step": 1658 + }, + { + "epoch": 1.2408376963350785, + "grad_norm": 2.4536004081982794, + "learning_rate": 3.446686404194337e-06, + "loss": 1.5139, + "step": 1659 + }, + { + "epoch": 1.2415856394913987, + "grad_norm": 2.152771841589576, + "learning_rate": 3.440811663486206e-06, + "loss": 1.6137, + "step": 1660 + }, + { + "epoch": 1.2423335826477189, + "grad_norm": 1.8538201048289418, + "learning_rate": 3.434939306134186e-06, + "loss": 1.6837, + "step": 1661 + }, + { + "epoch": 1.2430815258040389, + "grad_norm": 2.0190402569352095, + "learning_rate": 3.4290693411146882e-06, + "loss": 1.6993, + "step": 1662 + }, + { + "epoch": 1.243829468960359, + "grad_norm": 2.138760328028405, + "learning_rate": 3.423201777400469e-06, + "loss": 1.5996, + "step": 1663 + }, + { + "epoch": 1.244577412116679, + "grad_norm": 2.22781362176441, + "learning_rate": 3.4173366239606153e-06, + "loss": 1.6564, + "step": 1664 + }, + { + "epoch": 1.2453253552729993, + "grad_norm": 2.399419354917597, + "learning_rate": 3.4114738897605267e-06, + "loss": 1.5008, + "step": 1665 + }, + { + "epoch": 1.2460732984293195, + "grad_norm": 2.107397290810108, + "learning_rate": 3.4056135837619077e-06, + "loss": 1.8004, + "step": 1666 + }, + { + "epoch": 1.2468212415856395, + "grad_norm": 2.1521021969497145, + "learning_rate": 3.399755714922751e-06, + "loss": 1.5733, + "step": 1667 + }, + { + "epoch": 1.2475691847419597, + "grad_norm": 2.3099963031106836, + "learning_rate": 3.3939002921973197e-06, + "loss": 1.4852, + "step": 1668 + }, + { + "epoch": 1.2483171278982796, + "grad_norm": 1.9801009608800644, + "learning_rate": 3.388047324536143e-06, + "loss": 1.5452, + "step": 1669 + }, + { + "epoch": 1.2490650710545999, + "grad_norm": 1.8321029654636065, + "learning_rate": 3.3821968208859955e-06, + "loss": 1.5414, + "step": 1670 + }, + { + "epoch": 1.24981301421092, + "grad_norm": 2.0286699927351424, + "learning_rate": 3.376348790189886e-06, + "loss": 1.3151, + "step": 1671 + }, + { + "epoch": 1.25056095736724, + "grad_norm": 2.183180099857848, + "learning_rate": 3.3705032413870402e-06, + "loss": 1.6566, + "step": 1672 + }, + { + "epoch": 1.2513089005235603, + "grad_norm": 2.4832735526448624, + "learning_rate": 3.3646601834128924e-06, + "loss": 1.7245, + "step": 1673 + }, + { + "epoch": 1.2520568436798802, + "grad_norm": 2.4171200224868215, + "learning_rate": 3.358819625199071e-06, + "loss": 1.5168, + "step": 1674 + }, + { + "epoch": 1.2528047868362004, + "grad_norm": 3.3591896618265014, + "learning_rate": 3.3529815756733773e-06, + "loss": 1.4946, + "step": 1675 + }, + { + "epoch": 1.2528047868362004, + "eval_loss": 1.6712989807128906, + "eval_runtime": 68.965, + "eval_samples_per_second": 59.059, + "eval_steps_per_second": 3.698, + "step": 1675 + }, + { + "epoch": 1.2535527299925207, + "grad_norm": 1.9481467214917645, + "learning_rate": 3.347146043759787e-06, + "loss": 1.3298, + "step": 1676 + }, + { + "epoch": 1.2543006731488406, + "grad_norm": 1.9920886110776446, + "learning_rate": 3.34131303837842e-06, + "loss": 1.4474, + "step": 1677 + }, + { + "epoch": 1.2550486163051608, + "grad_norm": 2.104497889698879, + "learning_rate": 3.335482568445536e-06, + "loss": 1.5518, + "step": 1678 + }, + { + "epoch": 1.2557965594614808, + "grad_norm": 2.7621995141742484, + "learning_rate": 3.3296546428735206e-06, + "loss": 1.3706, + "step": 1679 + }, + { + "epoch": 1.256544502617801, + "grad_norm": 2.254031798451383, + "learning_rate": 3.3238292705708675e-06, + "loss": 1.5863, + "step": 1680 + }, + { + "epoch": 1.2572924457741212, + "grad_norm": 2.261530384619498, + "learning_rate": 3.318006460442173e-06, + "loss": 1.489, + "step": 1681 + }, + { + "epoch": 1.2580403889304412, + "grad_norm": 2.0547187351068903, + "learning_rate": 3.3121862213881093e-06, + "loss": 1.4568, + "step": 1682 + }, + { + "epoch": 1.2587883320867614, + "grad_norm": 2.5008877866514823, + "learning_rate": 3.306368562305422e-06, + "loss": 1.5741, + "step": 1683 + }, + { + "epoch": 1.2595362752430814, + "grad_norm": 1.9911232859387435, + "learning_rate": 3.3005534920869175e-06, + "loss": 1.62, + "step": 1684 + }, + { + "epoch": 1.2602842183994016, + "grad_norm": 2.1140436057246608, + "learning_rate": 3.2947410196214353e-06, + "loss": 1.3291, + "step": 1685 + }, + { + "epoch": 1.2610321615557218, + "grad_norm": 2.3243201154957593, + "learning_rate": 3.288931153793853e-06, + "loss": 1.4692, + "step": 1686 + }, + { + "epoch": 1.2617801047120418, + "grad_norm": 1.8299948280616072, + "learning_rate": 3.2831239034850593e-06, + "loss": 1.5196, + "step": 1687 + }, + { + "epoch": 1.262528047868362, + "grad_norm": 2.255754277513295, + "learning_rate": 3.2773192775719432e-06, + "loss": 1.6094, + "step": 1688 + }, + { + "epoch": 1.263275991024682, + "grad_norm": 2.075390758028022, + "learning_rate": 3.2715172849273867e-06, + "loss": 1.5118, + "step": 1689 + }, + { + "epoch": 1.2640239341810022, + "grad_norm": 2.4615405904921817, + "learning_rate": 3.265717934420244e-06, + "loss": 1.6022, + "step": 1690 + }, + { + "epoch": 1.2647718773373224, + "grad_norm": 2.4717882683213146, + "learning_rate": 3.259921234915332e-06, + "loss": 1.5316, + "step": 1691 + }, + { + "epoch": 1.2655198204936424, + "grad_norm": 2.2826094789192215, + "learning_rate": 3.254127195273412e-06, + "loss": 1.7106, + "step": 1692 + }, + { + "epoch": 1.2662677636499626, + "grad_norm": 2.1047091252420245, + "learning_rate": 3.248335824351183e-06, + "loss": 1.4425, + "step": 1693 + }, + { + "epoch": 1.2670157068062826, + "grad_norm": 2.4665674065588834, + "learning_rate": 3.2425471310012645e-06, + "loss": 1.4623, + "step": 1694 + }, + { + "epoch": 1.2677636499626028, + "grad_norm": 2.393787003441449, + "learning_rate": 3.2367611240721796e-06, + "loss": 1.6164, + "step": 1695 + }, + { + "epoch": 1.268511593118923, + "grad_norm": 2.258113903359061, + "learning_rate": 3.230977812408348e-06, + "loss": 1.622, + "step": 1696 + }, + { + "epoch": 1.269259536275243, + "grad_norm": 2.363449710184473, + "learning_rate": 3.22519720485007e-06, + "loss": 1.146, + "step": 1697 + }, + { + "epoch": 1.2700074794315632, + "grad_norm": 2.164353285897623, + "learning_rate": 3.219419310233509e-06, + "loss": 1.4688, + "step": 1698 + }, + { + "epoch": 1.2707554225878832, + "grad_norm": 3.0400332866219455, + "learning_rate": 3.213644137390686e-06, + "loss": 1.2771, + "step": 1699 + }, + { + "epoch": 1.2715033657442034, + "grad_norm": 2.1694455543962987, + "learning_rate": 3.207871695149459e-06, + "loss": 1.6192, + "step": 1700 + }, + { + "epoch": 1.2722513089005236, + "grad_norm": 2.222746150782225, + "learning_rate": 3.2021019923335093e-06, + "loss": 1.1453, + "step": 1701 + }, + { + "epoch": 1.2729992520568436, + "grad_norm": 2.0797366017775127, + "learning_rate": 3.196335037762337e-06, + "loss": 1.476, + "step": 1702 + }, + { + "epoch": 1.2737471952131638, + "grad_norm": 2.7924922620421246, + "learning_rate": 3.1905708402512347e-06, + "loss": 1.674, + "step": 1703 + }, + { + "epoch": 1.2744951383694838, + "grad_norm": 2.8012261627629162, + "learning_rate": 3.1848094086112887e-06, + "loss": 1.4342, + "step": 1704 + }, + { + "epoch": 1.275243081525804, + "grad_norm": 2.3926593543413213, + "learning_rate": 3.1790507516493473e-06, + "loss": 1.5141, + "step": 1705 + }, + { + "epoch": 1.2759910246821242, + "grad_norm": 2.0953232200361738, + "learning_rate": 3.173294878168025e-06, + "loss": 1.5372, + "step": 1706 + }, + { + "epoch": 1.2767389678384442, + "grad_norm": 2.2382690292909326, + "learning_rate": 3.167541796965681e-06, + "loss": 1.4189, + "step": 1707 + }, + { + "epoch": 1.2774869109947644, + "grad_norm": 2.2402080303217597, + "learning_rate": 3.1617915168363994e-06, + "loss": 1.4491, + "step": 1708 + }, + { + "epoch": 1.2782348541510844, + "grad_norm": 2.0533204519840487, + "learning_rate": 3.1560440465699924e-06, + "loss": 1.5113, + "step": 1709 + }, + { + "epoch": 1.2789827973074046, + "grad_norm": 2.0326272137836314, + "learning_rate": 3.150299394951972e-06, + "loss": 1.5449, + "step": 1710 + }, + { + "epoch": 1.2797307404637248, + "grad_norm": 2.3240725752367775, + "learning_rate": 3.1445575707635383e-06, + "loss": 1.7284, + "step": 1711 + }, + { + "epoch": 1.2804786836200448, + "grad_norm": 2.615343897074273, + "learning_rate": 3.1388185827815766e-06, + "loss": 1.5289, + "step": 1712 + }, + { + "epoch": 1.281226626776365, + "grad_norm": 2.0261255007725127, + "learning_rate": 3.1330824397786317e-06, + "loss": 1.6033, + "step": 1713 + }, + { + "epoch": 1.281974569932685, + "grad_norm": 2.4598098417035237, + "learning_rate": 3.127349150522903e-06, + "loss": 1.3465, + "step": 1714 + }, + { + "epoch": 1.2827225130890052, + "grad_norm": 2.0369692879527306, + "learning_rate": 3.121618723778225e-06, + "loss": 1.2828, + "step": 1715 + }, + { + "epoch": 1.2834704562453254, + "grad_norm": 1.8510456416359524, + "learning_rate": 3.115891168304055e-06, + "loss": 1.5925, + "step": 1716 + }, + { + "epoch": 1.2842183994016454, + "grad_norm": 1.9430472274932142, + "learning_rate": 3.110166492855468e-06, + "loss": 1.8234, + "step": 1717 + }, + { + "epoch": 1.2849663425579656, + "grad_norm": 2.2595696622174004, + "learning_rate": 3.104444706183129e-06, + "loss": 1.7027, + "step": 1718 + }, + { + "epoch": 1.2857142857142856, + "grad_norm": 2.155863943588656, + "learning_rate": 3.0987258170332935e-06, + "loss": 1.4707, + "step": 1719 + }, + { + "epoch": 1.2864622288706058, + "grad_norm": 2.114001510735076, + "learning_rate": 3.093009834147784e-06, + "loss": 1.294, + "step": 1720 + }, + { + "epoch": 1.287210172026926, + "grad_norm": 1.9839017697076802, + "learning_rate": 3.0872967662639795e-06, + "loss": 1.5257, + "step": 1721 + }, + { + "epoch": 1.2879581151832462, + "grad_norm": 2.0393129416401434, + "learning_rate": 3.081586622114809e-06, + "loss": 1.6673, + "step": 1722 + }, + { + "epoch": 1.2887060583395662, + "grad_norm": 1.938111220135879, + "learning_rate": 3.0758794104287275e-06, + "loss": 1.7307, + "step": 1723 + }, + { + "epoch": 1.2894540014958862, + "grad_norm": 2.0974851331960673, + "learning_rate": 3.0701751399297063e-06, + "loss": 1.6462, + "step": 1724 + }, + { + "epoch": 1.2902019446522064, + "grad_norm": 2.2897371378356275, + "learning_rate": 3.0644738193372246e-06, + "loss": 1.4889, + "step": 1725 + }, + { + "epoch": 1.2909498878085266, + "grad_norm": 2.786051541290913, + "learning_rate": 3.0587754573662508e-06, + "loss": 1.6228, + "step": 1726 + }, + { + "epoch": 1.2916978309648468, + "grad_norm": 1.921370330055876, + "learning_rate": 3.053080062727233e-06, + "loss": 1.4697, + "step": 1727 + }, + { + "epoch": 1.2924457741211668, + "grad_norm": 2.2621047352762256, + "learning_rate": 3.0473876441260786e-06, + "loss": 1.6333, + "step": 1728 + }, + { + "epoch": 1.2931937172774868, + "grad_norm": 2.45707148311356, + "learning_rate": 3.041698210264149e-06, + "loss": 1.662, + "step": 1729 + }, + { + "epoch": 1.293941660433807, + "grad_norm": 2.184862585548346, + "learning_rate": 3.0360117698382446e-06, + "loss": 1.4745, + "step": 1730 + }, + { + "epoch": 1.2946896035901272, + "grad_norm": 1.9377170006372952, + "learning_rate": 3.0303283315405853e-06, + "loss": 1.3894, + "step": 1731 + }, + { + "epoch": 1.2954375467464474, + "grad_norm": 1.9045528269727179, + "learning_rate": 3.0246479040588077e-06, + "loss": 1.4932, + "step": 1732 + }, + { + "epoch": 1.2961854899027674, + "grad_norm": 2.388888038731087, + "learning_rate": 3.0189704960759436e-06, + "loss": 1.5239, + "step": 1733 + }, + { + "epoch": 1.2969334330590876, + "grad_norm": 2.086665880947813, + "learning_rate": 3.0132961162704054e-06, + "loss": 1.7607, + "step": 1734 + }, + { + "epoch": 1.2976813762154076, + "grad_norm": 2.058859736568643, + "learning_rate": 3.0076247733159846e-06, + "loss": 1.5251, + "step": 1735 + }, + { + "epoch": 1.2984293193717278, + "grad_norm": 2.1321951482071966, + "learning_rate": 3.001956475881822e-06, + "loss": 1.5424, + "step": 1736 + }, + { + "epoch": 1.299177262528048, + "grad_norm": 2.3891564295649053, + "learning_rate": 2.996291232632412e-06, + "loss": 1.6653, + "step": 1737 + }, + { + "epoch": 1.299925205684368, + "grad_norm": 1.9243918248031042, + "learning_rate": 2.9906290522275738e-06, + "loss": 1.6631, + "step": 1738 + }, + { + "epoch": 1.3006731488406882, + "grad_norm": 2.448526793094096, + "learning_rate": 2.9849699433224423e-06, + "loss": 1.5982, + "step": 1739 + }, + { + "epoch": 1.3014210919970082, + "grad_norm": 1.8116003031683892, + "learning_rate": 2.97931391456747e-06, + "loss": 1.4758, + "step": 1740 + }, + { + "epoch": 1.3021690351533284, + "grad_norm": 1.9608994206796753, + "learning_rate": 2.9736609746083853e-06, + "loss": 1.58, + "step": 1741 + }, + { + "epoch": 1.3029169783096486, + "grad_norm": 2.18585360783782, + "learning_rate": 2.968011132086209e-06, + "loss": 1.3495, + "step": 1742 + }, + { + "epoch": 1.3036649214659686, + "grad_norm": 2.245764327735025, + "learning_rate": 2.962364395637216e-06, + "loss": 1.3265, + "step": 1743 + }, + { + "epoch": 1.3044128646222888, + "grad_norm": 2.1958799695995603, + "learning_rate": 2.9567207738929395e-06, + "loss": 1.7286, + "step": 1744 + }, + { + "epoch": 1.3051608077786088, + "grad_norm": 2.3001304505073152, + "learning_rate": 2.9510802754801526e-06, + "loss": 1.683, + "step": 1745 + }, + { + "epoch": 1.305908750934929, + "grad_norm": 2.346572307292022, + "learning_rate": 2.945442909020847e-06, + "loss": 1.4967, + "step": 1746 + }, + { + "epoch": 1.3066566940912492, + "grad_norm": 1.902159647418323, + "learning_rate": 2.939808683132238e-06, + "loss": 1.4605, + "step": 1747 + }, + { + "epoch": 1.3074046372475692, + "grad_norm": 2.0329813452887735, + "learning_rate": 2.93417760642673e-06, + "loss": 1.5743, + "step": 1748 + }, + { + "epoch": 1.3081525804038894, + "grad_norm": 2.1492338969494846, + "learning_rate": 2.928549687511917e-06, + "loss": 1.4491, + "step": 1749 + }, + { + "epoch": 1.3089005235602094, + "grad_norm": 1.9237915494716844, + "learning_rate": 2.9229249349905686e-06, + "loss": 1.5336, + "step": 1750 + }, + { + "epoch": 1.3096484667165296, + "grad_norm": 2.150681646170878, + "learning_rate": 2.9173033574606112e-06, + "loss": 1.398, + "step": 1751 + }, + { + "epoch": 1.3103964098728498, + "grad_norm": 1.9481269775835857, + "learning_rate": 2.911684963515119e-06, + "loss": 1.5551, + "step": 1752 + }, + { + "epoch": 1.3111443530291698, + "grad_norm": 2.458912548316881, + "learning_rate": 2.906069761742302e-06, + "loss": 1.1692, + "step": 1753 + }, + { + "epoch": 1.31189229618549, + "grad_norm": 2.1078326619990104, + "learning_rate": 2.9004577607254847e-06, + "loss": 1.6895, + "step": 1754 + }, + { + "epoch": 1.31264023934181, + "grad_norm": 1.8565445149993776, + "learning_rate": 2.8948489690431062e-06, + "loss": 1.3806, + "step": 1755 + }, + { + "epoch": 1.3133881824981302, + "grad_norm": 2.84904868215971, + "learning_rate": 2.889243395268696e-06, + "loss": 1.8232, + "step": 1756 + }, + { + "epoch": 1.3141361256544504, + "grad_norm": 2.390083139761547, + "learning_rate": 2.8836410479708625e-06, + "loss": 1.3284, + "step": 1757 + }, + { + "epoch": 1.3148840688107704, + "grad_norm": 2.364046787574681, + "learning_rate": 2.8780419357132864e-06, + "loss": 1.5502, + "step": 1758 + }, + { + "epoch": 1.3156320119670906, + "grad_norm": 1.9871285754130654, + "learning_rate": 2.872446067054703e-06, + "loss": 1.4965, + "step": 1759 + }, + { + "epoch": 1.3163799551234106, + "grad_norm": 2.1085007399000073, + "learning_rate": 2.8668534505488887e-06, + "loss": 1.677, + "step": 1760 + }, + { + "epoch": 1.3171278982797308, + "grad_norm": 2.234165086924829, + "learning_rate": 2.861264094744647e-06, + "loss": 1.5371, + "step": 1761 + }, + { + "epoch": 1.317875841436051, + "grad_norm": 1.9732692045769376, + "learning_rate": 2.8556780081857966e-06, + "loss": 1.3184, + "step": 1762 + }, + { + "epoch": 1.318623784592371, + "grad_norm": 2.513721805781147, + "learning_rate": 2.8500951994111647e-06, + "loss": 1.5283, + "step": 1763 + }, + { + "epoch": 1.3193717277486912, + "grad_norm": 2.004756240277015, + "learning_rate": 2.84451567695456e-06, + "loss": 1.5133, + "step": 1764 + }, + { + "epoch": 1.3201196709050111, + "grad_norm": 2.3077439296814437, + "learning_rate": 2.8389394493447732e-06, + "loss": 1.3235, + "step": 1765 + }, + { + "epoch": 1.3208676140613314, + "grad_norm": 2.6745914985862673, + "learning_rate": 2.833366525105561e-06, + "loss": 1.5631, + "step": 1766 + }, + { + "epoch": 1.3216155572176516, + "grad_norm": 2.245716186642266, + "learning_rate": 2.8277969127556204e-06, + "loss": 1.3607, + "step": 1767 + }, + { + "epoch": 1.3223635003739715, + "grad_norm": 2.184157045393329, + "learning_rate": 2.8222306208085988e-06, + "loss": 1.7719, + "step": 1768 + }, + { + "epoch": 1.3231114435302918, + "grad_norm": 1.8373951011940246, + "learning_rate": 2.8166676577730556e-06, + "loss": 1.5606, + "step": 1769 + }, + { + "epoch": 1.3238593866866117, + "grad_norm": 2.0423621209515876, + "learning_rate": 2.811108032152473e-06, + "loss": 1.5057, + "step": 1770 + }, + { + "epoch": 1.324607329842932, + "grad_norm": 1.9897868356427515, + "learning_rate": 2.805551752445222e-06, + "loss": 1.5885, + "step": 1771 + }, + { + "epoch": 1.3253552729992522, + "grad_norm": 2.0574818113631714, + "learning_rate": 2.7999988271445643e-06, + "loss": 1.7286, + "step": 1772 + }, + { + "epoch": 1.3261032161555721, + "grad_norm": 2.6130085661636033, + "learning_rate": 2.794449264738637e-06, + "loss": 1.363, + "step": 1773 + }, + { + "epoch": 1.3268511593118923, + "grad_norm": 2.013923304844127, + "learning_rate": 2.7889030737104283e-06, + "loss": 1.6081, + "step": 1774 + }, + { + "epoch": 1.3275991024682123, + "grad_norm": 1.9023818517377722, + "learning_rate": 2.7833602625377803e-06, + "loss": 1.7143, + "step": 1775 + }, + { + "epoch": 1.3283470456245325, + "grad_norm": 2.224244867758022, + "learning_rate": 2.7778208396933657e-06, + "loss": 1.6687, + "step": 1776 + }, + { + "epoch": 1.3290949887808527, + "grad_norm": 1.9836697751838352, + "learning_rate": 2.772284813644675e-06, + "loss": 1.5175, + "step": 1777 + }, + { + "epoch": 1.3298429319371727, + "grad_norm": 2.213220187812896, + "learning_rate": 2.766752192854012e-06, + "loss": 1.6144, + "step": 1778 + }, + { + "epoch": 1.330590875093493, + "grad_norm": 2.1195559934276376, + "learning_rate": 2.7612229857784743e-06, + "loss": 1.3845, + "step": 1779 + }, + { + "epoch": 1.331338818249813, + "grad_norm": 2.3903361350737464, + "learning_rate": 2.755697200869937e-06, + "loss": 1.6488, + "step": 1780 + }, + { + "epoch": 1.3320867614061331, + "grad_norm": 1.9483696790854605, + "learning_rate": 2.7501748465750498e-06, + "loss": 1.5203, + "step": 1781 + }, + { + "epoch": 1.3328347045624533, + "grad_norm": 2.4817071702187823, + "learning_rate": 2.7446559313352117e-06, + "loss": 1.3721, + "step": 1782 + }, + { + "epoch": 1.3335826477187733, + "grad_norm": 2.424772199505897, + "learning_rate": 2.7391404635865725e-06, + "loss": 1.5574, + "step": 1783 + }, + { + "epoch": 1.3343305908750935, + "grad_norm": 2.1670328689833043, + "learning_rate": 2.7336284517600053e-06, + "loss": 1.3787, + "step": 1784 + }, + { + "epoch": 1.3350785340314135, + "grad_norm": 2.3289483574484446, + "learning_rate": 2.728119904281105e-06, + "loss": 1.6332, + "step": 1785 + }, + { + "epoch": 1.3358264771877337, + "grad_norm": 2.3019488300330324, + "learning_rate": 2.7226148295701704e-06, + "loss": 1.2824, + "step": 1786 + }, + { + "epoch": 1.336574420344054, + "grad_norm": 1.8502529593852841, + "learning_rate": 2.7171132360421894e-06, + "loss": 1.681, + "step": 1787 + }, + { + "epoch": 1.337322363500374, + "grad_norm": 1.8069840740547785, + "learning_rate": 2.711615132106832e-06, + "loss": 1.7159, + "step": 1788 + }, + { + "epoch": 1.3380703066566941, + "grad_norm": 2.4537747355287247, + "learning_rate": 2.7061205261684307e-06, + "loss": 1.2363, + "step": 1789 + }, + { + "epoch": 1.3388182498130141, + "grad_norm": 2.0348916191513604, + "learning_rate": 2.7006294266259716e-06, + "loss": 1.4398, + "step": 1790 + }, + { + "epoch": 1.3395661929693343, + "grad_norm": 2.078270108839509, + "learning_rate": 2.6951418418730825e-06, + "loss": 1.2591, + "step": 1791 + }, + { + "epoch": 1.3403141361256545, + "grad_norm": 2.38942510154024, + "learning_rate": 2.689657780298019e-06, + "loss": 1.3316, + "step": 1792 + }, + { + "epoch": 1.3410620792819745, + "grad_norm": 2.0150126941175017, + "learning_rate": 2.6841772502836518e-06, + "loss": 1.5659, + "step": 1793 + }, + { + "epoch": 1.3418100224382947, + "grad_norm": 2.248481810496939, + "learning_rate": 2.678700260207449e-06, + "loss": 1.5626, + "step": 1794 + }, + { + "epoch": 1.3425579655946147, + "grad_norm": 3.0119038869722288, + "learning_rate": 2.6732268184414695e-06, + "loss": 1.3424, + "step": 1795 + }, + { + "epoch": 1.343305908750935, + "grad_norm": 2.0790275042133834, + "learning_rate": 2.667756933352352e-06, + "loss": 1.6456, + "step": 1796 + }, + { + "epoch": 1.3440538519072551, + "grad_norm": 1.936742075580005, + "learning_rate": 2.6622906133012926e-06, + "loss": 1.3556, + "step": 1797 + }, + { + "epoch": 1.3448017950635751, + "grad_norm": 2.0786269668174175, + "learning_rate": 2.6568278666440417e-06, + "loss": 1.5975, + "step": 1798 + }, + { + "epoch": 1.3455497382198953, + "grad_norm": 1.9697990741148197, + "learning_rate": 2.651368701730889e-06, + "loss": 1.5577, + "step": 1799 + }, + { + "epoch": 1.3462976813762153, + "grad_norm": 1.9949559827893473, + "learning_rate": 2.645913126906644e-06, + "loss": 1.5936, + "step": 1800 + }, + { + "epoch": 1.3470456245325355, + "grad_norm": 3.0282958135141933, + "learning_rate": 2.6404611505106337e-06, + "loss": 1.5092, + "step": 1801 + }, + { + "epoch": 1.3477935676888557, + "grad_norm": 2.3076951123272043, + "learning_rate": 2.6350127808766786e-06, + "loss": 1.8741, + "step": 1802 + }, + { + "epoch": 1.3485415108451757, + "grad_norm": 1.8603428307963321, + "learning_rate": 2.6295680263330936e-06, + "loss": 1.5627, + "step": 1803 + }, + { + "epoch": 1.349289454001496, + "grad_norm": 2.13472842597363, + "learning_rate": 2.6241268952026592e-06, + "loss": 1.2433, + "step": 1804 + }, + { + "epoch": 1.350037397157816, + "grad_norm": 2.235413445451521, + "learning_rate": 2.6186893958026245e-06, + "loss": 1.4329, + "step": 1805 + }, + { + "epoch": 1.350785340314136, + "grad_norm": 2.0530689114803677, + "learning_rate": 2.6132555364446856e-06, + "loss": 1.6219, + "step": 1806 + }, + { + "epoch": 1.3515332834704563, + "grad_norm": 2.108961349578589, + "learning_rate": 2.6078253254349706e-06, + "loss": 1.4341, + "step": 1807 + }, + { + "epoch": 1.3522812266267763, + "grad_norm": 2.4528545598280993, + "learning_rate": 2.6023987710740317e-06, + "loss": 1.5966, + "step": 1808 + }, + { + "epoch": 1.3530291697830965, + "grad_norm": 1.8649183200639123, + "learning_rate": 2.5969758816568367e-06, + "loss": 1.4782, + "step": 1809 + }, + { + "epoch": 1.3537771129394165, + "grad_norm": 2.1498606383837413, + "learning_rate": 2.591556665472744e-06, + "loss": 1.716, + "step": 1810 + }, + { + "epoch": 1.3545250560957367, + "grad_norm": 2.2762237106185115, + "learning_rate": 2.5861411308055038e-06, + "loss": 1.3742, + "step": 1811 + }, + { + "epoch": 1.355272999252057, + "grad_norm": 2.2052267739747684, + "learning_rate": 2.5807292859332354e-06, + "loss": 1.5475, + "step": 1812 + }, + { + "epoch": 1.356020942408377, + "grad_norm": 2.053216727644687, + "learning_rate": 2.5753211391284172e-06, + "loss": 1.4057, + "step": 1813 + }, + { + "epoch": 1.356768885564697, + "grad_norm": 4.798642846149934, + "learning_rate": 2.5699166986578773e-06, + "loss": 1.6018, + "step": 1814 + }, + { + "epoch": 1.357516828721017, + "grad_norm": 3.661433310983455, + "learning_rate": 2.564515972782774e-06, + "loss": 1.4399, + "step": 1815 + }, + { + "epoch": 1.3582647718773373, + "grad_norm": 1.9288554156053217, + "learning_rate": 2.559118969758595e-06, + "loss": 1.423, + "step": 1816 + }, + { + "epoch": 1.3590127150336575, + "grad_norm": 2.0046207010606483, + "learning_rate": 2.5537256978351276e-06, + "loss": 1.3176, + "step": 1817 + }, + { + "epoch": 1.3597606581899775, + "grad_norm": 2.0036708178220852, + "learning_rate": 2.548336165256463e-06, + "loss": 1.2276, + "step": 1818 + }, + { + "epoch": 1.3605086013462977, + "grad_norm": 2.0992150812950814, + "learning_rate": 2.5429503802609777e-06, + "loss": 1.3913, + "step": 1819 + }, + { + "epoch": 1.3612565445026177, + "grad_norm": 2.0532592714532103, + "learning_rate": 2.537568351081311e-06, + "loss": 1.2705, + "step": 1820 + }, + { + "epoch": 1.362004487658938, + "grad_norm": 2.4267292231391635, + "learning_rate": 2.5321900859443714e-06, + "loss": 1.3078, + "step": 1821 + }, + { + "epoch": 1.362752430815258, + "grad_norm": 2.4955603112297386, + "learning_rate": 2.526815593071306e-06, + "loss": 1.4022, + "step": 1822 + }, + { + "epoch": 1.363500373971578, + "grad_norm": 1.8518542130237632, + "learning_rate": 2.521444880677497e-06, + "loss": 1.4028, + "step": 1823 + }, + { + "epoch": 1.3642483171278983, + "grad_norm": 1.946867892052029, + "learning_rate": 2.5160779569725514e-06, + "loss": 1.7253, + "step": 1824 + }, + { + "epoch": 1.3649962602842183, + "grad_norm": 2.298424513651998, + "learning_rate": 2.5107148301602825e-06, + "loss": 1.601, + "step": 1825 + }, + { + "epoch": 1.3657442034405385, + "grad_norm": 2.0993672666151775, + "learning_rate": 2.5053555084387022e-06, + "loss": 1.3868, + "step": 1826 + }, + { + "epoch": 1.3664921465968587, + "grad_norm": 2.1388651581264866, + "learning_rate": 2.5000000000000015e-06, + "loss": 1.4429, + "step": 1827 + }, + { + "epoch": 1.3672400897531787, + "grad_norm": 2.3455010186018255, + "learning_rate": 2.494648313030542e-06, + "loss": 1.6724, + "step": 1828 + }, + { + "epoch": 1.3679880329094989, + "grad_norm": 1.9755795520153072, + "learning_rate": 2.4893004557108526e-06, + "loss": 1.7484, + "step": 1829 + }, + { + "epoch": 1.3687359760658189, + "grad_norm": 2.2377549673259662, + "learning_rate": 2.483956436215596e-06, + "loss": 1.3278, + "step": 1830 + }, + { + "epoch": 1.369483919222139, + "grad_norm": 2.654345607990908, + "learning_rate": 2.478616262713578e-06, + "loss": 1.4453, + "step": 1831 + }, + { + "epoch": 1.3702318623784593, + "grad_norm": 2.1406249489829, + "learning_rate": 2.4732799433677236e-06, + "loss": 1.4291, + "step": 1832 + }, + { + "epoch": 1.3709798055347795, + "grad_norm": 2.384341107235647, + "learning_rate": 2.467947486335061e-06, + "loss": 1.5006, + "step": 1833 + }, + { + "epoch": 1.3717277486910995, + "grad_norm": 2.1196230217643204, + "learning_rate": 2.4626188997667224e-06, + "loss": 1.5329, + "step": 1834 + }, + { + "epoch": 1.3724756918474195, + "grad_norm": 2.283506181406822, + "learning_rate": 2.4572941918079178e-06, + "loss": 1.6048, + "step": 1835 + }, + { + "epoch": 1.3732236350037397, + "grad_norm": 1.8162164011888162, + "learning_rate": 2.4519733705979283e-06, + "loss": 1.6128, + "step": 1836 + }, + { + "epoch": 1.3739715781600599, + "grad_norm": 2.151355764219005, + "learning_rate": 2.4466564442700974e-06, + "loss": 1.3818, + "step": 1837 + }, + { + "epoch": 1.37471952131638, + "grad_norm": 2.046437380311156, + "learning_rate": 2.4413434209518137e-06, + "loss": 1.6117, + "step": 1838 + }, + { + "epoch": 1.3754674644727, + "grad_norm": 3.1870675985151977, + "learning_rate": 2.436034308764501e-06, + "loss": 1.6306, + "step": 1839 + }, + { + "epoch": 1.37621540762902, + "grad_norm": 1.9788140583776086, + "learning_rate": 2.430729115823601e-06, + "loss": 1.537, + "step": 1840 + }, + { + "epoch": 1.3769633507853403, + "grad_norm": 2.2373759336063777, + "learning_rate": 2.425427850238565e-06, + "loss": 0.9794, + "step": 1841 + }, + { + "epoch": 1.3777112939416605, + "grad_norm": 2.202170228692596, + "learning_rate": 2.4201305201128444e-06, + "loss": 1.613, + "step": 1842 + }, + { + "epoch": 1.3784592370979807, + "grad_norm": 1.9628087633237494, + "learning_rate": 2.414837133543871e-06, + "loss": 1.487, + "step": 1843 + }, + { + "epoch": 1.3792071802543007, + "grad_norm": 2.430030605785805, + "learning_rate": 2.4095476986230516e-06, + "loss": 1.6484, + "step": 1844 + }, + { + "epoch": 1.3799551234106207, + "grad_norm": 1.9540861720799736, + "learning_rate": 2.4042622234357526e-06, + "loss": 1.5248, + "step": 1845 + }, + { + "epoch": 1.3807030665669409, + "grad_norm": 2.4403033639207927, + "learning_rate": 2.398980716061284e-06, + "loss": 1.4251, + "step": 1846 + }, + { + "epoch": 1.381451009723261, + "grad_norm": 2.1222729868219274, + "learning_rate": 2.393703184572895e-06, + "loss": 1.4327, + "step": 1847 + }, + { + "epoch": 1.3821989528795813, + "grad_norm": 2.464842490422499, + "learning_rate": 2.388429637037753e-06, + "loss": 1.3014, + "step": 1848 + }, + { + "epoch": 1.3829468960359013, + "grad_norm": 2.9388537821098106, + "learning_rate": 2.383160081516941e-06, + "loss": 1.5635, + "step": 1849 + }, + { + "epoch": 1.3836948391922215, + "grad_norm": 2.3004959620206553, + "learning_rate": 2.377894526065433e-06, + "loss": 1.3158, + "step": 1850 + }, + { + "epoch": 1.3844427823485415, + "grad_norm": 2.000232586213384, + "learning_rate": 2.3726329787320943e-06, + "loss": 1.6238, + "step": 1851 + }, + { + "epoch": 1.3851907255048617, + "grad_norm": 2.0793155245970225, + "learning_rate": 2.3673754475596634e-06, + "loss": 1.4936, + "step": 1852 + }, + { + "epoch": 1.3859386686611819, + "grad_norm": 3.4756392902504962, + "learning_rate": 2.3621219405847333e-06, + "loss": 1.4729, + "step": 1853 + }, + { + "epoch": 1.3866866118175019, + "grad_norm": 1.9821115133948681, + "learning_rate": 2.356872465837755e-06, + "loss": 1.1649, + "step": 1854 + }, + { + "epoch": 1.387434554973822, + "grad_norm": 2.1051322436884874, + "learning_rate": 2.3516270313430085e-06, + "loss": 1.5109, + "step": 1855 + }, + { + "epoch": 1.388182498130142, + "grad_norm": 2.09710297440993, + "learning_rate": 2.3463856451185986e-06, + "loss": 1.5303, + "step": 1856 + }, + { + "epoch": 1.3889304412864623, + "grad_norm": 2.015508338659588, + "learning_rate": 2.341148315176446e-06, + "loss": 1.5523, + "step": 1857 + }, + { + "epoch": 1.3896783844427825, + "grad_norm": 2.0645925119779984, + "learning_rate": 2.335915049522268e-06, + "loss": 1.7927, + "step": 1858 + }, + { + "epoch": 1.3904263275991025, + "grad_norm": 2.074106947223502, + "learning_rate": 2.3306858561555728e-06, + "loss": 1.6931, + "step": 1859 + }, + { + "epoch": 1.3911742707554227, + "grad_norm": 1.8274549179212949, + "learning_rate": 2.3254607430696393e-06, + "loss": 1.5687, + "step": 1860 + }, + { + "epoch": 1.3919222139117426, + "grad_norm": 2.0016264822558902, + "learning_rate": 2.320239718251508e-06, + "loss": 1.542, + "step": 1861 + }, + { + "epoch": 1.3926701570680629, + "grad_norm": 2.4381329710387374, + "learning_rate": 2.3150227896819782e-06, + "loss": 1.4198, + "step": 1862 + }, + { + "epoch": 1.393418100224383, + "grad_norm": 3.1359037131836662, + "learning_rate": 2.3098099653355804e-06, + "loss": 1.6044, + "step": 1863 + }, + { + "epoch": 1.394166043380703, + "grad_norm": 2.4240265995051633, + "learning_rate": 2.30460125318057e-06, + "loss": 1.476, + "step": 1864 + }, + { + "epoch": 1.3949139865370233, + "grad_norm": 1.8726943664346882, + "learning_rate": 2.2993966611789277e-06, + "loss": 1.6239, + "step": 1865 + }, + { + "epoch": 1.3956619296933432, + "grad_norm": 1.7986926618834298, + "learning_rate": 2.294196197286323e-06, + "loss": 1.651, + "step": 1866 + }, + { + "epoch": 1.3964098728496634, + "grad_norm": 2.0047941241984164, + "learning_rate": 2.2889998694521257e-06, + "loss": 1.589, + "step": 1867 + }, + { + "epoch": 1.3971578160059837, + "grad_norm": 2.1204169227107754, + "learning_rate": 2.2838076856193758e-06, + "loss": 1.3507, + "step": 1868 + }, + { + "epoch": 1.3979057591623036, + "grad_norm": 2.1007055170758755, + "learning_rate": 2.278619653724781e-06, + "loss": 1.6346, + "step": 1869 + }, + { + "epoch": 1.3986537023186238, + "grad_norm": 2.2396804223821416, + "learning_rate": 2.2734357816987046e-06, + "loss": 1.447, + "step": 1870 + }, + { + "epoch": 1.3994016454749438, + "grad_norm": 2.306493043690132, + "learning_rate": 2.2682560774651458e-06, + "loss": 1.3574, + "step": 1871 + }, + { + "epoch": 1.400149588631264, + "grad_norm": 3.2587616950010956, + "learning_rate": 2.2630805489417437e-06, + "loss": 1.826, + "step": 1872 + }, + { + "epoch": 1.4008975317875842, + "grad_norm": 1.963964535707808, + "learning_rate": 2.2579092040397436e-06, + "loss": 1.5968, + "step": 1873 + }, + { + "epoch": 1.4016454749439042, + "grad_norm": 1.9719654631516332, + "learning_rate": 2.252742050663998e-06, + "loss": 1.5723, + "step": 1874 + }, + { + "epoch": 1.4023934181002244, + "grad_norm": 2.0915152122514566, + "learning_rate": 2.247579096712958e-06, + "loss": 1.5192, + "step": 1875 + }, + { + "epoch": 1.4031413612565444, + "grad_norm": 1.717013523263439, + "learning_rate": 2.2424203500786473e-06, + "loss": 1.413, + "step": 1876 + }, + { + "epoch": 1.4038893044128646, + "grad_norm": 2.788440285418151, + "learning_rate": 2.2372658186466665e-06, + "loss": 1.767, + "step": 1877 + }, + { + "epoch": 1.4046372475691848, + "grad_norm": 2.0055975071446754, + "learning_rate": 2.232115510296165e-06, + "loss": 1.5107, + "step": 1878 + }, + { + "epoch": 1.4053851907255048, + "grad_norm": 2.2163128564286447, + "learning_rate": 2.226969432899843e-06, + "loss": 1.8729, + "step": 1879 + }, + { + "epoch": 1.406133133881825, + "grad_norm": 2.509659242581342, + "learning_rate": 2.2218275943239333e-06, + "loss": 1.6225, + "step": 1880 + }, + { + "epoch": 1.406881077038145, + "grad_norm": 1.9328854394405115, + "learning_rate": 2.2166900024281818e-06, + "loss": 1.5886, + "step": 1881 + }, + { + "epoch": 1.4076290201944652, + "grad_norm": 2.0152482615177583, + "learning_rate": 2.211556665065854e-06, + "loss": 1.6197, + "step": 1882 + }, + { + "epoch": 1.4083769633507854, + "grad_norm": 2.282483693274865, + "learning_rate": 2.206427590083703e-06, + "loss": 1.5908, + "step": 1883 + }, + { + "epoch": 1.4091249065071054, + "grad_norm": 2.3515501480321173, + "learning_rate": 2.201302785321968e-06, + "loss": 1.7414, + "step": 1884 + }, + { + "epoch": 1.4098728496634256, + "grad_norm": 1.928427549281249, + "learning_rate": 2.196182258614365e-06, + "loss": 1.534, + "step": 1885 + }, + { + "epoch": 1.4106207928197456, + "grad_norm": 2.900177254510004, + "learning_rate": 2.1910660177880675e-06, + "loss": 1.4838, + "step": 1886 + }, + { + "epoch": 1.4113687359760658, + "grad_norm": 2.405309655302665, + "learning_rate": 2.1859540706637007e-06, + "loss": 1.5117, + "step": 1887 + }, + { + "epoch": 1.412116679132386, + "grad_norm": 3.1217428343180287, + "learning_rate": 2.180846425055322e-06, + "loss": 1.3546, + "step": 1888 + }, + { + "epoch": 1.412864622288706, + "grad_norm": 2.012876363264464, + "learning_rate": 2.1757430887704135e-06, + "loss": 1.5483, + "step": 1889 + }, + { + "epoch": 1.4136125654450262, + "grad_norm": 2.349515800189081, + "learning_rate": 2.170644069609876e-06, + "loss": 1.3275, + "step": 1890 + }, + { + "epoch": 1.4143605086013462, + "grad_norm": 2.029927743704215, + "learning_rate": 2.165549375368004e-06, + "loss": 1.5241, + "step": 1891 + }, + { + "epoch": 1.4151084517576664, + "grad_norm": 3.53511554352164, + "learning_rate": 2.1604590138324866e-06, + "loss": 1.4204, + "step": 1892 + }, + { + "epoch": 1.4158563949139866, + "grad_norm": 2.4319717557708223, + "learning_rate": 2.1553729927843894e-06, + "loss": 1.2527, + "step": 1893 + }, + { + "epoch": 1.4166043380703066, + "grad_norm": 1.975371276566408, + "learning_rate": 2.150291319998137e-06, + "loss": 1.8591, + "step": 1894 + }, + { + "epoch": 1.4173522812266268, + "grad_norm": 2.1451330008097957, + "learning_rate": 2.1452140032415174e-06, + "loss": 1.3388, + "step": 1895 + }, + { + "epoch": 1.4181002243829468, + "grad_norm": 2.1686945294856694, + "learning_rate": 2.1401410502756513e-06, + "loss": 1.6217, + "step": 1896 + }, + { + "epoch": 1.418848167539267, + "grad_norm": 2.0525326302505955, + "learning_rate": 2.1350724688549906e-06, + "loss": 1.6787, + "step": 1897 + }, + { + "epoch": 1.4195961106955872, + "grad_norm": 2.155931156334961, + "learning_rate": 2.1300082667273087e-06, + "loss": 1.4561, + "step": 1898 + }, + { + "epoch": 1.4203440538519072, + "grad_norm": 2.2705711465685425, + "learning_rate": 2.1249484516336826e-06, + "loss": 1.6089, + "step": 1899 + }, + { + "epoch": 1.4210919970082274, + "grad_norm": 1.8367579578262165, + "learning_rate": 2.119893031308485e-06, + "loss": 1.2879, + "step": 1900 + }, + { + "epoch": 1.4218399401645474, + "grad_norm": 2.080213481776387, + "learning_rate": 2.114842013479367e-06, + "loss": 1.5011, + "step": 1901 + }, + { + "epoch": 1.4225878833208676, + "grad_norm": 2.508038563942065, + "learning_rate": 2.1097954058672515e-06, + "loss": 1.5325, + "step": 1902 + }, + { + "epoch": 1.4233358264771878, + "grad_norm": 2.2174384912191814, + "learning_rate": 2.1047532161863242e-06, + "loss": 1.6039, + "step": 1903 + }, + { + "epoch": 1.4240837696335078, + "grad_norm": 2.17105145841857, + "learning_rate": 2.09971545214401e-06, + "loss": 1.6393, + "step": 1904 + }, + { + "epoch": 1.424831712789828, + "grad_norm": 2.134575023441808, + "learning_rate": 2.094682121440975e-06, + "loss": 1.6525, + "step": 1905 + }, + { + "epoch": 1.425579655946148, + "grad_norm": 1.9678855242282483, + "learning_rate": 2.0896532317711087e-06, + "loss": 1.6554, + "step": 1906 + }, + { + "epoch": 1.4263275991024682, + "grad_norm": 2.4832300418249296, + "learning_rate": 2.0846287908215063e-06, + "loss": 1.5014, + "step": 1907 + }, + { + "epoch": 1.4270755422587884, + "grad_norm": 1.9677230316258774, + "learning_rate": 2.0796088062724697e-06, + "loss": 1.7481, + "step": 1908 + }, + { + "epoch": 1.4278234854151084, + "grad_norm": 2.014368872996576, + "learning_rate": 2.0745932857974825e-06, + "loss": 1.4575, + "step": 1909 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 2.38436878398276, + "learning_rate": 2.0695822370632106e-06, + "loss": 1.251, + "step": 1910 + }, + { + "epoch": 1.4293193717277486, + "grad_norm": 1.9199140674902129, + "learning_rate": 2.0645756677294788e-06, + "loss": 1.3212, + "step": 1911 + }, + { + "epoch": 1.4300673148840688, + "grad_norm": 2.4025578947281767, + "learning_rate": 2.0595735854492675e-06, + "loss": 1.5114, + "step": 1912 + }, + { + "epoch": 1.430815258040389, + "grad_norm": 2.218898999825128, + "learning_rate": 2.054575997868702e-06, + "loss": 1.3986, + "step": 1913 + }, + { + "epoch": 1.431563201196709, + "grad_norm": 2.872267734317338, + "learning_rate": 2.0495829126270277e-06, + "loss": 1.4441, + "step": 1914 + }, + { + "epoch": 1.4323111443530292, + "grad_norm": 2.534863554426513, + "learning_rate": 2.0445943373566178e-06, + "loss": 1.2627, + "step": 1915 + }, + { + "epoch": 1.4330590875093492, + "grad_norm": 2.5257980712001795, + "learning_rate": 2.0396102796829458e-06, + "loss": 1.5497, + "step": 1916 + }, + { + "epoch": 1.4338070306656694, + "grad_norm": 2.304487816494637, + "learning_rate": 2.034630747224578e-06, + "loss": 1.5618, + "step": 1917 + }, + { + "epoch": 1.4345549738219896, + "grad_norm": 1.9904958289049852, + "learning_rate": 2.029655747593169e-06, + "loss": 1.5124, + "step": 1918 + }, + { + "epoch": 1.4353029169783096, + "grad_norm": 1.9795399019835245, + "learning_rate": 2.0246852883934437e-06, + "loss": 1.612, + "step": 1919 + }, + { + "epoch": 1.4360508601346298, + "grad_norm": 1.8820570527925915, + "learning_rate": 2.019719377223181e-06, + "loss": 1.6145, + "step": 1920 + }, + { + "epoch": 1.4367988032909498, + "grad_norm": 2.535354137840176, + "learning_rate": 2.014758021673216e-06, + "loss": 1.3507, + "step": 1921 + }, + { + "epoch": 1.43754674644727, + "grad_norm": 2.081178810825882, + "learning_rate": 2.009801229327411e-06, + "loss": 1.6617, + "step": 1922 + }, + { + "epoch": 1.4382946896035902, + "grad_norm": 1.9432332786690594, + "learning_rate": 2.0048490077626624e-06, + "loss": 1.7644, + "step": 1923 + }, + { + "epoch": 1.4390426327599102, + "grad_norm": 1.9098860877138484, + "learning_rate": 1.9999013645488714e-06, + "loss": 1.5041, + "step": 1924 + }, + { + "epoch": 1.4397905759162304, + "grad_norm": 1.8846901275622205, + "learning_rate": 1.9949583072489455e-06, + "loss": 1.404, + "step": 1925 + }, + { + "epoch": 1.4405385190725504, + "grad_norm": 2.4287697079175965, + "learning_rate": 1.9900198434187838e-06, + "loss": 1.5515, + "step": 1926 + }, + { + "epoch": 1.4412864622288706, + "grad_norm": 1.9824440570562907, + "learning_rate": 1.9850859806072576e-06, + "loss": 1.6523, + "step": 1927 + }, + { + "epoch": 1.4420344053851908, + "grad_norm": 2.3350738189483926, + "learning_rate": 1.9801567263562126e-06, + "loss": 1.5465, + "step": 1928 + }, + { + "epoch": 1.4427823485415108, + "grad_norm": 2.561718589943007, + "learning_rate": 1.9752320882004445e-06, + "loss": 1.493, + "step": 1929 + }, + { + "epoch": 1.443530291697831, + "grad_norm": 2.0490317447031443, + "learning_rate": 1.970312073667693e-06, + "loss": 1.4748, + "step": 1930 + }, + { + "epoch": 1.444278234854151, + "grad_norm": 2.247093885150452, + "learning_rate": 1.9653966902786327e-06, + "loss": 1.3435, + "step": 1931 + }, + { + "epoch": 1.4450261780104712, + "grad_norm": 2.1030640572785475, + "learning_rate": 1.9604859455468587e-06, + "loss": 1.5336, + "step": 1932 + }, + { + "epoch": 1.4457741211667914, + "grad_norm": 1.9206027327693225, + "learning_rate": 1.9555798469788773e-06, + "loss": 1.4653, + "step": 1933 + }, + { + "epoch": 1.4465220643231114, + "grad_norm": 2.208260553964317, + "learning_rate": 1.9506784020740875e-06, + "loss": 1.3762, + "step": 1934 + }, + { + "epoch": 1.4472700074794316, + "grad_norm": 2.0486657755639777, + "learning_rate": 1.9457816183247767e-06, + "loss": 1.4418, + "step": 1935 + }, + { + "epoch": 1.4480179506357516, + "grad_norm": 2.5081014859888042, + "learning_rate": 1.9408895032161116e-06, + "loss": 1.376, + "step": 1936 + }, + { + "epoch": 1.4487658937920718, + "grad_norm": 1.912936341003927, + "learning_rate": 1.9360020642261155e-06, + "loss": 1.5419, + "step": 1937 + }, + { + "epoch": 1.449513836948392, + "grad_norm": 2.1026093805876096, + "learning_rate": 1.93111930882567e-06, + "loss": 1.5391, + "step": 1938 + }, + { + "epoch": 1.450261780104712, + "grad_norm": 2.3592518068469364, + "learning_rate": 1.926241244478496e-06, + "loss": 1.3805, + "step": 1939 + }, + { + "epoch": 1.4510097232610322, + "grad_norm": 2.1885925267875765, + "learning_rate": 1.9213678786411392e-06, + "loss": 1.4269, + "step": 1940 + }, + { + "epoch": 1.4517576664173522, + "grad_norm": 2.210552397252292, + "learning_rate": 1.9164992187629705e-06, + "loss": 1.3821, + "step": 1941 + }, + { + "epoch": 1.4525056095736724, + "grad_norm": 2.41944980228632, + "learning_rate": 1.9116352722861596e-06, + "loss": 1.3473, + "step": 1942 + }, + { + "epoch": 1.4532535527299926, + "grad_norm": 2.057516218373771, + "learning_rate": 1.906776046645679e-06, + "loss": 1.6038, + "step": 1943 + }, + { + "epoch": 1.4540014958863128, + "grad_norm": 2.4291228076156073, + "learning_rate": 1.9019215492692778e-06, + "loss": 1.6245, + "step": 1944 + }, + { + "epoch": 1.4547494390426328, + "grad_norm": 2.0781355930360963, + "learning_rate": 1.897071787577482e-06, + "loss": 1.5747, + "step": 1945 + }, + { + "epoch": 1.4554973821989527, + "grad_norm": 2.3214114644183375, + "learning_rate": 1.8922267689835806e-06, + "loss": 1.3999, + "step": 1946 + }, + { + "epoch": 1.456245325355273, + "grad_norm": 2.4665154360354595, + "learning_rate": 1.8873865008936065e-06, + "loss": 1.7056, + "step": 1947 + }, + { + "epoch": 1.4569932685115932, + "grad_norm": 2.069509724846586, + "learning_rate": 1.8825509907063328e-06, + "loss": 1.2437, + "step": 1948 + }, + { + "epoch": 1.4577412116679134, + "grad_norm": 2.576206192081514, + "learning_rate": 1.8777202458132644e-06, + "loss": 1.5328, + "step": 1949 + }, + { + "epoch": 1.4584891548242334, + "grad_norm": 1.9592319416118262, + "learning_rate": 1.8728942735986149e-06, + "loss": 1.5207, + "step": 1950 + }, + { + "epoch": 1.4592370979805533, + "grad_norm": 2.0691084440826764, + "learning_rate": 1.8680730814393084e-06, + "loss": 1.3262, + "step": 1951 + }, + { + "epoch": 1.4599850411368736, + "grad_norm": 2.2944642818521457, + "learning_rate": 1.8632566767049603e-06, + "loss": 1.6698, + "step": 1952 + }, + { + "epoch": 1.4607329842931938, + "grad_norm": 2.2308949316113025, + "learning_rate": 1.8584450667578656e-06, + "loss": 1.7158, + "step": 1953 + }, + { + "epoch": 1.461480927449514, + "grad_norm": 2.1721731437723593, + "learning_rate": 1.8536382589529944e-06, + "loss": 1.7328, + "step": 1954 + }, + { + "epoch": 1.462228870605834, + "grad_norm": 2.0274820291693287, + "learning_rate": 1.8488362606379713e-06, + "loss": 1.6915, + "step": 1955 + }, + { + "epoch": 1.462976813762154, + "grad_norm": 2.1989227047510997, + "learning_rate": 1.844039079153075e-06, + "loss": 1.5153, + "step": 1956 + }, + { + "epoch": 1.4637247569184741, + "grad_norm": 2.044549306906558, + "learning_rate": 1.839246721831215e-06, + "loss": 1.4015, + "step": 1957 + }, + { + "epoch": 1.4644727000747944, + "grad_norm": 2.049261981044707, + "learning_rate": 1.8344591959979302e-06, + "loss": 1.6035, + "step": 1958 + }, + { + "epoch": 1.4652206432311146, + "grad_norm": 2.7749350741498846, + "learning_rate": 1.829676508971377e-06, + "loss": 1.3278, + "step": 1959 + }, + { + "epoch": 1.4659685863874345, + "grad_norm": 2.518743386133657, + "learning_rate": 1.8248986680623077e-06, + "loss": 1.2719, + "step": 1960 + }, + { + "epoch": 1.4667165295437548, + "grad_norm": 2.535489609495528, + "learning_rate": 1.8201256805740752e-06, + "loss": 1.4722, + "step": 1961 + }, + { + "epoch": 1.4674644727000747, + "grad_norm": 1.9409369437444248, + "learning_rate": 1.8153575538026064e-06, + "loss": 1.5208, + "step": 1962 + }, + { + "epoch": 1.468212415856395, + "grad_norm": 2.163196924007279, + "learning_rate": 1.8105942950364008e-06, + "loss": 1.3582, + "step": 1963 + }, + { + "epoch": 1.4689603590127152, + "grad_norm": 2.3311309689838686, + "learning_rate": 1.805835911556519e-06, + "loss": 1.5527, + "step": 1964 + }, + { + "epoch": 1.4697083021690351, + "grad_norm": 1.981444441055877, + "learning_rate": 1.801082410636566e-06, + "loss": 1.6182, + "step": 1965 + }, + { + "epoch": 1.4704562453253553, + "grad_norm": 1.9076930128965328, + "learning_rate": 1.7963337995426876e-06, + "loss": 1.5364, + "step": 1966 + }, + { + "epoch": 1.4712041884816753, + "grad_norm": 2.1932920653921486, + "learning_rate": 1.7915900855335506e-06, + "loss": 1.5603, + "step": 1967 + }, + { + "epoch": 1.4719521316379955, + "grad_norm": 2.1110241650593076, + "learning_rate": 1.7868512758603362e-06, + "loss": 1.6285, + "step": 1968 + }, + { + "epoch": 1.4727000747943157, + "grad_norm": 3.265038547666587, + "learning_rate": 1.7821173777667333e-06, + "loss": 1.5625, + "step": 1969 + }, + { + "epoch": 1.4734480179506357, + "grad_norm": 2.089459825608497, + "learning_rate": 1.7773883984889178e-06, + "loss": 1.5548, + "step": 1970 + }, + { + "epoch": 1.474195961106956, + "grad_norm": 2.1535980282546374, + "learning_rate": 1.7726643452555504e-06, + "loss": 1.4189, + "step": 1971 + }, + { + "epoch": 1.474943904263276, + "grad_norm": 2.292470107115746, + "learning_rate": 1.7679452252877622e-06, + "loss": 1.4304, + "step": 1972 + }, + { + "epoch": 1.4756918474195961, + "grad_norm": 2.1622013668345126, + "learning_rate": 1.7632310457991387e-06, + "loss": 1.4352, + "step": 1973 + }, + { + "epoch": 1.4764397905759163, + "grad_norm": 2.1647387852793325, + "learning_rate": 1.7585218139957205e-06, + "loss": 1.5877, + "step": 1974 + }, + { + "epoch": 1.4771877337322363, + "grad_norm": 2.5175504910159217, + "learning_rate": 1.7538175370759797e-06, + "loss": 1.6657, + "step": 1975 + }, + { + "epoch": 1.4779356768885565, + "grad_norm": 1.8835114656958258, + "learning_rate": 1.7491182222308139e-06, + "loss": 1.387, + "step": 1976 + }, + { + "epoch": 1.4786836200448765, + "grad_norm": 2.042540829318312, + "learning_rate": 1.74442387664354e-06, + "loss": 1.4949, + "step": 1977 + }, + { + "epoch": 1.4794315632011967, + "grad_norm": 2.174863510658964, + "learning_rate": 1.739734507489877e-06, + "loss": 1.5079, + "step": 1978 + }, + { + "epoch": 1.480179506357517, + "grad_norm": 2.1261306261400765, + "learning_rate": 1.7350501219379384e-06, + "loss": 1.4513, + "step": 1979 + }, + { + "epoch": 1.480927449513837, + "grad_norm": 2.131251344322679, + "learning_rate": 1.7303707271482161e-06, + "loss": 1.4842, + "step": 1980 + }, + { + "epoch": 1.4816753926701571, + "grad_norm": 2.1352810139246854, + "learning_rate": 1.7256963302735752e-06, + "loss": 1.4929, + "step": 1981 + }, + { + "epoch": 1.4824233358264771, + "grad_norm": 2.4443903323264786, + "learning_rate": 1.7210269384592433e-06, + "loss": 1.2187, + "step": 1982 + }, + { + "epoch": 1.4831712789827973, + "grad_norm": 2.087458139976212, + "learning_rate": 1.7163625588427918e-06, + "loss": 1.687, + "step": 1983 + }, + { + "epoch": 1.4839192221391175, + "grad_norm": 2.177170094782141, + "learning_rate": 1.7117031985541366e-06, + "loss": 1.7355, + "step": 1984 + }, + { + "epoch": 1.4846671652954375, + "grad_norm": 2.5395262713072326, + "learning_rate": 1.7070488647155204e-06, + "loss": 1.5756, + "step": 1985 + }, + { + "epoch": 1.4854151084517577, + "grad_norm": 2.0899767137036895, + "learning_rate": 1.702399564441496e-06, + "loss": 1.6573, + "step": 1986 + }, + { + "epoch": 1.4861630516080777, + "grad_norm": 2.331064119486473, + "learning_rate": 1.6977553048389306e-06, + "loss": 1.3597, + "step": 1987 + }, + { + "epoch": 1.486910994764398, + "grad_norm": 2.5498923612749085, + "learning_rate": 1.6931160930069789e-06, + "loss": 1.214, + "step": 1988 + }, + { + "epoch": 1.4876589379207181, + "grad_norm": 2.088366747906741, + "learning_rate": 1.6884819360370868e-06, + "loss": 1.3994, + "step": 1989 + }, + { + "epoch": 1.4884068810770381, + "grad_norm": 2.3057259842257944, + "learning_rate": 1.6838528410129662e-06, + "loss": 1.6883, + "step": 1990 + }, + { + "epoch": 1.4891548242333583, + "grad_norm": 1.9876318429686965, + "learning_rate": 1.6792288150105957e-06, + "loss": 1.6244, + "step": 1991 + }, + { + "epoch": 1.4899027673896783, + "grad_norm": 2.2788041735314644, + "learning_rate": 1.6746098650982072e-06, + "loss": 1.6892, + "step": 1992 + }, + { + "epoch": 1.4906507105459985, + "grad_norm": 2.113567308720945, + "learning_rate": 1.6699959983362668e-06, + "loss": 1.4739, + "step": 1993 + }, + { + "epoch": 1.4913986537023187, + "grad_norm": 2.04802647324031, + "learning_rate": 1.665387221777478e-06, + "loss": 1.5985, + "step": 1994 + }, + { + "epoch": 1.4921465968586387, + "grad_norm": 1.949724994924677, + "learning_rate": 1.6607835424667578e-06, + "loss": 1.4634, + "step": 1995 + }, + { + "epoch": 1.492894540014959, + "grad_norm": 1.8145129264667208, + "learning_rate": 1.6561849674412327e-06, + "loss": 1.4097, + "step": 1996 + }, + { + "epoch": 1.493642483171279, + "grad_norm": 2.0764177307094185, + "learning_rate": 1.6515915037302298e-06, + "loss": 1.6607, + "step": 1997 + }, + { + "epoch": 1.494390426327599, + "grad_norm": 3.3339006757603493, + "learning_rate": 1.647003158355262e-06, + "loss": 1.1687, + "step": 1998 + }, + { + "epoch": 1.4951383694839193, + "grad_norm": 2.1581386604854873, + "learning_rate": 1.6424199383300149e-06, + "loss": 1.483, + "step": 1999 + }, + { + "epoch": 1.4958863126402393, + "grad_norm": 2.3546452976973606, + "learning_rate": 1.6378418506603456e-06, + "loss": 1.1961, + "step": 2000 + }, + { + "epoch": 1.4966342557965595, + "grad_norm": 1.9499859694179191, + "learning_rate": 1.633268902344259e-06, + "loss": 1.4743, + "step": 2001 + }, + { + "epoch": 1.4973821989528795, + "grad_norm": 2.023067693487027, + "learning_rate": 1.6287011003719105e-06, + "loss": 1.5234, + "step": 2002 + }, + { + "epoch": 1.4981301421091997, + "grad_norm": 2.3150949618565804, + "learning_rate": 1.6241384517255854e-06, + "loss": 1.6006, + "step": 2003 + }, + { + "epoch": 1.49887808526552, + "grad_norm": 1.957795897184621, + "learning_rate": 1.6195809633796889e-06, + "loss": 1.3778, + "step": 2004 + }, + { + "epoch": 1.49962602842184, + "grad_norm": 2.1053635070509364, + "learning_rate": 1.6150286423007472e-06, + "loss": 1.4287, + "step": 2005 + }, + { + "epoch": 1.50037397157816, + "grad_norm": 2.074055188600626, + "learning_rate": 1.610481495447378e-06, + "loss": 1.5427, + "step": 2006 + }, + { + "epoch": 1.50112191473448, + "grad_norm": 2.1849767946220657, + "learning_rate": 1.6059395297702967e-06, + "loss": 1.6621, + "step": 2007 + }, + { + "epoch": 1.5018698578908003, + "grad_norm": 2.0256878149878945, + "learning_rate": 1.601402752212296e-06, + "loss": 1.8033, + "step": 2008 + }, + { + "epoch": 1.5026178010471205, + "grad_norm": 2.3786505246298506, + "learning_rate": 1.596871169708235e-06, + "loss": 1.5411, + "step": 2009 + }, + { + "epoch": 1.5033657442034405, + "grad_norm": 2.569578467448889, + "learning_rate": 1.5923447891850396e-06, + "loss": 1.4687, + "step": 2010 + }, + { + "epoch": 1.5033657442034405, + "eval_loss": 1.669525384902954, + "eval_runtime": 68.8275, + "eval_samples_per_second": 59.177, + "eval_steps_per_second": 3.705, + "step": 2010 + }, + { + "epoch": 1.5041136873597607, + "grad_norm": 2.01529732458853, + "learning_rate": 1.5878236175616736e-06, + "loss": 1.4803, + "step": 2011 + }, + { + "epoch": 1.5048616305160807, + "grad_norm": 2.0770225171395755, + "learning_rate": 1.583307661749151e-06, + "loss": 1.6217, + "step": 2012 + }, + { + "epoch": 1.505609573672401, + "grad_norm": 1.9037020366701678, + "learning_rate": 1.5787969286505028e-06, + "loss": 1.4308, + "step": 2013 + }, + { + "epoch": 1.506357516828721, + "grad_norm": 2.162458104127453, + "learning_rate": 1.5742914251607794e-06, + "loss": 1.4533, + "step": 2014 + }, + { + "epoch": 1.507105459985041, + "grad_norm": 2.264666692791258, + "learning_rate": 1.56979115816704e-06, + "loss": 1.6626, + "step": 2015 + }, + { + "epoch": 1.5078534031413613, + "grad_norm": 1.9313263647199042, + "learning_rate": 1.5652961345483353e-06, + "loss": 1.3014, + "step": 2016 + }, + { + "epoch": 1.5086013462976813, + "grad_norm": 2.078364718059357, + "learning_rate": 1.5608063611757058e-06, + "loss": 1.7025, + "step": 2017 + }, + { + "epoch": 1.5093492894540015, + "grad_norm": 1.9482336559075386, + "learning_rate": 1.5563218449121603e-06, + "loss": 1.4035, + "step": 2018 + }, + { + "epoch": 1.5100972326103217, + "grad_norm": 2.392669942324436, + "learning_rate": 1.5518425926126774e-06, + "loss": 1.6509, + "step": 2019 + }, + { + "epoch": 1.510845175766642, + "grad_norm": 2.6826737025612446, + "learning_rate": 1.5473686111241881e-06, + "loss": 1.4395, + "step": 2020 + }, + { + "epoch": 1.5115931189229619, + "grad_norm": 2.121261560580008, + "learning_rate": 1.5428999072855626e-06, + "loss": 1.6871, + "step": 2021 + }, + { + "epoch": 1.5123410620792819, + "grad_norm": 2.5501559544045302, + "learning_rate": 1.5384364879276104e-06, + "loss": 1.8164, + "step": 2022 + }, + { + "epoch": 1.513089005235602, + "grad_norm": 2.13283113380617, + "learning_rate": 1.5339783598730568e-06, + "loss": 1.4784, + "step": 2023 + }, + { + "epoch": 1.5138369483919223, + "grad_norm": 1.9623361442771394, + "learning_rate": 1.5295255299365397e-06, + "loss": 1.6706, + "step": 2024 + }, + { + "epoch": 1.5145848915482425, + "grad_norm": 2.010763217670828, + "learning_rate": 1.5250780049246028e-06, + "loss": 1.5295, + "step": 2025 + }, + { + "epoch": 1.5153328347045625, + "grad_norm": 2.1141170432288217, + "learning_rate": 1.5206357916356796e-06, + "loss": 1.5146, + "step": 2026 + }, + { + "epoch": 1.5160807778608825, + "grad_norm": 2.1956650340873995, + "learning_rate": 1.5161988968600787e-06, + "loss": 1.4436, + "step": 2027 + }, + { + "epoch": 1.5168287210172027, + "grad_norm": 1.9537099580791772, + "learning_rate": 1.5117673273799866e-06, + "loss": 1.6963, + "step": 2028 + }, + { + "epoch": 1.5175766641735229, + "grad_norm": 4.673456221889441, + "learning_rate": 1.507341089969443e-06, + "loss": 1.5494, + "step": 2029 + }, + { + "epoch": 1.518324607329843, + "grad_norm": 1.9547747265176358, + "learning_rate": 1.5029201913943425e-06, + "loss": 1.6501, + "step": 2030 + }, + { + "epoch": 1.519072550486163, + "grad_norm": 1.9777013384651896, + "learning_rate": 1.498504638412413e-06, + "loss": 1.5147, + "step": 2031 + }, + { + "epoch": 1.519820493642483, + "grad_norm": 2.1410739027181034, + "learning_rate": 1.4940944377732168e-06, + "loss": 1.417, + "step": 2032 + }, + { + "epoch": 1.5205684367988033, + "grad_norm": 2.449901538235843, + "learning_rate": 1.4896895962181335e-06, + "loss": 1.5553, + "step": 2033 + }, + { + "epoch": 1.5213163799551235, + "grad_norm": 2.465846558275372, + "learning_rate": 1.485290120480346e-06, + "loss": 1.395, + "step": 2034 + }, + { + "epoch": 1.5220643231114437, + "grad_norm": 2.1914822912544434, + "learning_rate": 1.480896017284843e-06, + "loss": 1.3996, + "step": 2035 + }, + { + "epoch": 1.5228122662677637, + "grad_norm": 1.922468489283271, + "learning_rate": 1.4765072933483949e-06, + "loss": 1.4046, + "step": 2036 + }, + { + "epoch": 1.5235602094240837, + "grad_norm": 2.5273100270778426, + "learning_rate": 1.4721239553795485e-06, + "loss": 1.3415, + "step": 2037 + }, + { + "epoch": 1.5243081525804039, + "grad_norm": 1.9807413506033016, + "learning_rate": 1.4677460100786228e-06, + "loss": 1.6258, + "step": 2038 + }, + { + "epoch": 1.525056095736724, + "grad_norm": 2.30402812152116, + "learning_rate": 1.4633734641376913e-06, + "loss": 1.5004, + "step": 2039 + }, + { + "epoch": 1.5258040388930443, + "grad_norm": 2.087158790483457, + "learning_rate": 1.4590063242405745e-06, + "loss": 1.6873, + "step": 2040 + }, + { + "epoch": 1.5265519820493643, + "grad_norm": 2.118372398888837, + "learning_rate": 1.4546445970628275e-06, + "loss": 1.4234, + "step": 2041 + }, + { + "epoch": 1.5272999252056843, + "grad_norm": 2.1511130775609577, + "learning_rate": 1.4502882892717319e-06, + "loss": 1.3769, + "step": 2042 + }, + { + "epoch": 1.5280478683620045, + "grad_norm": 2.1884873661871573, + "learning_rate": 1.445937407526289e-06, + "loss": 1.3031, + "step": 2043 + }, + { + "epoch": 1.5287958115183247, + "grad_norm": 2.7765683795980953, + "learning_rate": 1.4415919584771999e-06, + "loss": 1.2352, + "step": 2044 + }, + { + "epoch": 1.5295437546746449, + "grad_norm": 2.028614180489609, + "learning_rate": 1.4372519487668657e-06, + "loss": 1.52, + "step": 2045 + }, + { + "epoch": 1.5302916978309649, + "grad_norm": 2.1020556899143257, + "learning_rate": 1.4329173850293738e-06, + "loss": 1.4951, + "step": 2046 + }, + { + "epoch": 1.5310396409872848, + "grad_norm": 2.3179778593302705, + "learning_rate": 1.4285882738904822e-06, + "loss": 1.666, + "step": 2047 + }, + { + "epoch": 1.531787584143605, + "grad_norm": 1.8994067679213271, + "learning_rate": 1.4242646219676193e-06, + "loss": 1.6448, + "step": 2048 + }, + { + "epoch": 1.5325355272999253, + "grad_norm": 2.0454601816558253, + "learning_rate": 1.4199464358698638e-06, + "loss": 1.4336, + "step": 2049 + }, + { + "epoch": 1.5332834704562455, + "grad_norm": 4.034733378919713, + "learning_rate": 1.415633722197945e-06, + "loss": 1.5265, + "step": 2050 + }, + { + "epoch": 1.5340314136125655, + "grad_norm": 2.1840571361544496, + "learning_rate": 1.4113264875442201e-06, + "loss": 1.4714, + "step": 2051 + }, + { + "epoch": 1.5347793567688854, + "grad_norm": 2.1325523260134327, + "learning_rate": 1.4070247384926771e-06, + "loss": 1.5952, + "step": 2052 + }, + { + "epoch": 1.5355272999252056, + "grad_norm": 2.1829128355440375, + "learning_rate": 1.4027284816189179e-06, + "loss": 1.4191, + "step": 2053 + }, + { + "epoch": 1.5362752430815259, + "grad_norm": 2.3110684101874583, + "learning_rate": 1.398437723490147e-06, + "loss": 1.1543, + "step": 2054 + }, + { + "epoch": 1.537023186237846, + "grad_norm": 1.9307045661436306, + "learning_rate": 1.3941524706651628e-06, + "loss": 1.4791, + "step": 2055 + }, + { + "epoch": 1.537771129394166, + "grad_norm": 2.429861582858043, + "learning_rate": 1.3898727296943526e-06, + "loss": 1.6646, + "step": 2056 + }, + { + "epoch": 1.538519072550486, + "grad_norm": 2.7512757537400905, + "learning_rate": 1.3855985071196725e-06, + "loss": 1.5062, + "step": 2057 + }, + { + "epoch": 1.5392670157068062, + "grad_norm": 2.001231482032362, + "learning_rate": 1.3813298094746491e-06, + "loss": 1.5845, + "step": 2058 + }, + { + "epoch": 1.5400149588631264, + "grad_norm": 2.2801937796005483, + "learning_rate": 1.3770666432843622e-06, + "loss": 1.5957, + "step": 2059 + }, + { + "epoch": 1.5407629020194467, + "grad_norm": 1.9208552584529661, + "learning_rate": 1.3728090150654316e-06, + "loss": 1.4244, + "step": 2060 + }, + { + "epoch": 1.5415108451757666, + "grad_norm": 2.151548894205283, + "learning_rate": 1.3685569313260194e-06, + "loss": 1.5491, + "step": 2061 + }, + { + "epoch": 1.5422587883320866, + "grad_norm": 2.390831839817829, + "learning_rate": 1.3643103985658047e-06, + "loss": 1.4333, + "step": 2062 + }, + { + "epoch": 1.5430067314884068, + "grad_norm": 1.9192378068090175, + "learning_rate": 1.3600694232759886e-06, + "loss": 1.579, + "step": 2063 + }, + { + "epoch": 1.543754674644727, + "grad_norm": 1.893232497778761, + "learning_rate": 1.3558340119392704e-06, + "loss": 1.5006, + "step": 2064 + }, + { + "epoch": 1.5445026178010473, + "grad_norm": 2.188522166822069, + "learning_rate": 1.35160417102985e-06, + "loss": 1.3074, + "step": 2065 + }, + { + "epoch": 1.5452505609573672, + "grad_norm": 1.8621414738132571, + "learning_rate": 1.3473799070134107e-06, + "loss": 1.2361, + "step": 2066 + }, + { + "epoch": 1.5459985041136872, + "grad_norm": 2.8513243525230902, + "learning_rate": 1.3431612263471082e-06, + "loss": 1.4059, + "step": 2067 + }, + { + "epoch": 1.5467464472700074, + "grad_norm": 2.0755993358090263, + "learning_rate": 1.3389481354795691e-06, + "loss": 1.5627, + "step": 2068 + }, + { + "epoch": 1.5474943904263276, + "grad_norm": 2.3032305515898424, + "learning_rate": 1.3347406408508695e-06, + "loss": 1.3109, + "step": 2069 + }, + { + "epoch": 1.5482423335826478, + "grad_norm": 2.1786092282114433, + "learning_rate": 1.3305387488925342e-06, + "loss": 1.4748, + "step": 2070 + }, + { + "epoch": 1.5489902767389678, + "grad_norm": 1.9189865130459391, + "learning_rate": 1.3263424660275237e-06, + "loss": 1.5392, + "step": 2071 + }, + { + "epoch": 1.5497382198952878, + "grad_norm": 1.9955444713968478, + "learning_rate": 1.3221517986702249e-06, + "loss": 1.4127, + "step": 2072 + }, + { + "epoch": 1.550486163051608, + "grad_norm": 2.4806914110830482, + "learning_rate": 1.3179667532264417e-06, + "loss": 1.4067, + "step": 2073 + }, + { + "epoch": 1.5512341062079282, + "grad_norm": 2.2880020610767375, + "learning_rate": 1.3137873360933812e-06, + "loss": 1.6302, + "step": 2074 + }, + { + "epoch": 1.5519820493642484, + "grad_norm": 2.080948123417047, + "learning_rate": 1.3096135536596488e-06, + "loss": 1.5073, + "step": 2075 + }, + { + "epoch": 1.5527299925205684, + "grad_norm": 2.250742986038551, + "learning_rate": 1.3054454123052385e-06, + "loss": 1.5661, + "step": 2076 + }, + { + "epoch": 1.5534779356768884, + "grad_norm": 2.071871078036196, + "learning_rate": 1.301282918401518e-06, + "loss": 1.6158, + "step": 2077 + }, + { + "epoch": 1.5542258788332086, + "grad_norm": 2.356103767701609, + "learning_rate": 1.2971260783112249e-06, + "loss": 1.55, + "step": 2078 + }, + { + "epoch": 1.5549738219895288, + "grad_norm": 2.34541366575487, + "learning_rate": 1.292974898388456e-06, + "loss": 1.5303, + "step": 2079 + }, + { + "epoch": 1.555721765145849, + "grad_norm": 5.551661482791568, + "learning_rate": 1.2888293849786503e-06, + "loss": 1.6955, + "step": 2080 + }, + { + "epoch": 1.556469708302169, + "grad_norm": 3.1378060851157903, + "learning_rate": 1.284689544418592e-06, + "loss": 1.4477, + "step": 2081 + }, + { + "epoch": 1.557217651458489, + "grad_norm": 2.4868630798613283, + "learning_rate": 1.2805553830363886e-06, + "loss": 1.3623, + "step": 2082 + }, + { + "epoch": 1.5579655946148092, + "grad_norm": 2.321848445644338, + "learning_rate": 1.276426907151468e-06, + "loss": 1.375, + "step": 2083 + }, + { + "epoch": 1.5587135377711294, + "grad_norm": 1.8411816939068943, + "learning_rate": 1.2723041230745676e-06, + "loss": 1.6435, + "step": 2084 + }, + { + "epoch": 1.5594614809274496, + "grad_norm": 1.8280265567262899, + "learning_rate": 1.2681870371077249e-06, + "loss": 1.5065, + "step": 2085 + }, + { + "epoch": 1.5602094240837696, + "grad_norm": 2.1390419672413805, + "learning_rate": 1.2640756555442684e-06, + "loss": 1.6877, + "step": 2086 + }, + { + "epoch": 1.5609573672400896, + "grad_norm": 2.312225460991533, + "learning_rate": 1.259969984668804e-06, + "loss": 1.5867, + "step": 2087 + }, + { + "epoch": 1.5617053103964098, + "grad_norm": 2.023565926672615, + "learning_rate": 1.2558700307572085e-06, + "loss": 1.4586, + "step": 2088 + }, + { + "epoch": 1.56245325355273, + "grad_norm": 2.481329847007712, + "learning_rate": 1.2517758000766234e-06, + "loss": 1.4491, + "step": 2089 + }, + { + "epoch": 1.5632011967090502, + "grad_norm": 1.8530137928131867, + "learning_rate": 1.2476872988854366e-06, + "loss": 1.6519, + "step": 2090 + }, + { + "epoch": 1.5639491398653702, + "grad_norm": 2.3015682087548055, + "learning_rate": 1.2436045334332824e-06, + "loss": 1.5559, + "step": 2091 + }, + { + "epoch": 1.5646970830216902, + "grad_norm": 1.9594291960209633, + "learning_rate": 1.2395275099610272e-06, + "loss": 1.5442, + "step": 2092 + }, + { + "epoch": 1.5654450261780104, + "grad_norm": 2.0295724414825465, + "learning_rate": 1.235456234700756e-06, + "loss": 1.6427, + "step": 2093 + }, + { + "epoch": 1.5661929693343306, + "grad_norm": 2.1951017494124434, + "learning_rate": 1.2313907138757724e-06, + "loss": 1.6232, + "step": 2094 + }, + { + "epoch": 1.5669409124906508, + "grad_norm": 2.150747782116089, + "learning_rate": 1.2273309537005801e-06, + "loss": 1.5395, + "step": 2095 + }, + { + "epoch": 1.5676888556469708, + "grad_norm": 2.365291313327555, + "learning_rate": 1.2232769603808797e-06, + "loss": 1.6926, + "step": 2096 + }, + { + "epoch": 1.5684367988032908, + "grad_norm": 2.0171479223174393, + "learning_rate": 1.2192287401135538e-06, + "loss": 1.604, + "step": 2097 + }, + { + "epoch": 1.569184741959611, + "grad_norm": 2.2920715980662227, + "learning_rate": 1.215186299086663e-06, + "loss": 1.4308, + "step": 2098 + }, + { + "epoch": 1.5699326851159312, + "grad_norm": 1.9759158844121194, + "learning_rate": 1.2111496434794345e-06, + "loss": 1.5651, + "step": 2099 + }, + { + "epoch": 1.5706806282722514, + "grad_norm": 2.1176044607008566, + "learning_rate": 1.207118779462248e-06, + "loss": 1.46, + "step": 2100 + }, + { + "epoch": 1.5714285714285714, + "grad_norm": 1.9525428361737172, + "learning_rate": 1.2030937131966346e-06, + "loss": 1.4852, + "step": 2101 + }, + { + "epoch": 1.5721765145848914, + "grad_norm": 2.150673865229284, + "learning_rate": 1.1990744508352604e-06, + "loss": 1.4776, + "step": 2102 + }, + { + "epoch": 1.5729244577412116, + "grad_norm": 2.286189446167476, + "learning_rate": 1.195060998521918e-06, + "loss": 1.4376, + "step": 2103 + }, + { + "epoch": 1.5736724008975318, + "grad_norm": 1.9643791439487077, + "learning_rate": 1.1910533623915244e-06, + "loss": 1.5579, + "step": 2104 + }, + { + "epoch": 1.574420344053852, + "grad_norm": 2.090201909029723, + "learning_rate": 1.1870515485701013e-06, + "loss": 1.3907, + "step": 2105 + }, + { + "epoch": 1.575168287210172, + "grad_norm": 2.4858276363695984, + "learning_rate": 1.1830555631747736e-06, + "loss": 1.4391, + "step": 2106 + }, + { + "epoch": 1.5759162303664922, + "grad_norm": 4.771625815701116, + "learning_rate": 1.1790654123137552e-06, + "loss": 1.5822, + "step": 2107 + }, + { + "epoch": 1.5766641735228122, + "grad_norm": 2.103680121833824, + "learning_rate": 1.175081102086339e-06, + "loss": 1.3806, + "step": 2108 + }, + { + "epoch": 1.5774121166791324, + "grad_norm": 1.8371589955875713, + "learning_rate": 1.171102638582896e-06, + "loss": 1.6798, + "step": 2109 + }, + { + "epoch": 1.5781600598354526, + "grad_norm": 2.019482758074796, + "learning_rate": 1.1671300278848536e-06, + "loss": 1.5921, + "step": 2110 + }, + { + "epoch": 1.5789080029917726, + "grad_norm": 2.3308933015055646, + "learning_rate": 1.1631632760646965e-06, + "loss": 1.4098, + "step": 2111 + }, + { + "epoch": 1.5796559461480928, + "grad_norm": 1.9217202404885947, + "learning_rate": 1.1592023891859554e-06, + "loss": 1.3521, + "step": 2112 + }, + { + "epoch": 1.5804038893044128, + "grad_norm": 2.774324475626262, + "learning_rate": 1.1552473733031893e-06, + "loss": 1.3592, + "step": 2113 + }, + { + "epoch": 1.581151832460733, + "grad_norm": 2.2326856797682204, + "learning_rate": 1.1512982344619904e-06, + "loss": 1.5789, + "step": 2114 + }, + { + "epoch": 1.5818997756170532, + "grad_norm": 2.7541007926007386, + "learning_rate": 1.1473549786989618e-06, + "loss": 1.6674, + "step": 2115 + }, + { + "epoch": 1.5826477187733732, + "grad_norm": 2.5056124053627276, + "learning_rate": 1.1434176120417152e-06, + "loss": 1.5896, + "step": 2116 + }, + { + "epoch": 1.5833956619296934, + "grad_norm": 2.0938758283478625, + "learning_rate": 1.139486140508861e-06, + "loss": 1.5515, + "step": 2117 + }, + { + "epoch": 1.5841436050860134, + "grad_norm": 2.048187760383302, + "learning_rate": 1.1355605701099985e-06, + "loss": 1.5174, + "step": 2118 + }, + { + "epoch": 1.5848915482423336, + "grad_norm": 1.9963537866236605, + "learning_rate": 1.1316409068457084e-06, + "loss": 1.3715, + "step": 2119 + }, + { + "epoch": 1.5856394913986538, + "grad_norm": 2.002200319983981, + "learning_rate": 1.1277271567075377e-06, + "loss": 1.2683, + "step": 2120 + }, + { + "epoch": 1.5863874345549738, + "grad_norm": 2.115917396651828, + "learning_rate": 1.1238193256779955e-06, + "loss": 1.8217, + "step": 2121 + }, + { + "epoch": 1.587135377711294, + "grad_norm": 1.9316812440659903, + "learning_rate": 1.1199174197305473e-06, + "loss": 1.6885, + "step": 2122 + }, + { + "epoch": 1.587883320867614, + "grad_norm": 2.1584325551601093, + "learning_rate": 1.1160214448295958e-06, + "loss": 1.3139, + "step": 2123 + }, + { + "epoch": 1.5886312640239342, + "grad_norm": 2.049422653462758, + "learning_rate": 1.112131406930481e-06, + "loss": 1.5131, + "step": 2124 + }, + { + "epoch": 1.5893792071802544, + "grad_norm": 2.0707368091692215, + "learning_rate": 1.1082473119794695e-06, + "loss": 1.5736, + "step": 2125 + }, + { + "epoch": 1.5901271503365744, + "grad_norm": 2.6363712596781146, + "learning_rate": 1.1043691659137372e-06, + "loss": 1.5876, + "step": 2126 + }, + { + "epoch": 1.5908750934928946, + "grad_norm": 2.145421268210254, + "learning_rate": 1.1004969746613736e-06, + "loss": 1.3829, + "step": 2127 + }, + { + "epoch": 1.5916230366492146, + "grad_norm": 3.967330668473747, + "learning_rate": 1.0966307441413598e-06, + "loss": 1.3868, + "step": 2128 + }, + { + "epoch": 1.5923709798055348, + "grad_norm": 2.317144912891404, + "learning_rate": 1.0927704802635708e-06, + "loss": 1.6398, + "step": 2129 + }, + { + "epoch": 1.593118922961855, + "grad_norm": 2.770292651808177, + "learning_rate": 1.0889161889287563e-06, + "loss": 1.2247, + "step": 2130 + }, + { + "epoch": 1.5938668661181752, + "grad_norm": 2.1244618596117544, + "learning_rate": 1.0850678760285399e-06, + "loss": 1.4294, + "step": 2131 + }, + { + "epoch": 1.5946148092744952, + "grad_norm": 2.3225139932511887, + "learning_rate": 1.0812255474454059e-06, + "loss": 1.7446, + "step": 2132 + }, + { + "epoch": 1.5953627524308152, + "grad_norm": 2.2937539561790827, + "learning_rate": 1.0773892090526888e-06, + "loss": 1.7315, + "step": 2133 + }, + { + "epoch": 1.5961106955871354, + "grad_norm": 2.3054476246482674, + "learning_rate": 1.0735588667145702e-06, + "loss": 1.4365, + "step": 2134 + }, + { + "epoch": 1.5968586387434556, + "grad_norm": 2.0715728117779326, + "learning_rate": 1.0697345262860638e-06, + "loss": 1.4164, + "step": 2135 + }, + { + "epoch": 1.5976065818997758, + "grad_norm": 2.4497269148721004, + "learning_rate": 1.0659161936130069e-06, + "loss": 1.4766, + "step": 2136 + }, + { + "epoch": 1.5983545250560958, + "grad_norm": 1.9944878937742365, + "learning_rate": 1.0621038745320579e-06, + "loss": 1.6179, + "step": 2137 + }, + { + "epoch": 1.5991024682124158, + "grad_norm": 2.427545623930715, + "learning_rate": 1.0582975748706819e-06, + "loss": 1.5383, + "step": 2138 + }, + { + "epoch": 1.599850411368736, + "grad_norm": 2.800526838620709, + "learning_rate": 1.0544973004471398e-06, + "loss": 1.4265, + "step": 2139 + }, + { + "epoch": 1.6005983545250562, + "grad_norm": 2.188643272770338, + "learning_rate": 1.0507030570704858e-06, + "loss": 1.6179, + "step": 2140 + }, + { + "epoch": 1.6013462976813764, + "grad_norm": 1.9893579139927962, + "learning_rate": 1.0469148505405518e-06, + "loss": 1.4985, + "step": 2141 + }, + { + "epoch": 1.6020942408376964, + "grad_norm": 2.3556983547218118, + "learning_rate": 1.0431326866479457e-06, + "loss": 1.5558, + "step": 2142 + }, + { + "epoch": 1.6028421839940163, + "grad_norm": 1.960531311438138, + "learning_rate": 1.0393565711740361e-06, + "loss": 1.4561, + "step": 2143 + }, + { + "epoch": 1.6035901271503366, + "grad_norm": 2.0221893371283914, + "learning_rate": 1.0355865098909452e-06, + "loss": 1.6758, + "step": 2144 + }, + { + "epoch": 1.6043380703066568, + "grad_norm": 2.1277465021737623, + "learning_rate": 1.031822508561543e-06, + "loss": 1.4399, + "step": 2145 + }, + { + "epoch": 1.605086013462977, + "grad_norm": 1.9262742942924351, + "learning_rate": 1.0280645729394368e-06, + "loss": 1.1343, + "step": 2146 + }, + { + "epoch": 1.605833956619297, + "grad_norm": 2.2329861638774946, + "learning_rate": 1.0243127087689609e-06, + "loss": 1.6903, + "step": 2147 + }, + { + "epoch": 1.606581899775617, + "grad_norm": 2.176312927309546, + "learning_rate": 1.0205669217851678e-06, + "loss": 1.4519, + "step": 2148 + }, + { + "epoch": 1.6073298429319371, + "grad_norm": 2.16659832611242, + "learning_rate": 1.01682721771382e-06, + "loss": 1.6261, + "step": 2149 + }, + { + "epoch": 1.6080777860882574, + "grad_norm": 2.0976862577077364, + "learning_rate": 1.0130936022713861e-06, + "loss": 1.6877, + "step": 2150 + }, + { + "epoch": 1.6088257292445776, + "grad_norm": 1.939584876661489, + "learning_rate": 1.0093660811650218e-06, + "loss": 1.3856, + "step": 2151 + }, + { + "epoch": 1.6095736724008975, + "grad_norm": 2.2676664988401476, + "learning_rate": 1.0056446600925718e-06, + "loss": 1.4251, + "step": 2152 + }, + { + "epoch": 1.6103216155572175, + "grad_norm": 2.145938527934006, + "learning_rate": 1.0019293447425555e-06, + "loss": 1.4787, + "step": 2153 + }, + { + "epoch": 1.6110695587135377, + "grad_norm": 2.191866876542668, + "learning_rate": 9.98220140794155e-07, + "loss": 1.4853, + "step": 2154 + }, + { + "epoch": 1.611817501869858, + "grad_norm": 1.8826253500739696, + "learning_rate": 9.945170539172171e-07, + "loss": 1.4449, + "step": 2155 + }, + { + "epoch": 1.6125654450261782, + "grad_norm": 1.8642234228698031, + "learning_rate": 9.908200897722332e-07, + "loss": 1.5634, + "step": 2156 + }, + { + "epoch": 1.6133133881824981, + "grad_norm": 2.2083791197499907, + "learning_rate": 9.871292540103377e-07, + "loss": 1.2387, + "step": 2157 + }, + { + "epoch": 1.6140613313388181, + "grad_norm": 2.246696133912523, + "learning_rate": 9.834445522732954e-07, + "loss": 1.5777, + "step": 2158 + }, + { + "epoch": 1.6148092744951383, + "grad_norm": 2.1306363773619084, + "learning_rate": 9.797659901934981e-07, + "loss": 1.4712, + "step": 2159 + }, + { + "epoch": 1.6155572176514585, + "grad_norm": 2.110217518367613, + "learning_rate": 9.760935733939497e-07, + "loss": 1.6651, + "step": 2160 + }, + { + "epoch": 1.6163051608077788, + "grad_norm": 2.179439682117969, + "learning_rate": 9.724273074882606e-07, + "loss": 1.5486, + "step": 2161 + }, + { + "epoch": 1.6170531039640987, + "grad_norm": 2.3179067663104775, + "learning_rate": 9.687671980806417e-07, + "loss": 1.3613, + "step": 2162 + }, + { + "epoch": 1.6178010471204187, + "grad_norm": 2.348305645880337, + "learning_rate": 9.6511325076589e-07, + "loss": 1.4775, + "step": 2163 + }, + { + "epoch": 1.618548990276739, + "grad_norm": 2.12237233050043, + "learning_rate": 9.61465471129383e-07, + "loss": 1.6526, + "step": 2164 + }, + { + "epoch": 1.6192969334330591, + "grad_norm": 2.7863540466603873, + "learning_rate": 9.578238647470732e-07, + "loss": 1.753, + "step": 2165 + }, + { + "epoch": 1.6200448765893793, + "grad_norm": 2.8117900603452624, + "learning_rate": 9.541884371854765e-07, + "loss": 1.3075, + "step": 2166 + }, + { + "epoch": 1.6207928197456993, + "grad_norm": 2.9808754156423913, + "learning_rate": 9.505591940016601e-07, + "loss": 1.7768, + "step": 2167 + }, + { + "epoch": 1.6215407629020193, + "grad_norm": 2.2253427463625037, + "learning_rate": 9.469361407432431e-07, + "loss": 1.2346, + "step": 2168 + }, + { + "epoch": 1.6222887060583395, + "grad_norm": 1.9660250920355822, + "learning_rate": 9.433192829483778e-07, + "loss": 1.7252, + "step": 2169 + }, + { + "epoch": 1.6230366492146597, + "grad_norm": 1.9696938280983747, + "learning_rate": 9.397086261457511e-07, + "loss": 1.5365, + "step": 2170 + }, + { + "epoch": 1.62378459237098, + "grad_norm": 2.0336667940435658, + "learning_rate": 9.361041758545659e-07, + "loss": 1.4823, + "step": 2171 + }, + { + "epoch": 1.6245325355273, + "grad_norm": 2.078893292213059, + "learning_rate": 9.325059375845414e-07, + "loss": 1.5899, + "step": 2172 + }, + { + "epoch": 1.62528047868362, + "grad_norm": 2.064045638993278, + "learning_rate": 9.289139168359019e-07, + "loss": 1.494, + "step": 2173 + }, + { + "epoch": 1.6260284218399401, + "grad_norm": 2.3357496008273686, + "learning_rate": 9.25328119099364e-07, + "loss": 1.4058, + "step": 2174 + }, + { + "epoch": 1.6267763649962603, + "grad_norm": 1.864211142107165, + "learning_rate": 9.217485498561363e-07, + "loss": 1.6405, + "step": 2175 + }, + { + "epoch": 1.6275243081525805, + "grad_norm": 2.0429471914541377, + "learning_rate": 9.181752145779027e-07, + "loss": 1.7318, + "step": 2176 + }, + { + "epoch": 1.6282722513089005, + "grad_norm": 2.3775017338185345, + "learning_rate": 9.146081187268185e-07, + "loss": 1.5041, + "step": 2177 + }, + { + "epoch": 1.6290201944652205, + "grad_norm": 2.048186881151589, + "learning_rate": 9.110472677555043e-07, + "loss": 1.716, + "step": 2178 + }, + { + "epoch": 1.6297681376215407, + "grad_norm": 2.1741173044607165, + "learning_rate": 9.074926671070322e-07, + "loss": 1.4728, + "step": 2179 + }, + { + "epoch": 1.630516080777861, + "grad_norm": 2.454760412908578, + "learning_rate": 9.039443222149231e-07, + "loss": 1.3175, + "step": 2180 + }, + { + "epoch": 1.6312640239341811, + "grad_norm": 2.4185780241697565, + "learning_rate": 9.004022385031314e-07, + "loss": 1.1176, + "step": 2181 + }, + { + "epoch": 1.6320119670905011, + "grad_norm": 2.8053770758620904, + "learning_rate": 8.968664213860417e-07, + "loss": 1.6738, + "step": 2182 + }, + { + "epoch": 1.632759910246821, + "grad_norm": 1.9248602732773943, + "learning_rate": 8.933368762684635e-07, + "loss": 1.4058, + "step": 2183 + }, + { + "epoch": 1.6335078534031413, + "grad_norm": 2.66307893152923, + "learning_rate": 8.898136085456127e-07, + "loss": 1.4401, + "step": 2184 + }, + { + "epoch": 1.6342557965594615, + "grad_norm": 1.8691673152522987, + "learning_rate": 8.862966236031151e-07, + "loss": 1.5475, + "step": 2185 + }, + { + "epoch": 1.6350037397157817, + "grad_norm": 2.072631510306256, + "learning_rate": 8.827859268169908e-07, + "loss": 1.4773, + "step": 2186 + }, + { + "epoch": 1.6357516828721017, + "grad_norm": 2.07283817930023, + "learning_rate": 8.792815235536467e-07, + "loss": 1.6592, + "step": 2187 + }, + { + "epoch": 1.6364996260284217, + "grad_norm": 2.1957104428034016, + "learning_rate": 8.757834191698716e-07, + "loss": 1.7366, + "step": 2188 + }, + { + "epoch": 1.637247569184742, + "grad_norm": 2.4193729099420045, + "learning_rate": 8.722916190128233e-07, + "loss": 1.552, + "step": 2189 + }, + { + "epoch": 1.637995512341062, + "grad_norm": 2.288413942947409, + "learning_rate": 8.688061284200266e-07, + "loss": 1.609, + "step": 2190 + }, + { + "epoch": 1.6387434554973823, + "grad_norm": 1.8976559001008242, + "learning_rate": 8.65326952719357e-07, + "loss": 1.4093, + "step": 2191 + }, + { + "epoch": 1.6394913986537023, + "grad_norm": 1.9280380470343506, + "learning_rate": 8.618540972290407e-07, + "loss": 1.5503, + "step": 2192 + }, + { + "epoch": 1.6402393418100223, + "grad_norm": 1.992669877756859, + "learning_rate": 8.583875672576441e-07, + "loss": 1.533, + "step": 2193 + }, + { + "epoch": 1.6409872849663425, + "grad_norm": 2.3323221094146684, + "learning_rate": 8.549273681040598e-07, + "loss": 1.4534, + "step": 2194 + }, + { + "epoch": 1.6417352281226627, + "grad_norm": 2.4194926940396613, + "learning_rate": 8.514735050575058e-07, + "loss": 1.5271, + "step": 2195 + }, + { + "epoch": 1.642483171278983, + "grad_norm": 2.315044870414895, + "learning_rate": 8.480259833975152e-07, + "loss": 1.5602, + "step": 2196 + }, + { + "epoch": 1.643231114435303, + "grad_norm": 2.152178686999804, + "learning_rate": 8.445848083939267e-07, + "loss": 1.5071, + "step": 2197 + }, + { + "epoch": 1.6439790575916229, + "grad_norm": 2.186468764255944, + "learning_rate": 8.411499853068783e-07, + "loss": 1.5809, + "step": 2198 + }, + { + "epoch": 1.644727000747943, + "grad_norm": 2.1672692267940583, + "learning_rate": 8.377215193867999e-07, + "loss": 1.6152, + "step": 2199 + }, + { + "epoch": 1.6454749439042633, + "grad_norm": 2.373874308901926, + "learning_rate": 8.342994158743994e-07, + "loss": 1.4325, + "step": 2200 + }, + { + "epoch": 1.6462228870605835, + "grad_norm": 2.103827883547905, + "learning_rate": 8.308836800006648e-07, + "loss": 1.5703, + "step": 2201 + }, + { + "epoch": 1.6469708302169035, + "grad_norm": 1.920174493943408, + "learning_rate": 8.27474316986846e-07, + "loss": 1.3326, + "step": 2202 + }, + { + "epoch": 1.6477187733732235, + "grad_norm": 2.126345194124045, + "learning_rate": 8.240713320444549e-07, + "loss": 1.5462, + "step": 2203 + }, + { + "epoch": 1.6484667165295437, + "grad_norm": 2.401700557408496, + "learning_rate": 8.206747303752505e-07, + "loss": 1.392, + "step": 2204 + }, + { + "epoch": 1.649214659685864, + "grad_norm": 2.375560043329784, + "learning_rate": 8.172845171712379e-07, + "loss": 1.4746, + "step": 2205 + }, + { + "epoch": 1.649962602842184, + "grad_norm": 1.9887369661137317, + "learning_rate": 8.139006976146557e-07, + "loss": 1.4899, + "step": 2206 + }, + { + "epoch": 1.650710545998504, + "grad_norm": 2.340444455173702, + "learning_rate": 8.10523276877967e-07, + "loss": 1.5384, + "step": 2207 + }, + { + "epoch": 1.651458489154824, + "grad_norm": 1.938363080811646, + "learning_rate": 8.071522601238574e-07, + "loss": 1.4859, + "step": 2208 + }, + { + "epoch": 1.6522064323111443, + "grad_norm": 3.4676867799954456, + "learning_rate": 8.037876525052213e-07, + "loss": 1.3032, + "step": 2209 + }, + { + "epoch": 1.6529543754674645, + "grad_norm": 26.3462088751488, + "learning_rate": 8.004294591651546e-07, + "loss": 1.3372, + "step": 2210 + }, + { + "epoch": 1.6537023186237847, + "grad_norm": 2.0282212688549466, + "learning_rate": 7.970776852369522e-07, + "loss": 1.4244, + "step": 2211 + }, + { + "epoch": 1.6544502617801047, + "grad_norm": 2.1074679861911036, + "learning_rate": 7.937323358440935e-07, + "loss": 1.4449, + "step": 2212 + }, + { + "epoch": 1.6551982049364247, + "grad_norm": 2.1314998894711477, + "learning_rate": 7.903934161002408e-07, + "loss": 1.6667, + "step": 2213 + }, + { + "epoch": 1.6559461480927449, + "grad_norm": 2.1106606318151906, + "learning_rate": 7.870609311092236e-07, + "loss": 1.2174, + "step": 2214 + }, + { + "epoch": 1.656694091249065, + "grad_norm": 1.9655596221959257, + "learning_rate": 7.837348859650367e-07, + "loss": 1.5765, + "step": 2215 + }, + { + "epoch": 1.6574420344053853, + "grad_norm": 3.654291938813619, + "learning_rate": 7.804152857518338e-07, + "loss": 1.692, + "step": 2216 + }, + { + "epoch": 1.6581899775617053, + "grad_norm": 2.5435290050380734, + "learning_rate": 7.771021355439135e-07, + "loss": 1.7061, + "step": 2217 + }, + { + "epoch": 1.6589379207180255, + "grad_norm": 2.6469087728519805, + "learning_rate": 7.737954404057163e-07, + "loss": 1.614, + "step": 2218 + }, + { + "epoch": 1.6596858638743455, + "grad_norm": 2.0283420172354014, + "learning_rate": 7.70495205391818e-07, + "loss": 1.6225, + "step": 2219 + }, + { + "epoch": 1.6604338070306657, + "grad_norm": 2.138400277522585, + "learning_rate": 7.672014355469143e-07, + "loss": 1.4287, + "step": 2220 + }, + { + "epoch": 1.6611817501869859, + "grad_norm": 2.2207595376557845, + "learning_rate": 7.639141359058238e-07, + "loss": 1.747, + "step": 2221 + }, + { + "epoch": 1.6619296933433059, + "grad_norm": 2.0715397362545915, + "learning_rate": 7.606333114934711e-07, + "loss": 1.4796, + "step": 2222 + }, + { + "epoch": 1.662677636499626, + "grad_norm": 2.050389971736127, + "learning_rate": 7.573589673248833e-07, + "loss": 1.4302, + "step": 2223 + }, + { + "epoch": 1.663425579655946, + "grad_norm": 2.0152074629928456, + "learning_rate": 7.540911084051827e-07, + "loss": 1.7126, + "step": 2224 + }, + { + "epoch": 1.6641735228122663, + "grad_norm": 2.308795399193383, + "learning_rate": 7.508297397295794e-07, + "loss": 1.6204, + "step": 2225 + }, + { + "epoch": 1.6649214659685865, + "grad_norm": 2.0057676546762675, + "learning_rate": 7.475748662833615e-07, + "loss": 1.5077, + "step": 2226 + }, + { + "epoch": 1.6656694091249065, + "grad_norm": 1.8441767845386274, + "learning_rate": 7.443264930418886e-07, + "loss": 1.5619, + "step": 2227 + }, + { + "epoch": 1.6664173522812267, + "grad_norm": 6.0637008807425135, + "learning_rate": 7.410846249705816e-07, + "loss": 1.3901, + "step": 2228 + }, + { + "epoch": 1.6671652954375467, + "grad_norm": 2.2495288457968377, + "learning_rate": 7.378492670249238e-07, + "loss": 1.4326, + "step": 2229 + }, + { + "epoch": 1.6679132385938669, + "grad_norm": 2.8172238528443367, + "learning_rate": 7.346204241504401e-07, + "loss": 1.6174, + "step": 2230 + }, + { + "epoch": 1.668661181750187, + "grad_norm": 2.4062786107752863, + "learning_rate": 7.313981012827015e-07, + "loss": 1.3883, + "step": 2231 + }, + { + "epoch": 1.669409124906507, + "grad_norm": 2.0139660316069894, + "learning_rate": 7.281823033473129e-07, + "loss": 1.5814, + "step": 2232 + }, + { + "epoch": 1.6701570680628273, + "grad_norm": 2.325695509369395, + "learning_rate": 7.249730352599e-07, + "loss": 1.5786, + "step": 2233 + }, + { + "epoch": 1.6709050112191473, + "grad_norm": 2.1411918304127555, + "learning_rate": 7.217703019261135e-07, + "loss": 1.6059, + "step": 2234 + }, + { + "epoch": 1.6716529543754675, + "grad_norm": 3.005764908024603, + "learning_rate": 7.185741082416103e-07, + "loss": 1.3597, + "step": 2235 + }, + { + "epoch": 1.6724008975317877, + "grad_norm": 2.246724205753057, + "learning_rate": 7.153844590920539e-07, + "loss": 1.0876, + "step": 2236 + }, + { + "epoch": 1.6731488406881077, + "grad_norm": 1.9791294586051822, + "learning_rate": 7.122013593531019e-07, + "loss": 1.4752, + "step": 2237 + }, + { + "epoch": 1.6738967838444279, + "grad_norm": 2.1713007074265995, + "learning_rate": 7.090248138904011e-07, + "loss": 1.7311, + "step": 2238 + }, + { + "epoch": 1.6746447270007478, + "grad_norm": 1.9703533003046083, + "learning_rate": 7.058548275595816e-07, + "loss": 1.4348, + "step": 2239 + }, + { + "epoch": 1.675392670157068, + "grad_norm": 2.8401957317239783, + "learning_rate": 7.026914052062433e-07, + "loss": 1.6551, + "step": 2240 + }, + { + "epoch": 1.6761406133133883, + "grad_norm": 2.1049734260271986, + "learning_rate": 6.995345516659574e-07, + "loss": 1.5742, + "step": 2241 + }, + { + "epoch": 1.6768885564697082, + "grad_norm": 1.994686874217122, + "learning_rate": 6.963842717642488e-07, + "loss": 1.4926, + "step": 2242 + }, + { + "epoch": 1.6776364996260285, + "grad_norm": 2.2660617231825015, + "learning_rate": 6.932405703165973e-07, + "loss": 1.3686, + "step": 2243 + }, + { + "epoch": 1.6783844427823484, + "grad_norm": 1.8185714106634254, + "learning_rate": 6.901034521284261e-07, + "loss": 1.4037, + "step": 2244 + }, + { + "epoch": 1.6791323859386686, + "grad_norm": 2.251443612914309, + "learning_rate": 6.86972921995096e-07, + "loss": 1.5318, + "step": 2245 + }, + { + "epoch": 1.6798803290949889, + "grad_norm": 2.0469959027080304, + "learning_rate": 6.838489847018975e-07, + "loss": 1.4058, + "step": 2246 + }, + { + "epoch": 1.680628272251309, + "grad_norm": 2.8312465365358497, + "learning_rate": 6.807316450240425e-07, + "loss": 1.4591, + "step": 2247 + }, + { + "epoch": 1.681376215407629, + "grad_norm": 2.050335036045211, + "learning_rate": 6.77620907726656e-07, + "loss": 1.4649, + "step": 2248 + }, + { + "epoch": 1.682124158563949, + "grad_norm": 2.681281902137062, + "learning_rate": 6.745167775647754e-07, + "loss": 1.3808, + "step": 2249 + }, + { + "epoch": 1.6828721017202692, + "grad_norm": 2.3613740536883028, + "learning_rate": 6.714192592833335e-07, + "loss": 1.5617, + "step": 2250 + }, + { + "epoch": 1.6836200448765894, + "grad_norm": 5.220570063801175, + "learning_rate": 6.6832835761716e-07, + "loss": 1.5503, + "step": 2251 + }, + { + "epoch": 1.6843679880329097, + "grad_norm": 2.1751798004383813, + "learning_rate": 6.652440772909708e-07, + "loss": 1.3181, + "step": 2252 + }, + { + "epoch": 1.6851159311892296, + "grad_norm": 1.861716676213943, + "learning_rate": 6.621664230193553e-07, + "loss": 1.5701, + "step": 2253 + }, + { + "epoch": 1.6858638743455496, + "grad_norm": 2.0590658250920937, + "learning_rate": 6.590953995067812e-07, + "loss": 1.4029, + "step": 2254 + }, + { + "epoch": 1.6866118175018698, + "grad_norm": 1.9879999545237546, + "learning_rate": 6.560310114475754e-07, + "loss": 1.6427, + "step": 2255 + }, + { + "epoch": 1.68735976065819, + "grad_norm": 2.357557448082484, + "learning_rate": 6.529732635259234e-07, + "loss": 1.3745, + "step": 2256 + }, + { + "epoch": 1.6881077038145103, + "grad_norm": 2.3026248632923294, + "learning_rate": 6.499221604158623e-07, + "loss": 1.2441, + "step": 2257 + }, + { + "epoch": 1.6888556469708302, + "grad_norm": 2.3038073852241303, + "learning_rate": 6.468777067812699e-07, + "loss": 1.6666, + "step": 2258 + }, + { + "epoch": 1.6896035901271502, + "grad_norm": 2.0187727692615507, + "learning_rate": 6.438399072758622e-07, + "loss": 1.5187, + "step": 2259 + }, + { + "epoch": 1.6903515332834704, + "grad_norm": 2.2074646107498026, + "learning_rate": 6.408087665431806e-07, + "loss": 1.5125, + "step": 2260 + }, + { + "epoch": 1.6910994764397906, + "grad_norm": 2.112342725031505, + "learning_rate": 6.377842892165892e-07, + "loss": 1.7615, + "step": 2261 + }, + { + "epoch": 1.6918474195961108, + "grad_norm": 2.050165864693556, + "learning_rate": 6.347664799192688e-07, + "loss": 1.4094, + "step": 2262 + }, + { + "epoch": 1.6925953627524308, + "grad_norm": 2.2248623930881917, + "learning_rate": 6.317553432642032e-07, + "loss": 1.3822, + "step": 2263 + }, + { + "epoch": 1.6933433059087508, + "grad_norm": 1.9962055574940987, + "learning_rate": 6.287508838541795e-07, + "loss": 1.3045, + "step": 2264 + }, + { + "epoch": 1.694091249065071, + "grad_norm": 2.123589242476627, + "learning_rate": 6.257531062817795e-07, + "loss": 1.6171, + "step": 2265 + }, + { + "epoch": 1.6948391922213912, + "grad_norm": 2.16169672971795, + "learning_rate": 6.227620151293662e-07, + "loss": 1.2726, + "step": 2266 + }, + { + "epoch": 1.6955871353777114, + "grad_norm": 2.1112939507335255, + "learning_rate": 6.197776149690871e-07, + "loss": 1.4156, + "step": 2267 + }, + { + "epoch": 1.6963350785340314, + "grad_norm": 2.106678910908053, + "learning_rate": 6.167999103628569e-07, + "loss": 1.5394, + "step": 2268 + }, + { + "epoch": 1.6970830216903514, + "grad_norm": 1.9638365504681343, + "learning_rate": 6.138289058623615e-07, + "loss": 1.3869, + "step": 2269 + }, + { + "epoch": 1.6978309648466716, + "grad_norm": 2.0360450501004608, + "learning_rate": 6.1086460600904e-07, + "loss": 1.5003, + "step": 2270 + }, + { + "epoch": 1.6985789080029918, + "grad_norm": 2.3961754854474635, + "learning_rate": 6.079070153340827e-07, + "loss": 1.5857, + "step": 2271 + }, + { + "epoch": 1.699326851159312, + "grad_norm": 2.040073849100188, + "learning_rate": 6.049561383584301e-07, + "loss": 1.4951, + "step": 2272 + }, + { + "epoch": 1.700074794315632, + "grad_norm": 2.213961217421706, + "learning_rate": 6.020119795927542e-07, + "loss": 1.5235, + "step": 2273 + }, + { + "epoch": 1.700822737471952, + "grad_norm": 2.1136435116959955, + "learning_rate": 5.990745435374612e-07, + "loss": 1.5609, + "step": 2274 + }, + { + "epoch": 1.7015706806282722, + "grad_norm": 2.0120351664375065, + "learning_rate": 5.961438346826792e-07, + "loss": 1.4428, + "step": 2275 + }, + { + "epoch": 1.7023186237845924, + "grad_norm": 2.532682732752495, + "learning_rate": 5.932198575082526e-07, + "loss": 1.5325, + "step": 2276 + }, + { + "epoch": 1.7030665669409126, + "grad_norm": 2.4397470807838615, + "learning_rate": 5.903026164837389e-07, + "loss": 1.4296, + "step": 2277 + }, + { + "epoch": 1.7038145100972326, + "grad_norm": 2.3331974285766117, + "learning_rate": 5.873921160683943e-07, + "loss": 1.411, + "step": 2278 + }, + { + "epoch": 1.7045624532535526, + "grad_norm": 2.347991377455073, + "learning_rate": 5.844883607111751e-07, + "loss": 1.5973, + "step": 2279 + }, + { + "epoch": 1.7053103964098728, + "grad_norm": 2.6595757446677672, + "learning_rate": 5.81591354850728e-07, + "loss": 1.5303, + "step": 2280 + }, + { + "epoch": 1.706058339566193, + "grad_norm": 2.4394408444529363, + "learning_rate": 5.787011029153777e-07, + "loss": 1.4248, + "step": 2281 + }, + { + "epoch": 1.7068062827225132, + "grad_norm": 1.8242589418190063, + "learning_rate": 5.758176093231294e-07, + "loss": 1.514, + "step": 2282 + }, + { + "epoch": 1.7075542258788332, + "grad_norm": 2.657919817840494, + "learning_rate": 5.729408784816559e-07, + "loss": 1.6412, + "step": 2283 + }, + { + "epoch": 1.7083021690351532, + "grad_norm": 2.2375123648550286, + "learning_rate": 5.70070914788291e-07, + "loss": 1.618, + "step": 2284 + }, + { + "epoch": 1.7090501121914734, + "grad_norm": 1.930886704307051, + "learning_rate": 5.672077226300265e-07, + "loss": 1.4256, + "step": 2285 + }, + { + "epoch": 1.7097980553477936, + "grad_norm": 2.335189863714475, + "learning_rate": 5.643513063835038e-07, + "loss": 1.4353, + "step": 2286 + }, + { + "epoch": 1.7105459985041138, + "grad_norm": 2.17589884890653, + "learning_rate": 5.615016704150056e-07, + "loss": 1.5447, + "step": 2287 + }, + { + "epoch": 1.7112939416604338, + "grad_norm": 2.0405114374864985, + "learning_rate": 5.586588190804499e-07, + "loss": 1.2703, + "step": 2288 + }, + { + "epoch": 1.7120418848167538, + "grad_norm": 2.4920929804707, + "learning_rate": 5.558227567253832e-07, + "loss": 1.3687, + "step": 2289 + }, + { + "epoch": 1.712789827973074, + "grad_norm": 2.3332865373910994, + "learning_rate": 5.529934876849774e-07, + "loss": 1.4464, + "step": 2290 + }, + { + "epoch": 1.7135377711293942, + "grad_norm": 2.087083567467807, + "learning_rate": 5.501710162840162e-07, + "loss": 1.6636, + "step": 2291 + }, + { + "epoch": 1.7142857142857144, + "grad_norm": 2.20640731659084, + "learning_rate": 5.473553468368963e-07, + "loss": 1.5365, + "step": 2292 + }, + { + "epoch": 1.7150336574420344, + "grad_norm": 1.9950588833496596, + "learning_rate": 5.445464836476155e-07, + "loss": 1.5039, + "step": 2293 + }, + { + "epoch": 1.7157816005983544, + "grad_norm": 1.9188068245276764, + "learning_rate": 5.417444310097664e-07, + "loss": 1.6118, + "step": 2294 + }, + { + "epoch": 1.7165295437546746, + "grad_norm": 2.2101521869015466, + "learning_rate": 5.389491932065338e-07, + "loss": 1.516, + "step": 2295 + }, + { + "epoch": 1.7172774869109948, + "grad_norm": 2.430038230291678, + "learning_rate": 5.361607745106817e-07, + "loss": 1.7822, + "step": 2296 + }, + { + "epoch": 1.718025430067315, + "grad_norm": 2.112333110606934, + "learning_rate": 5.333791791845549e-07, + "loss": 1.3096, + "step": 2297 + }, + { + "epoch": 1.718773373223635, + "grad_norm": 2.081653521841784, + "learning_rate": 5.306044114800635e-07, + "loss": 1.7073, + "step": 2298 + }, + { + "epoch": 1.719521316379955, + "grad_norm": 2.1842501291564536, + "learning_rate": 5.278364756386839e-07, + "loss": 1.5219, + "step": 2299 + }, + { + "epoch": 1.7202692595362752, + "grad_norm": 2.0379383505135378, + "learning_rate": 5.250753758914506e-07, + "loss": 1.6205, + "step": 2300 + }, + { + "epoch": 1.7210172026925954, + "grad_norm": 2.009245245559314, + "learning_rate": 5.223211164589448e-07, + "loss": 1.5887, + "step": 2301 + }, + { + "epoch": 1.7217651458489156, + "grad_norm": 2.015197943816919, + "learning_rate": 5.195737015512947e-07, + "loss": 1.5484, + "step": 2302 + }, + { + "epoch": 1.7225130890052356, + "grad_norm": 2.289269434852466, + "learning_rate": 5.168331353681643e-07, + "loss": 1.6589, + "step": 2303 + }, + { + "epoch": 1.7232610321615556, + "grad_norm": 2.2148654412346813, + "learning_rate": 5.140994220987489e-07, + "loss": 1.4966, + "step": 2304 + }, + { + "epoch": 1.7240089753178758, + "grad_norm": 2.8308057735450536, + "learning_rate": 5.113725659217689e-07, + "loss": 1.3664, + "step": 2305 + }, + { + "epoch": 1.724756918474196, + "grad_norm": 1.8291437001482367, + "learning_rate": 5.086525710054652e-07, + "loss": 1.4066, + "step": 2306 + }, + { + "epoch": 1.7255048616305162, + "grad_norm": 2.1004806097977156, + "learning_rate": 5.059394415075858e-07, + "loss": 1.4753, + "step": 2307 + }, + { + "epoch": 1.7262528047868362, + "grad_norm": 1.9827819773302846, + "learning_rate": 5.032331815753894e-07, + "loss": 1.5981, + "step": 2308 + }, + { + "epoch": 1.7270007479431562, + "grad_norm": 1.9093652465515325, + "learning_rate": 5.005337953456297e-07, + "loss": 1.5399, + "step": 2309 + }, + { + "epoch": 1.7277486910994764, + "grad_norm": 1.9946621829241813, + "learning_rate": 4.97841286944557e-07, + "loss": 1.6904, + "step": 2310 + }, + { + "epoch": 1.7284966342557966, + "grad_norm": 1.8974388918563863, + "learning_rate": 4.951556604879049e-07, + "loss": 1.4393, + "step": 2311 + }, + { + "epoch": 1.7292445774121168, + "grad_norm": 2.525078932471866, + "learning_rate": 4.924769200808893e-07, + "loss": 1.5531, + "step": 2312 + }, + { + "epoch": 1.7299925205684368, + "grad_norm": 1.9763926001047016, + "learning_rate": 4.898050698182017e-07, + "loss": 1.626, + "step": 2313 + }, + { + "epoch": 1.7307404637247568, + "grad_norm": 2.162763647085165, + "learning_rate": 4.871401137839971e-07, + "loss": 1.5252, + "step": 2314 + }, + { + "epoch": 1.731488406881077, + "grad_norm": 2.9196439849867017, + "learning_rate": 4.844820560518959e-07, + "loss": 1.6514, + "step": 2315 + }, + { + "epoch": 1.7322363500373972, + "grad_norm": 2.303977386213056, + "learning_rate": 4.81830900684973e-07, + "loss": 1.6474, + "step": 2316 + }, + { + "epoch": 1.7329842931937174, + "grad_norm": 2.0975082750648313, + "learning_rate": 4.791866517357491e-07, + "loss": 1.7261, + "step": 2317 + }, + { + "epoch": 1.7337322363500374, + "grad_norm": 1.8465511642228039, + "learning_rate": 4.765493132461929e-07, + "loss": 1.558, + "step": 2318 + }, + { + "epoch": 1.7344801795063574, + "grad_norm": 2.129757428707769, + "learning_rate": 4.739188892477076e-07, + "loss": 1.67, + "step": 2319 + }, + { + "epoch": 1.7352281226626776, + "grad_norm": 2.068726386273408, + "learning_rate": 4.7129538376112673e-07, + "loss": 1.607, + "step": 2320 + }, + { + "epoch": 1.7359760658189978, + "grad_norm": 18.905947827533527, + "learning_rate": 4.686788007967086e-07, + "loss": 1.5004, + "step": 2321 + }, + { + "epoch": 1.736724008975318, + "grad_norm": 2.369508435565922, + "learning_rate": 4.660691443541282e-07, + "loss": 1.4297, + "step": 2322 + }, + { + "epoch": 1.737471952131638, + "grad_norm": 2.3957086726273915, + "learning_rate": 4.634664184224769e-07, + "loss": 1.6055, + "step": 2323 + }, + { + "epoch": 1.738219895287958, + "grad_norm": 2.0471892254773585, + "learning_rate": 4.608706269802471e-07, + "loss": 1.7835, + "step": 2324 + }, + { + "epoch": 1.7389678384442782, + "grad_norm": 2.382922615267553, + "learning_rate": 4.5828177399533436e-07, + "loss": 1.356, + "step": 2325 + }, + { + "epoch": 1.7397157816005984, + "grad_norm": 2.1891049071767, + "learning_rate": 4.556998634250287e-07, + "loss": 1.4159, + "step": 2326 + }, + { + "epoch": 1.7404637247569186, + "grad_norm": 1.9331144046495055, + "learning_rate": 4.531248992160053e-07, + "loss": 1.4635, + "step": 2327 + }, + { + "epoch": 1.7412116679132386, + "grad_norm": 2.0436693673752657, + "learning_rate": 4.5055688530432407e-07, + "loss": 1.4752, + "step": 2328 + }, + { + "epoch": 1.7419596110695588, + "grad_norm": 2.041595778582549, + "learning_rate": 4.4799582561541934e-07, + "loss": 1.4355, + "step": 2329 + }, + { + "epoch": 1.7427075542258788, + "grad_norm": 1.9542156272337414, + "learning_rate": 4.4544172406409345e-07, + "loss": 1.5779, + "step": 2330 + }, + { + "epoch": 1.743455497382199, + "grad_norm": 2.1501114342989016, + "learning_rate": 4.428945845545168e-07, + "loss": 1.6094, + "step": 2331 + }, + { + "epoch": 1.7442034405385192, + "grad_norm": 2.061999824959966, + "learning_rate": 4.403544109802144e-07, + "loss": 1.4771, + "step": 2332 + }, + { + "epoch": 1.7449513836948392, + "grad_norm": 1.9545383981115723, + "learning_rate": 4.3782120722406565e-07, + "loss": 1.541, + "step": 2333 + }, + { + "epoch": 1.7456993268511594, + "grad_norm": 2.094784851137542, + "learning_rate": 4.352949771582937e-07, + "loss": 1.4103, + "step": 2334 + }, + { + "epoch": 1.7464472700074793, + "grad_norm": 2.0460991957834382, + "learning_rate": 4.3277572464446247e-07, + "loss": 1.7565, + "step": 2335 + }, + { + "epoch": 1.7471952131637996, + "grad_norm": 2.0029284277094774, + "learning_rate": 4.30263453533471e-07, + "loss": 1.1959, + "step": 2336 + }, + { + "epoch": 1.7479431563201198, + "grad_norm": 2.2713781736682144, + "learning_rate": 4.277581676655451e-07, + "loss": 1.3789, + "step": 2337 + }, + { + "epoch": 1.7486910994764397, + "grad_norm": 2.248376102147184, + "learning_rate": 4.2525987087023433e-07, + "loss": 1.7068, + "step": 2338 + }, + { + "epoch": 1.74943904263276, + "grad_norm": 1.8823939960226306, + "learning_rate": 4.2276856696640557e-07, + "loss": 1.283, + "step": 2339 + }, + { + "epoch": 1.75018698578908, + "grad_norm": 1.986118825334705, + "learning_rate": 4.202842597622336e-07, + "loss": 1.2941, + "step": 2340 + }, + { + "epoch": 1.7509349289454001, + "grad_norm": 1.9351158915360367, + "learning_rate": 4.1780695305520147e-07, + "loss": 1.7014, + "step": 2341 + }, + { + "epoch": 1.7516828721017204, + "grad_norm": 2.04589170911797, + "learning_rate": 4.1533665063208773e-07, + "loss": 1.5926, + "step": 2342 + }, + { + "epoch": 1.7524308152580403, + "grad_norm": 2.205033790810646, + "learning_rate": 4.128733562689685e-07, + "loss": 1.6614, + "step": 2343 + }, + { + "epoch": 1.7531787584143605, + "grad_norm": 1.8765143832186768, + "learning_rate": 4.1041707373120354e-07, + "loss": 1.41, + "step": 2344 + }, + { + "epoch": 1.7539267015706805, + "grad_norm": 2.058808129420275, + "learning_rate": 4.0796780677343606e-07, + "loss": 1.4387, + "step": 2345 + }, + { + "epoch": 1.7539267015706805, + "eval_loss": 1.6688328981399536, + "eval_runtime": 68.9525, + "eval_samples_per_second": 59.07, + "eval_steps_per_second": 3.698, + "step": 2345 + }, + { + "epoch": 1.7546746447270007, + "grad_norm": 2.319592158151184, + "learning_rate": 4.055255591395879e-07, + "loss": 1.5642, + "step": 2346 + }, + { + "epoch": 1.755422587883321, + "grad_norm": 2.1403019309862823, + "learning_rate": 4.0309033456284565e-07, + "loss": 1.239, + "step": 2347 + }, + { + "epoch": 1.756170531039641, + "grad_norm": 2.4454244036987722, + "learning_rate": 4.0066213676566544e-07, + "loss": 1.5766, + "step": 2348 + }, + { + "epoch": 1.7569184741959611, + "grad_norm": 2.521745352391224, + "learning_rate": 3.9824096945976056e-07, + "loss": 1.3297, + "step": 2349 + }, + { + "epoch": 1.7576664173522811, + "grad_norm": 2.089093300262014, + "learning_rate": 3.95826836346096e-07, + "loss": 1.6272, + "step": 2350 + }, + { + "epoch": 1.7584143605086013, + "grad_norm": 2.3873647114983205, + "learning_rate": 3.934197411148871e-07, + "loss": 1.4455, + "step": 2351 + }, + { + "epoch": 1.7591623036649215, + "grad_norm": 2.61685914266429, + "learning_rate": 3.910196874455896e-07, + "loss": 1.4837, + "step": 2352 + }, + { + "epoch": 1.7599102468212415, + "grad_norm": 2.1774199142940245, + "learning_rate": 3.886266790068971e-07, + "loss": 1.2578, + "step": 2353 + }, + { + "epoch": 1.7606581899775617, + "grad_norm": 1.8991607451871677, + "learning_rate": 3.8624071945673246e-07, + "loss": 1.4584, + "step": 2354 + }, + { + "epoch": 1.7614061331338817, + "grad_norm": 2.0064481250760395, + "learning_rate": 3.8386181244224274e-07, + "loss": 1.6983, + "step": 2355 + }, + { + "epoch": 1.762154076290202, + "grad_norm": 2.2805817119740017, + "learning_rate": 3.8148996159979803e-07, + "loss": 1.4238, + "step": 2356 + }, + { + "epoch": 1.7629020194465221, + "grad_norm": 2.5770654545694796, + "learning_rate": 3.7912517055497756e-07, + "loss": 1.5476, + "step": 2357 + }, + { + "epoch": 1.7636499626028423, + "grad_norm": 2.122699515520187, + "learning_rate": 3.767674429225743e-07, + "loss": 1.6015, + "step": 2358 + }, + { + "epoch": 1.7643979057591623, + "grad_norm": 2.030592131596824, + "learning_rate": 3.744167823065814e-07, + "loss": 1.4118, + "step": 2359 + }, + { + "epoch": 1.7651458489154823, + "grad_norm": 2.447574683180195, + "learning_rate": 3.720731923001886e-07, + "loss": 1.4286, + "step": 2360 + }, + { + "epoch": 1.7658937920718025, + "grad_norm": 1.9298590282706602, + "learning_rate": 3.6973667648578016e-07, + "loss": 1.4301, + "step": 2361 + }, + { + "epoch": 1.7666417352281227, + "grad_norm": 2.0656551770427987, + "learning_rate": 3.674072384349242e-07, + "loss": 1.4756, + "step": 2362 + }, + { + "epoch": 1.767389678384443, + "grad_norm": 1.9071168219914683, + "learning_rate": 3.6508488170837133e-07, + "loss": 1.5234, + "step": 2363 + }, + { + "epoch": 1.768137621540763, + "grad_norm": 2.283471694900322, + "learning_rate": 3.6276960985604736e-07, + "loss": 1.5522, + "step": 2364 + }, + { + "epoch": 1.768885564697083, + "grad_norm": 1.8961515311009058, + "learning_rate": 3.604614264170486e-07, + "loss": 1.604, + "step": 2365 + }, + { + "epoch": 1.7696335078534031, + "grad_norm": 2.5855963180985104, + "learning_rate": 3.581603349196372e-07, + "loss": 1.6643, + "step": 2366 + }, + { + "epoch": 1.7703814510097233, + "grad_norm": 2.28491546054092, + "learning_rate": 3.5586633888123166e-07, + "loss": 1.5401, + "step": 2367 + }, + { + "epoch": 1.7711293941660435, + "grad_norm": 1.9771464331495074, + "learning_rate": 3.5357944180840544e-07, + "loss": 1.6284, + "step": 2368 + }, + { + "epoch": 1.7718773373223635, + "grad_norm": 2.0056559061589874, + "learning_rate": 3.5129964719688383e-07, + "loss": 1.554, + "step": 2369 + }, + { + "epoch": 1.7726252804786835, + "grad_norm": 5.001381907223179, + "learning_rate": 3.4902695853152977e-07, + "loss": 1.7014, + "step": 2370 + }, + { + "epoch": 1.7733732236350037, + "grad_norm": 4.967003682739623, + "learning_rate": 3.467613792863489e-07, + "loss": 1.4919, + "step": 2371 + }, + { + "epoch": 1.774121166791324, + "grad_norm": 2.4835765449448033, + "learning_rate": 3.445029129244787e-07, + "loss": 1.4775, + "step": 2372 + }, + { + "epoch": 1.7748691099476441, + "grad_norm": 1.8666399750398865, + "learning_rate": 3.4225156289818096e-07, + "loss": 1.6447, + "step": 2373 + }, + { + "epoch": 1.7756170531039641, + "grad_norm": 2.266451191540591, + "learning_rate": 3.4000733264884343e-07, + "loss": 1.2904, + "step": 2374 + }, + { + "epoch": 1.776364996260284, + "grad_norm": 2.7739339345970286, + "learning_rate": 3.377702256069676e-07, + "loss": 1.4961, + "step": 2375 + }, + { + "epoch": 1.7771129394166043, + "grad_norm": 2.0793911901571906, + "learning_rate": 3.355402451921685e-07, + "loss": 1.2204, + "step": 2376 + }, + { + "epoch": 1.7778608825729245, + "grad_norm": 2.6235710512380175, + "learning_rate": 3.333173948131663e-07, + "loss": 1.333, + "step": 2377 + }, + { + "epoch": 1.7786088257292447, + "grad_norm": 2.0173501711245123, + "learning_rate": 3.311016778677828e-07, + "loss": 1.4236, + "step": 2378 + }, + { + "epoch": 1.7793567688855647, + "grad_norm": 2.1861828810447754, + "learning_rate": 3.2889309774293666e-07, + "loss": 1.4611, + "step": 2379 + }, + { + "epoch": 1.7801047120418847, + "grad_norm": 2.3235755693742415, + "learning_rate": 3.26691657814634e-07, + "loss": 1.6128, + "step": 2380 + }, + { + "epoch": 1.780852655198205, + "grad_norm": 2.0121248663417273, + "learning_rate": 3.2449736144797183e-07, + "loss": 1.6771, + "step": 2381 + }, + { + "epoch": 1.7816005983545251, + "grad_norm": 2.039961935687293, + "learning_rate": 3.2231021199712274e-07, + "loss": 1.296, + "step": 2382 + }, + { + "epoch": 1.7823485415108453, + "grad_norm": 1.9397187805237894, + "learning_rate": 3.2013021280533576e-07, + "loss": 1.5943, + "step": 2383 + }, + { + "epoch": 1.7830964846671653, + "grad_norm": 2.0084751389207343, + "learning_rate": 3.1795736720493184e-07, + "loss": 1.6198, + "step": 2384 + }, + { + "epoch": 1.7838444278234853, + "grad_norm": 1.8405659314682121, + "learning_rate": 3.157916785172971e-07, + "loss": 1.1826, + "step": 2385 + }, + { + "epoch": 1.7845923709798055, + "grad_norm": 2.012441098007774, + "learning_rate": 3.13633150052875e-07, + "loss": 1.4214, + "step": 2386 + }, + { + "epoch": 1.7853403141361257, + "grad_norm": 2.183369130351531, + "learning_rate": 3.1148178511116624e-07, + "loss": 1.409, + "step": 2387 + }, + { + "epoch": 1.786088257292446, + "grad_norm": 2.26311001075309, + "learning_rate": 3.0933758698072023e-07, + "loss": 1.6134, + "step": 2388 + }, + { + "epoch": 1.786836200448766, + "grad_norm": 2.1333484251099066, + "learning_rate": 3.07200558939133e-07, + "loss": 1.4159, + "step": 2389 + }, + { + "epoch": 1.7875841436050859, + "grad_norm": 2.0686668820666645, + "learning_rate": 3.050707042530382e-07, + "loss": 1.6594, + "step": 2390 + }, + { + "epoch": 1.788332086761406, + "grad_norm": 2.160487561349476, + "learning_rate": 3.029480261781054e-07, + "loss": 1.5126, + "step": 2391 + }, + { + "epoch": 1.7890800299177263, + "grad_norm": 2.0925293129477467, + "learning_rate": 3.008325279590357e-07, + "loss": 1.6313, + "step": 2392 + }, + { + "epoch": 1.7898279730740465, + "grad_norm": 2.342206550468256, + "learning_rate": 2.98724212829552e-07, + "loss": 1.5335, + "step": 2393 + }, + { + "epoch": 1.7905759162303665, + "grad_norm": 2.0710699044926244, + "learning_rate": 2.966230840124007e-07, + "loss": 1.7452, + "step": 2394 + }, + { + "epoch": 1.7913238593866865, + "grad_norm": 2.7879304732061585, + "learning_rate": 2.945291447193399e-07, + "loss": 1.6661, + "step": 2395 + }, + { + "epoch": 1.7920718025430067, + "grad_norm": 1.949859368348173, + "learning_rate": 2.9244239815113874e-07, + "loss": 1.3709, + "step": 2396 + }, + { + "epoch": 1.792819745699327, + "grad_norm": 2.074036794319722, + "learning_rate": 2.903628474975739e-07, + "loss": 1.572, + "step": 2397 + }, + { + "epoch": 1.793567688855647, + "grad_norm": 2.067015729291377, + "learning_rate": 2.882904959374205e-07, + "loss": 1.5709, + "step": 2398 + }, + { + "epoch": 1.794315632011967, + "grad_norm": 2.1053641282228504, + "learning_rate": 2.862253466384507e-07, + "loss": 1.6078, + "step": 2399 + }, + { + "epoch": 1.795063575168287, + "grad_norm": 2.076743896915406, + "learning_rate": 2.8416740275742484e-07, + "loss": 1.187, + "step": 2400 + }, + { + "epoch": 1.7958115183246073, + "grad_norm": 2.228569645979376, + "learning_rate": 2.821166674400905e-07, + "loss": 1.4548, + "step": 2401 + }, + { + "epoch": 1.7965594614809275, + "grad_norm": 2.228585777996344, + "learning_rate": 2.8007314382117733e-07, + "loss": 1.3066, + "step": 2402 + }, + { + "epoch": 1.7973074046372477, + "grad_norm": 2.0400704468200304, + "learning_rate": 2.780368350243895e-07, + "loss": 1.6899, + "step": 2403 + }, + { + "epoch": 1.7980553477935677, + "grad_norm": 2.098014459731113, + "learning_rate": 2.760077441624043e-07, + "loss": 1.6164, + "step": 2404 + }, + { + "epoch": 1.7988032909498877, + "grad_norm": 2.1701612189779134, + "learning_rate": 2.7398587433686563e-07, + "loss": 1.7181, + "step": 2405 + }, + { + "epoch": 1.7995512341062079, + "grad_norm": 2.0541965702219844, + "learning_rate": 2.719712286383769e-07, + "loss": 1.6211, + "step": 2406 + }, + { + "epoch": 1.800299177262528, + "grad_norm": 1.9831162856955524, + "learning_rate": 2.6996381014650353e-07, + "loss": 1.5731, + "step": 2407 + }, + { + "epoch": 1.8010471204188483, + "grad_norm": 2.576561913164843, + "learning_rate": 2.6796362192975766e-07, + "loss": 1.3179, + "step": 2408 + }, + { + "epoch": 1.8017950635751683, + "grad_norm": 2.354050506299727, + "learning_rate": 2.659706670456047e-07, + "loss": 1.4444, + "step": 2409 + }, + { + "epoch": 1.8025430067314883, + "grad_norm": 2.0421691464123017, + "learning_rate": 2.6398494854045055e-07, + "loss": 1.3968, + "step": 2410 + }, + { + "epoch": 1.8032909498878085, + "grad_norm": 3.224200758972294, + "learning_rate": 2.620064694496377e-07, + "loss": 1.4404, + "step": 2411 + }, + { + "epoch": 1.8040388930441287, + "grad_norm": 2.202292749578127, + "learning_rate": 2.600352327974487e-07, + "loss": 1.6281, + "step": 2412 + }, + { + "epoch": 1.8047868362004489, + "grad_norm": 2.096637537152423, + "learning_rate": 2.5807124159708987e-07, + "loss": 1.6135, + "step": 2413 + }, + { + "epoch": 1.8055347793567689, + "grad_norm": 2.3053166099476665, + "learning_rate": 2.561144988506931e-07, + "loss": 1.7426, + "step": 2414 + }, + { + "epoch": 1.8062827225130889, + "grad_norm": 1.9561597317954442, + "learning_rate": 2.5416500754931294e-07, + "loss": 1.5421, + "step": 2415 + }, + { + "epoch": 1.807030665669409, + "grad_norm": 2.021985617884195, + "learning_rate": 2.5222277067291633e-07, + "loss": 1.5353, + "step": 2416 + }, + { + "epoch": 1.8077786088257293, + "grad_norm": 2.22656381218575, + "learning_rate": 2.502877911903839e-07, + "loss": 1.603, + "step": 2417 + }, + { + "epoch": 1.8085265519820495, + "grad_norm": 2.369215655955664, + "learning_rate": 2.4836007205950087e-07, + "loss": 1.2252, + "step": 2418 + }, + { + "epoch": 1.8092744951383695, + "grad_norm": 2.0372693866576346, + "learning_rate": 2.4643961622695444e-07, + "loss": 1.4582, + "step": 2419 + }, + { + "epoch": 1.8100224382946895, + "grad_norm": 2.102706616708393, + "learning_rate": 2.4452642662833146e-07, + "loss": 1.5663, + "step": 2420 + }, + { + "epoch": 1.8107703814510097, + "grad_norm": 2.3348578546445693, + "learning_rate": 2.426205061881082e-07, + "loss": 1.5671, + "step": 2421 + }, + { + "epoch": 1.8115183246073299, + "grad_norm": 2.417348772169172, + "learning_rate": 2.407218578196524e-07, + "loss": 1.5059, + "step": 2422 + }, + { + "epoch": 1.81226626776365, + "grad_norm": 2.1467859261969644, + "learning_rate": 2.388304844252143e-07, + "loss": 1.4867, + "step": 2423 + }, + { + "epoch": 1.81301421091997, + "grad_norm": 2.0305000804611075, + "learning_rate": 2.3694638889592336e-07, + "loss": 1.5388, + "step": 2424 + }, + { + "epoch": 1.81376215407629, + "grad_norm": 2.0685117182259303, + "learning_rate": 2.3506957411178478e-07, + "loss": 1.4428, + "step": 2425 + }, + { + "epoch": 1.8145100972326103, + "grad_norm": 1.9781902728112788, + "learning_rate": 2.3320004294167587e-07, + "loss": 1.6421, + "step": 2426 + }, + { + "epoch": 1.8152580403889305, + "grad_norm": 2.18780772988903, + "learning_rate": 2.3133779824333847e-07, + "loss": 1.3969, + "step": 2427 + }, + { + "epoch": 1.8160059835452507, + "grad_norm": 2.3656139756492656, + "learning_rate": 2.2948284286337663e-07, + "loss": 1.5319, + "step": 2428 + }, + { + "epoch": 1.8167539267015707, + "grad_norm": 2.427152742268917, + "learning_rate": 2.2763517963725169e-07, + "loss": 1.3634, + "step": 2429 + }, + { + "epoch": 1.8175018698578906, + "grad_norm": 2.0212928717626326, + "learning_rate": 2.257948113892805e-07, + "loss": 1.4752, + "step": 2430 + }, + { + "epoch": 1.8182498130142108, + "grad_norm": 2.3080615206248325, + "learning_rate": 2.23961740932625e-07, + "loss": 1.4858, + "step": 2431 + }, + { + "epoch": 1.818997756170531, + "grad_norm": 2.266724158450612, + "learning_rate": 2.2213597106929608e-07, + "loss": 1.5216, + "step": 2432 + }, + { + "epoch": 1.8197456993268513, + "grad_norm": 2.315194041456356, + "learning_rate": 2.2031750459014345e-07, + "loss": 1.4459, + "step": 2433 + }, + { + "epoch": 1.8204936424831712, + "grad_norm": 2.173002333180081, + "learning_rate": 2.1850634427485085e-07, + "loss": 1.5157, + "step": 2434 + }, + { + "epoch": 1.8212415856394912, + "grad_norm": 2.023273273667913, + "learning_rate": 2.16702492891937e-07, + "loss": 1.6484, + "step": 2435 + }, + { + "epoch": 1.8219895287958114, + "grad_norm": 2.25756714438817, + "learning_rate": 2.1490595319874574e-07, + "loss": 1.5203, + "step": 2436 + }, + { + "epoch": 1.8227374719521316, + "grad_norm": 2.3704401352472186, + "learning_rate": 2.13116727941447e-07, + "loss": 1.4886, + "step": 2437 + }, + { + "epoch": 1.8234854151084519, + "grad_norm": 1.9188911857818283, + "learning_rate": 2.1133481985502747e-07, + "loss": 1.6691, + "step": 2438 + }, + { + "epoch": 1.8242333582647718, + "grad_norm": 2.5303597321320512, + "learning_rate": 2.0956023166328942e-07, + "loss": 1.6859, + "step": 2439 + }, + { + "epoch": 1.824981301421092, + "grad_norm": 1.994120302545141, + "learning_rate": 2.0779296607884857e-07, + "loss": 1.3873, + "step": 2440 + }, + { + "epoch": 1.825729244577412, + "grad_norm": 2.1021462404522495, + "learning_rate": 2.0603302580312402e-07, + "loss": 1.4373, + "step": 2441 + }, + { + "epoch": 1.8264771877337322, + "grad_norm": 2.4175922152277884, + "learning_rate": 2.0428041352633766e-07, + "loss": 1.6351, + "step": 2442 + }, + { + "epoch": 1.8272251308900525, + "grad_norm": 2.077638615023113, + "learning_rate": 2.0253513192751374e-07, + "loss": 1.5694, + "step": 2443 + }, + { + "epoch": 1.8279730740463724, + "grad_norm": 2.4672676233755944, + "learning_rate": 2.0079718367446544e-07, + "loss": 1.456, + "step": 2444 + }, + { + "epoch": 1.8287210172026926, + "grad_norm": 1.8512923948188096, + "learning_rate": 1.9906657142380105e-07, + "loss": 1.6352, + "step": 2445 + }, + { + "epoch": 1.8294689603590126, + "grad_norm": 2.045219502557189, + "learning_rate": 1.9734329782091277e-07, + "loss": 1.5429, + "step": 2446 + }, + { + "epoch": 1.8302169035153328, + "grad_norm": 1.8958097545597172, + "learning_rate": 1.9562736549997463e-07, + "loss": 1.3833, + "step": 2447 + }, + { + "epoch": 1.830964846671653, + "grad_norm": 2.107863631486667, + "learning_rate": 1.9391877708394125e-07, + "loss": 1.6978, + "step": 2448 + }, + { + "epoch": 1.831712789827973, + "grad_norm": 2.2721231327626885, + "learning_rate": 1.922175351845379e-07, + "loss": 1.4508, + "step": 2449 + }, + { + "epoch": 1.8324607329842932, + "grad_norm": 1.9738959123626467, + "learning_rate": 1.905236424022633e-07, + "loss": 1.355, + "step": 2450 + }, + { + "epoch": 1.8332086761406132, + "grad_norm": 2.5516654189474606, + "learning_rate": 1.8883710132638068e-07, + "loss": 1.6248, + "step": 2451 + }, + { + "epoch": 1.8339566192969334, + "grad_norm": 2.0802416413084734, + "learning_rate": 1.8715791453491562e-07, + "loss": 1.6369, + "step": 2452 + }, + { + "epoch": 1.8347045624532536, + "grad_norm": 1.8893170133970048, + "learning_rate": 1.8548608459465377e-07, + "loss": 1.5162, + "step": 2453 + }, + { + "epoch": 1.8354525056095736, + "grad_norm": 1.9762527201230442, + "learning_rate": 1.8382161406113208e-07, + "loss": 1.4369, + "step": 2454 + }, + { + "epoch": 1.8362004487658938, + "grad_norm": 2.132743685048627, + "learning_rate": 1.821645054786414e-07, + "loss": 1.6177, + "step": 2455 + }, + { + "epoch": 1.8369483919222138, + "grad_norm": 2.02169005180573, + "learning_rate": 1.8051476138021663e-07, + "loss": 1.3597, + "step": 2456 + }, + { + "epoch": 1.837696335078534, + "grad_norm": 1.9550928396379987, + "learning_rate": 1.7887238428763553e-07, + "loss": 1.5221, + "step": 2457 + }, + { + "epoch": 1.8384442782348542, + "grad_norm": 2.3698575018485344, + "learning_rate": 1.7723737671141662e-07, + "loss": 1.536, + "step": 2458 + }, + { + "epoch": 1.8391922213911742, + "grad_norm": 1.9233879570946208, + "learning_rate": 1.7560974115081176e-07, + "loss": 1.4209, + "step": 2459 + }, + { + "epoch": 1.8399401645474944, + "grad_norm": 2.0891450204197093, + "learning_rate": 1.7398948009380578e-07, + "loss": 1.4308, + "step": 2460 + }, + { + "epoch": 1.8406881077038144, + "grad_norm": 1.8656830899864787, + "learning_rate": 1.7237659601710866e-07, + "loss": 1.5637, + "step": 2461 + }, + { + "epoch": 1.8414360508601346, + "grad_norm": 1.8730461144031647, + "learning_rate": 1.7077109138615544e-07, + "loss": 1.7021, + "step": 2462 + }, + { + "epoch": 1.8421839940164548, + "grad_norm": 2.010613971451108, + "learning_rate": 1.691729686551008e-07, + "loss": 1.5149, + "step": 2463 + }, + { + "epoch": 1.8429319371727748, + "grad_norm": 2.1841305174750287, + "learning_rate": 1.6758223026681507e-07, + "loss": 1.7783, + "step": 2464 + }, + { + "epoch": 1.843679880329095, + "grad_norm": 2.2534317204664815, + "learning_rate": 1.659988786528821e-07, + "loss": 1.6514, + "step": 2465 + }, + { + "epoch": 1.844427823485415, + "grad_norm": 2.3230680872363214, + "learning_rate": 1.6442291623359364e-07, + "loss": 1.6212, + "step": 2466 + }, + { + "epoch": 1.8451757666417352, + "grad_norm": 2.0027527020382445, + "learning_rate": 1.6285434541794598e-07, + "loss": 1.4173, + "step": 2467 + }, + { + "epoch": 1.8459237097980554, + "grad_norm": 1.9111806727839125, + "learning_rate": 1.612931686036373e-07, + "loss": 1.2863, + "step": 2468 + }, + { + "epoch": 1.8466716529543756, + "grad_norm": 2.3308297471326114, + "learning_rate": 1.597393881770637e-07, + "loss": 1.3739, + "step": 2469 + }, + { + "epoch": 1.8474195961106956, + "grad_norm": 2.4492018969006257, + "learning_rate": 1.5819300651331416e-07, + "loss": 1.3417, + "step": 2470 + }, + { + "epoch": 1.8481675392670156, + "grad_norm": 2.086544571884775, + "learning_rate": 1.5665402597616842e-07, + "loss": 1.645, + "step": 2471 + }, + { + "epoch": 1.8489154824233358, + "grad_norm": 2.6788751592652162, + "learning_rate": 1.5512244891809413e-07, + "loss": 1.6024, + "step": 2472 + }, + { + "epoch": 1.849663425579656, + "grad_norm": 2.4786864838155154, + "learning_rate": 1.5359827768024138e-07, + "loss": 1.5882, + "step": 2473 + }, + { + "epoch": 1.8504113687359762, + "grad_norm": 2.084405925819725, + "learning_rate": 1.520815145924387e-07, + "loss": 1.6049, + "step": 2474 + }, + { + "epoch": 1.8511593118922962, + "grad_norm": 1.9890977995210823, + "learning_rate": 1.5057216197319147e-07, + "loss": 1.5055, + "step": 2475 + }, + { + "epoch": 1.8519072550486162, + "grad_norm": 2.0792081641103612, + "learning_rate": 1.4907022212967803e-07, + "loss": 1.4499, + "step": 2476 + }, + { + "epoch": 1.8526551982049364, + "grad_norm": 1.9878383333594964, + "learning_rate": 1.475756973577447e-07, + "loss": 1.6029, + "step": 2477 + }, + { + "epoch": 1.8534031413612566, + "grad_norm": 2.0269128457855725, + "learning_rate": 1.4608858994190344e-07, + "loss": 1.3996, + "step": 2478 + }, + { + "epoch": 1.8541510845175768, + "grad_norm": 1.870177896619736, + "learning_rate": 1.4460890215532986e-07, + "loss": 1.4707, + "step": 2479 + }, + { + "epoch": 1.8548990276738968, + "grad_norm": 3.0763184711521676, + "learning_rate": 1.4313663625985464e-07, + "loss": 1.3151, + "step": 2480 + }, + { + "epoch": 1.8556469708302168, + "grad_norm": 2.069248213020571, + "learning_rate": 1.4167179450596647e-07, + "loss": 1.5046, + "step": 2481 + }, + { + "epoch": 1.856394913986537, + "grad_norm": 2.6315663611807274, + "learning_rate": 1.4021437913280366e-07, + "loss": 1.3871, + "step": 2482 + }, + { + "epoch": 1.8571428571428572, + "grad_norm": 2.139446584978334, + "learning_rate": 1.3876439236815475e-07, + "loss": 1.6815, + "step": 2483 + }, + { + "epoch": 1.8578908002991774, + "grad_norm": 1.8958290997975762, + "learning_rate": 1.373218364284501e-07, + "loss": 1.3583, + "step": 2484 + }, + { + "epoch": 1.8586387434554974, + "grad_norm": 2.2274030381576084, + "learning_rate": 1.358867135187636e-07, + "loss": 1.6577, + "step": 2485 + }, + { + "epoch": 1.8593866866118174, + "grad_norm": 2.561927231029186, + "learning_rate": 1.3445902583280656e-07, + "loss": 1.4427, + "step": 2486 + }, + { + "epoch": 1.8601346297681376, + "grad_norm": 2.059722770448202, + "learning_rate": 1.3303877555292443e-07, + "loss": 1.6209, + "step": 2487 + }, + { + "epoch": 1.8608825729244578, + "grad_norm": 2.352762394350394, + "learning_rate": 1.316259648500956e-07, + "loss": 1.7115, + "step": 2488 + }, + { + "epoch": 1.861630516080778, + "grad_norm": 2.3447485493804745, + "learning_rate": 1.3022059588392367e-07, + "loss": 1.4313, + "step": 2489 + }, + { + "epoch": 1.862378459237098, + "grad_norm": 2.0124929413587815, + "learning_rate": 1.288226708026391e-07, + "loss": 1.6913, + "step": 2490 + }, + { + "epoch": 1.863126402393418, + "grad_norm": 1.882253708474767, + "learning_rate": 1.2743219174309208e-07, + "loss": 1.4561, + "step": 2491 + }, + { + "epoch": 1.8638743455497382, + "grad_norm": 2.3353878293694597, + "learning_rate": 1.2604916083075236e-07, + "loss": 1.6191, + "step": 2492 + }, + { + "epoch": 1.8646222887060584, + "grad_norm": 1.8120542122254377, + "learning_rate": 1.2467358017970498e-07, + "loss": 1.3571, + "step": 2493 + }, + { + "epoch": 1.8653702318623786, + "grad_norm": 2.0957708172029688, + "learning_rate": 1.2330545189264464e-07, + "loss": 1.8676, + "step": 2494 + }, + { + "epoch": 1.8661181750186986, + "grad_norm": 2.3705676217360243, + "learning_rate": 1.2194477806087458e-07, + "loss": 1.7173, + "step": 2495 + }, + { + "epoch": 1.8668661181750186, + "grad_norm": 2.2700505986313844, + "learning_rate": 1.2059156076430545e-07, + "loss": 1.3277, + "step": 2496 + }, + { + "epoch": 1.8676140613313388, + "grad_norm": 1.930280312465175, + "learning_rate": 1.192458020714482e-07, + "loss": 1.445, + "step": 2497 + }, + { + "epoch": 1.868362004487659, + "grad_norm": 2.042380081354311, + "learning_rate": 1.1790750403941231e-07, + "loss": 1.4421, + "step": 2498 + }, + { + "epoch": 1.8691099476439792, + "grad_norm": 3.0220932010479054, + "learning_rate": 1.1657666871390471e-07, + "loss": 1.6, + "step": 2499 + }, + { + "epoch": 1.8698578908002992, + "grad_norm": 2.3629709755556085, + "learning_rate": 1.1525329812922315e-07, + "loss": 1.5142, + "step": 2500 + }, + { + "epoch": 1.8706058339566192, + "grad_norm": 2.113366555054515, + "learning_rate": 1.1393739430825667e-07, + "loss": 1.5559, + "step": 2501 + }, + { + "epoch": 1.8713537771129394, + "grad_norm": 2.3210212329969853, + "learning_rate": 1.1262895926247852e-07, + "loss": 1.4588, + "step": 2502 + }, + { + "epoch": 1.8721017202692596, + "grad_norm": 2.3308502477583284, + "learning_rate": 1.1132799499194658e-07, + "loss": 1.3267, + "step": 2503 + }, + { + "epoch": 1.8728496634255798, + "grad_norm": 2.2112250519263834, + "learning_rate": 1.1003450348529898e-07, + "loss": 1.5781, + "step": 2504 + }, + { + "epoch": 1.8735976065818998, + "grad_norm": 2.235342543571027, + "learning_rate": 1.087484867197508e-07, + "loss": 1.3073, + "step": 2505 + }, + { + "epoch": 1.8743455497382198, + "grad_norm": 2.2925676911964565, + "learning_rate": 1.0746994666109234e-07, + "loss": 1.42, + "step": 2506 + }, + { + "epoch": 1.87509349289454, + "grad_norm": 2.2350292945175205, + "learning_rate": 1.0619888526368361e-07, + "loss": 1.4201, + "step": 2507 + }, + { + "epoch": 1.8758414360508602, + "grad_norm": 2.4048384231032887, + "learning_rate": 1.0493530447045263e-07, + "loss": 1.474, + "step": 2508 + }, + { + "epoch": 1.8765893792071804, + "grad_norm": 1.9065102668989864, + "learning_rate": 1.0367920621289496e-07, + "loss": 1.7467, + "step": 2509 + }, + { + "epoch": 1.8773373223635004, + "grad_norm": 2.2842321925044935, + "learning_rate": 1.0243059241106524e-07, + "loss": 1.3505, + "step": 2510 + }, + { + "epoch": 1.8780852655198204, + "grad_norm": 2.0692649018196785, + "learning_rate": 1.0118946497358006e-07, + "loss": 1.6113, + "step": 2511 + }, + { + "epoch": 1.8788332086761406, + "grad_norm": 2.1561240161257857, + "learning_rate": 9.995582579761243e-08, + "loss": 1.3589, + "step": 2512 + }, + { + "epoch": 1.8795811518324608, + "grad_norm": 2.200385097306354, + "learning_rate": 9.872967676888611e-08, + "loss": 1.6561, + "step": 2513 + }, + { + "epoch": 1.880329094988781, + "grad_norm": 1.8408564068288356, + "learning_rate": 9.751101976167964e-08, + "loss": 1.5851, + "step": 2514 + }, + { + "epoch": 1.881077038145101, + "grad_norm": 2.0321416427905343, + "learning_rate": 9.629985663881514e-08, + "loss": 1.7287, + "step": 2515 + }, + { + "epoch": 1.881824981301421, + "grad_norm": 3.9064538583393498, + "learning_rate": 9.509618925166276e-08, + "loss": 1.4152, + "step": 2516 + }, + { + "epoch": 1.8825729244577412, + "grad_norm": 2.0513623577852624, + "learning_rate": 9.390001944013293e-08, + "loss": 1.6303, + "step": 2517 + }, + { + "epoch": 1.8833208676140614, + "grad_norm": 1.9192755777504555, + "learning_rate": 9.271134903267642e-08, + "loss": 1.5531, + "step": 2518 + }, + { + "epoch": 1.8840688107703816, + "grad_norm": 2.0540158611585984, + "learning_rate": 9.153017984628143e-08, + "loss": 1.6352, + "step": 2519 + }, + { + "epoch": 1.8848167539267016, + "grad_norm": 1.9376549595384829, + "learning_rate": 9.035651368646647e-08, + "loss": 1.5551, + "step": 2520 + }, + { + "epoch": 1.8855646970830215, + "grad_norm": 1.8906019765617859, + "learning_rate": 8.919035234728479e-08, + "loss": 1.6306, + "step": 2521 + }, + { + "epoch": 1.8863126402393418, + "grad_norm": 1.9494320812906338, + "learning_rate": 8.803169761131547e-08, + "loss": 1.4863, + "step": 2522 + }, + { + "epoch": 1.887060583395662, + "grad_norm": 2.076352589613822, + "learning_rate": 8.688055124966454e-08, + "loss": 1.4592, + "step": 2523 + }, + { + "epoch": 1.8878085265519822, + "grad_norm": 11.758653309166217, + "learning_rate": 8.573691502195891e-08, + "loss": 1.3073, + "step": 2524 + }, + { + "epoch": 1.8885564697083022, + "grad_norm": 2.169220954909985, + "learning_rate": 8.460079067634853e-08, + "loss": 1.5424, + "step": 2525 + }, + { + "epoch": 1.8893044128646221, + "grad_norm": 2.016719173280026, + "learning_rate": 8.347217994949808e-08, + "loss": 1.7898, + "step": 2526 + }, + { + "epoch": 1.8900523560209423, + "grad_norm": 2.0032725931115434, + "learning_rate": 8.235108456658814e-08, + "loss": 1.3478, + "step": 2527 + }, + { + "epoch": 1.8908002991772626, + "grad_norm": 2.2394359553135676, + "learning_rate": 8.123750624131121e-08, + "loss": 1.4446, + "step": 2528 + }, + { + "epoch": 1.8915482423335828, + "grad_norm": 2.2471854737882526, + "learning_rate": 8.013144667587014e-08, + "loss": 1.429, + "step": 2529 + }, + { + "epoch": 1.8922961854899027, + "grad_norm": 2.0654881199161044, + "learning_rate": 7.903290756097304e-08, + "loss": 1.653, + "step": 2530 + }, + { + "epoch": 1.8930441286462227, + "grad_norm": 2.0941433102108413, + "learning_rate": 7.794189057583335e-08, + "loss": 1.6704, + "step": 2531 + }, + { + "epoch": 1.893792071802543, + "grad_norm": 2.1667503976283977, + "learning_rate": 7.685839738816703e-08, + "loss": 1.611, + "step": 2532 + }, + { + "epoch": 1.8945400149588631, + "grad_norm": 2.1908361321979353, + "learning_rate": 7.578242965418814e-08, + "loss": 1.7807, + "step": 2533 + }, + { + "epoch": 1.8952879581151834, + "grad_norm": 2.1393884326725687, + "learning_rate": 7.471398901860772e-08, + "loss": 1.6303, + "step": 2534 + }, + { + "epoch": 1.8960359012715033, + "grad_norm": 2.0563929352906807, + "learning_rate": 7.365307711463043e-08, + "loss": 1.8149, + "step": 2535 + }, + { + "epoch": 1.8967838444278233, + "grad_norm": 2.0092001339675676, + "learning_rate": 7.259969556395352e-08, + "loss": 1.6891, + "step": 2536 + }, + { + "epoch": 1.8975317875841435, + "grad_norm": 2.052406476984988, + "learning_rate": 7.155384597676285e-08, + "loss": 1.6826, + "step": 2537 + }, + { + "epoch": 1.8982797307404637, + "grad_norm": 2.1931106829553646, + "learning_rate": 7.051552995173128e-08, + "loss": 1.4058, + "step": 2538 + }, + { + "epoch": 1.899027673896784, + "grad_norm": 2.9969291467811567, + "learning_rate": 6.948474907601699e-08, + "loss": 1.321, + "step": 2539 + }, + { + "epoch": 1.899775617053104, + "grad_norm": 2.2543851961560764, + "learning_rate": 6.846150492525739e-08, + "loss": 1.1527, + "step": 2540 + }, + { + "epoch": 1.900523560209424, + "grad_norm": 2.189392631455757, + "learning_rate": 6.744579906357185e-08, + "loss": 1.4943, + "step": 2541 + }, + { + "epoch": 1.9012715033657441, + "grad_norm": 2.411026523544033, + "learning_rate": 6.643763304355566e-08, + "loss": 1.6952, + "step": 2542 + }, + { + "epoch": 1.9020194465220643, + "grad_norm": 2.370616949766179, + "learning_rate": 6.543700840627832e-08, + "loss": 1.6353, + "step": 2543 + }, + { + "epoch": 1.9027673896783845, + "grad_norm": 2.1291391915505122, + "learning_rate": 6.444392668128352e-08, + "loss": 1.3714, + "step": 2544 + }, + { + "epoch": 1.9035153328347045, + "grad_norm": 2.000153791861807, + "learning_rate": 6.34583893865831e-08, + "loss": 1.603, + "step": 2545 + }, + { + "epoch": 1.9042632759910245, + "grad_norm": 2.120763564157661, + "learning_rate": 6.248039802865702e-08, + "loss": 1.3873, + "step": 2546 + }, + { + "epoch": 1.9050112191473447, + "grad_norm": 2.105682350472335, + "learning_rate": 6.150995410245164e-08, + "loss": 1.6092, + "step": 2547 + }, + { + "epoch": 1.905759162303665, + "grad_norm": 1.9772200876314592, + "learning_rate": 6.054705909137426e-08, + "loss": 1.4124, + "step": 2548 + }, + { + "epoch": 1.9065071054599851, + "grad_norm": 2.0507577887778656, + "learning_rate": 5.959171446729584e-08, + "loss": 1.4907, + "step": 2549 + }, + { + "epoch": 1.9072550486163051, + "grad_norm": 1.9101201601536706, + "learning_rate": 5.864392169054378e-08, + "loss": 1.4017, + "step": 2550 + }, + { + "epoch": 1.9080029917726253, + "grad_norm": 2.149495957263669, + "learning_rate": 5.770368220990252e-08, + "loss": 1.4593, + "step": 2551 + }, + { + "epoch": 1.9087509349289453, + "grad_norm": 2.1016389038222476, + "learning_rate": 5.677099746261017e-08, + "loss": 1.7695, + "step": 2552 + }, + { + "epoch": 1.9094988780852655, + "grad_norm": 1.9878762245813333, + "learning_rate": 5.584586887435739e-08, + "loss": 1.5649, + "step": 2553 + }, + { + "epoch": 1.9102468212415857, + "grad_norm": 2.034922138096175, + "learning_rate": 5.492829785928466e-08, + "loss": 1.3326, + "step": 2554 + }, + { + "epoch": 1.9109947643979057, + "grad_norm": 2.0904400335465114, + "learning_rate": 5.401828581997948e-08, + "loss": 1.4136, + "step": 2555 + }, + { + "epoch": 1.911742707554226, + "grad_norm": 2.1077572082108205, + "learning_rate": 5.311583414747523e-08, + "loss": 1.5586, + "step": 2556 + }, + { + "epoch": 1.912490650710546, + "grad_norm": 2.1279721599589703, + "learning_rate": 5.222094422124846e-08, + "loss": 1.795, + "step": 2557 + }, + { + "epoch": 1.9132385938668661, + "grad_norm": 2.021807476946668, + "learning_rate": 5.133361740921605e-08, + "loss": 1.6109, + "step": 2558 + }, + { + "epoch": 1.9139865370231863, + "grad_norm": 2.061084155726516, + "learning_rate": 5.0453855067735815e-08, + "loss": 1.487, + "step": 2559 + }, + { + "epoch": 1.9147344801795063, + "grad_norm": 1.9532563659104871, + "learning_rate": 4.9581658541602015e-08, + "loss": 1.6234, + "step": 2560 + }, + { + "epoch": 1.9154824233358265, + "grad_norm": 2.1681292726247663, + "learning_rate": 4.871702916404208e-08, + "loss": 1.7606, + "step": 2561 + }, + { + "epoch": 1.9162303664921465, + "grad_norm": 2.46127736704092, + "learning_rate": 4.7859968256719344e-08, + "loss": 1.1645, + "step": 2562 + }, + { + "epoch": 1.9169783096484667, + "grad_norm": 1.923190669728051, + "learning_rate": 4.701047712972584e-08, + "loss": 1.4212, + "step": 2563 + }, + { + "epoch": 1.917726252804787, + "grad_norm": 2.1173811193924164, + "learning_rate": 4.6168557081582854e-08, + "loss": 1.401, + "step": 2564 + }, + { + "epoch": 1.918474195961107, + "grad_norm": 2.3581555737571294, + "learning_rate": 4.533420939923927e-08, + "loss": 1.4333, + "step": 2565 + }, + { + "epoch": 1.9192221391174271, + "grad_norm": 2.0952299424546723, + "learning_rate": 4.450743535806934e-08, + "loss": 1.6925, + "step": 2566 + }, + { + "epoch": 1.919970082273747, + "grad_norm": 2.074253110994616, + "learning_rate": 4.368823622186935e-08, + "loss": 1.7039, + "step": 2567 + }, + { + "epoch": 1.9207180254300673, + "grad_norm": 2.2131426331068287, + "learning_rate": 4.287661324285652e-08, + "loss": 1.4525, + "step": 2568 + }, + { + "epoch": 1.9214659685863875, + "grad_norm": 2.007225735935588, + "learning_rate": 4.207256766166845e-08, + "loss": 1.7826, + "step": 2569 + }, + { + "epoch": 1.9222139117427075, + "grad_norm": 2.7312538191974025, + "learning_rate": 4.127610070735921e-08, + "loss": 1.5068, + "step": 2570 + }, + { + "epoch": 1.9229618548990277, + "grad_norm": 1.8446487150854436, + "learning_rate": 4.0487213597397715e-08, + "loss": 1.2583, + "step": 2571 + }, + { + "epoch": 1.9237097980553477, + "grad_norm": 2.4070462410342084, + "learning_rate": 3.970590753766712e-08, + "loss": 1.6542, + "step": 2572 + }, + { + "epoch": 1.924457741211668, + "grad_norm": 2.049305177835651, + "learning_rate": 3.8932183722463765e-08, + "loss": 1.4355, + "step": 2573 + }, + { + "epoch": 1.9252056843679881, + "grad_norm": 3.448410094160127, + "learning_rate": 3.8166043334490475e-08, + "loss": 1.9686, + "step": 2574 + }, + { + "epoch": 1.925953627524308, + "grad_norm": 2.0931464566968243, + "learning_rate": 3.7407487544861565e-08, + "loss": 1.4219, + "step": 2575 + }, + { + "epoch": 1.9267015706806283, + "grad_norm": 1.941232077692635, + "learning_rate": 3.665651751309451e-08, + "loss": 1.6122, + "step": 2576 + }, + { + "epoch": 1.9274495138369483, + "grad_norm": 1.9593237082589532, + "learning_rate": 3.5913134387114414e-08, + "loss": 1.4996, + "step": 2577 + }, + { + "epoch": 1.9281974569932685, + "grad_norm": 2.1376789675853196, + "learning_rate": 3.517733930324674e-08, + "loss": 1.2579, + "step": 2578 + }, + { + "epoch": 1.9289454001495887, + "grad_norm": 1.921017518960848, + "learning_rate": 3.444913338621847e-08, + "loss": 1.7095, + "step": 2579 + }, + { + "epoch": 1.929693343305909, + "grad_norm": 2.0337286238272045, + "learning_rate": 3.372851774915697e-08, + "loss": 1.437, + "step": 2580 + }, + { + "epoch": 1.930441286462229, + "grad_norm": 2.0492885128334692, + "learning_rate": 3.301549349358668e-08, + "loss": 1.6249, + "step": 2581 + }, + { + "epoch": 1.9311892296185489, + "grad_norm": 2.491805426541738, + "learning_rate": 3.2310061709426296e-08, + "loss": 1.3983, + "step": 2582 + }, + { + "epoch": 1.931937172774869, + "grad_norm": 2.0217660976382965, + "learning_rate": 3.16122234749916e-08, + "loss": 1.5697, + "step": 2583 + }, + { + "epoch": 1.9326851159311893, + "grad_norm": 1.9891517176645082, + "learning_rate": 3.092197985698875e-08, + "loss": 1.4296, + "step": 2584 + }, + { + "epoch": 1.9334330590875095, + "grad_norm": 2.0693555781486643, + "learning_rate": 3.023933191051598e-08, + "loss": 1.638, + "step": 2585 + }, + { + "epoch": 1.9341810022438295, + "grad_norm": 2.173553417050226, + "learning_rate": 2.9564280679060255e-08, + "loss": 1.5776, + "step": 2586 + }, + { + "epoch": 1.9349289454001495, + "grad_norm": 2.1116174763640023, + "learning_rate": 2.8896827194496713e-08, + "loss": 1.5417, + "step": 2587 + }, + { + "epoch": 1.9356768885564697, + "grad_norm": 2.005349497354917, + "learning_rate": 2.823697247708701e-08, + "loss": 1.5073, + "step": 2588 + }, + { + "epoch": 1.93642483171279, + "grad_norm": 2.173594816109451, + "learning_rate": 2.758471753547598e-08, + "loss": 1.5407, + "step": 2589 + }, + { + "epoch": 1.93717277486911, + "grad_norm": 1.9770234604951777, + "learning_rate": 2.6940063366693303e-08, + "loss": 1.5238, + "step": 2590 + }, + { + "epoch": 1.93792071802543, + "grad_norm": 2.2447870046917515, + "learning_rate": 2.6303010956149065e-08, + "loss": 1.5141, + "step": 2591 + }, + { + "epoch": 1.93866866118175, + "grad_norm": 1.832034117615565, + "learning_rate": 2.5673561277633764e-08, + "loss": 1.7258, + "step": 2592 + }, + { + "epoch": 1.9394166043380703, + "grad_norm": 1.9015803490517178, + "learning_rate": 2.5051715293316624e-08, + "loss": 1.7561, + "step": 2593 + }, + { + "epoch": 1.9401645474943905, + "grad_norm": 2.288933970751832, + "learning_rate": 2.4437473953743962e-08, + "loss": 1.6154, + "step": 2594 + }, + { + "epoch": 1.9409124906507107, + "grad_norm": 1.887381298154917, + "learning_rate": 2.3830838197836937e-08, + "loss": 1.3549, + "step": 2595 + }, + { + "epoch": 1.9416604338070307, + "grad_norm": 1.9034380822963641, + "learning_rate": 2.3231808952891566e-08, + "loss": 1.0891, + "step": 2596 + }, + { + "epoch": 1.9424083769633507, + "grad_norm": 2.418301838222331, + "learning_rate": 2.264038713457706e-08, + "loss": 1.4152, + "step": 2597 + }, + { + "epoch": 1.9431563201196709, + "grad_norm": 2.158313946687743, + "learning_rate": 2.2056573646933033e-08, + "loss": 1.3225, + "step": 2598 + }, + { + "epoch": 1.943904263275991, + "grad_norm": 1.9071908339752015, + "learning_rate": 2.148036938237008e-08, + "loss": 1.4265, + "step": 2599 + }, + { + "epoch": 1.9446522064323113, + "grad_norm": 2.1939012212361426, + "learning_rate": 2.091177522166643e-08, + "loss": 1.278, + "step": 2600 + }, + { + "epoch": 1.9454001495886313, + "grad_norm": 2.1685470957785107, + "learning_rate": 2.0350792033969057e-08, + "loss": 1.499, + "step": 2601 + }, + { + "epoch": 1.9461480927449513, + "grad_norm": 2.1090109803804307, + "learning_rate": 1.9797420676788692e-08, + "loss": 1.5162, + "step": 2602 + }, + { + "epoch": 1.9468960359012715, + "grad_norm": 2.21845659249394, + "learning_rate": 1.925166199600259e-08, + "loss": 1.657, + "step": 2603 + }, + { + "epoch": 1.9476439790575917, + "grad_norm": 1.8867533530028455, + "learning_rate": 1.8713516825851207e-08, + "loss": 1.3365, + "step": 2604 + }, + { + "epoch": 1.9483919222139119, + "grad_norm": 2.2119589275674123, + "learning_rate": 1.818298598893542e-08, + "loss": 1.47, + "step": 2605 + }, + { + "epoch": 1.9491398653702319, + "grad_norm": 2.1452924917918343, + "learning_rate": 1.7660070296219856e-08, + "loss": 1.6347, + "step": 2606 + }, + { + "epoch": 1.9498878085265519, + "grad_norm": 2.1590613464475163, + "learning_rate": 1.7144770547025125e-08, + "loss": 1.5485, + "step": 2607 + }, + { + "epoch": 1.950635751682872, + "grad_norm": 2.1568498896416846, + "learning_rate": 1.6637087529033925e-08, + "loss": 1.6246, + "step": 2608 + }, + { + "epoch": 1.9513836948391923, + "grad_norm": 2.2516754884339574, + "learning_rate": 1.613702201828271e-08, + "loss": 1.6034, + "step": 2609 + }, + { + "epoch": 1.9521316379955125, + "grad_norm": 2.11429115257337, + "learning_rate": 1.5644574779166698e-08, + "loss": 1.5933, + "step": 2610 + }, + { + "epoch": 1.9528795811518325, + "grad_norm": 1.9961727380844603, + "learning_rate": 1.51597465644332e-08, + "loss": 1.674, + "step": 2611 + }, + { + "epoch": 1.9536275243081525, + "grad_norm": 2.019161520630904, + "learning_rate": 1.4682538115186052e-08, + "loss": 1.4463, + "step": 2612 + }, + { + "epoch": 1.9543754674644727, + "grad_norm": 2.1433439781163433, + "learning_rate": 1.421295016088009e-08, + "loss": 1.5443, + "step": 2613 + }, + { + "epoch": 1.9551234106207929, + "grad_norm": 2.0338978363774003, + "learning_rate": 1.3750983419321129e-08, + "loss": 1.4779, + "step": 2614 + }, + { + "epoch": 1.955871353777113, + "grad_norm": 1.9940800312194462, + "learning_rate": 1.3296638596665968e-08, + "loss": 1.8729, + "step": 2615 + }, + { + "epoch": 1.956619296933433, + "grad_norm": 2.0581854168134677, + "learning_rate": 1.2849916387420724e-08, + "loss": 1.6849, + "step": 2616 + }, + { + "epoch": 1.957367240089753, + "grad_norm": 2.204370376432356, + "learning_rate": 1.241081747443862e-08, + "loss": 1.396, + "step": 2617 + }, + { + "epoch": 1.9581151832460733, + "grad_norm": 1.9982629327312829, + "learning_rate": 1.1979342528922189e-08, + "loss": 1.3521, + "step": 2618 + }, + { + "epoch": 1.9588631264023935, + "grad_norm": 2.22609025836769, + "learning_rate": 1.1555492210418295e-08, + "loss": 1.7076, + "step": 2619 + }, + { + "epoch": 1.9596110695587137, + "grad_norm": 1.9265211876276735, + "learning_rate": 1.113926716681868e-08, + "loss": 1.2411, + "step": 2620 + }, + { + "epoch": 1.9603590127150337, + "grad_norm": 2.8981222696501043, + "learning_rate": 1.0730668034360514e-08, + "loss": 1.389, + "step": 2621 + }, + { + "epoch": 1.9611069558713536, + "grad_norm": 2.1513616919725123, + "learning_rate": 1.0329695437623077e-08, + "loss": 1.4699, + "step": 2622 + }, + { + "epoch": 1.9618548990276738, + "grad_norm": 2.0298385610893934, + "learning_rate": 9.936349989528305e-09, + "loss": 1.6175, + "step": 2623 + }, + { + "epoch": 1.962602842183994, + "grad_norm": 2.3925048333558228, + "learning_rate": 9.550632291339124e-09, + "loss": 1.7143, + "step": 2624 + }, + { + "epoch": 1.9633507853403143, + "grad_norm": 2.108136899799544, + "learning_rate": 9.1725429326589e-09, + "loss": 1.5954, + "step": 2625 + }, + { + "epoch": 1.9640987284966342, + "grad_norm": 1.9942546144516533, + "learning_rate": 8.802082491430331e-09, + "loss": 1.486, + "step": 2626 + }, + { + "epoch": 1.9648466716529542, + "grad_norm": 1.9467039355122975, + "learning_rate": 8.439251533934323e-09, + "loss": 1.3811, + "step": 2627 + }, + { + "epoch": 1.9655946148092744, + "grad_norm": 2.161247297257024, + "learning_rate": 8.084050614791117e-09, + "loss": 1.6054, + "step": 2628 + }, + { + "epoch": 1.9663425579655947, + "grad_norm": 2.193505732532727, + "learning_rate": 7.736480276954728e-09, + "loss": 1.564, + "step": 2629 + }, + { + "epoch": 1.9670905011219149, + "grad_norm": 2.223528246992852, + "learning_rate": 7.3965410517179426e-09, + "loss": 1.4947, + "step": 2630 + }, + { + "epoch": 1.9678384442782348, + "grad_norm": 2.2751117091136313, + "learning_rate": 7.064233458707326e-09, + "loss": 1.3088, + "step": 2631 + }, + { + "epoch": 1.9685863874345548, + "grad_norm": 2.1666676546732915, + "learning_rate": 6.739558005884883e-09, + "loss": 1.4731, + "step": 2632 + }, + { + "epoch": 1.969334330590875, + "grad_norm": 1.9013755082249237, + "learning_rate": 6.4225151895447315e-09, + "loss": 1.4462, + "step": 2633 + }, + { + "epoch": 1.9700822737471952, + "grad_norm": 2.1481257142846166, + "learning_rate": 6.113105494314209e-09, + "loss": 1.6358, + "step": 2634 + }, + { + "epoch": 1.9708302169035155, + "grad_norm": 2.029582214794534, + "learning_rate": 5.81132939315443e-09, + "loss": 1.4807, + "step": 2635 + }, + { + "epoch": 1.9715781600598354, + "grad_norm": 2.557505723207599, + "learning_rate": 5.517187347355846e-09, + "loss": 1.3978, + "step": 2636 + }, + { + "epoch": 1.9723261032161554, + "grad_norm": 1.8490627750401885, + "learning_rate": 5.230679806540462e-09, + "loss": 1.546, + "step": 2637 + }, + { + "epoch": 1.9730740463724756, + "grad_norm": 1.8659813822785831, + "learning_rate": 4.951807208660176e-09, + "loss": 1.5311, + "step": 2638 + }, + { + "epoch": 1.9738219895287958, + "grad_norm": 2.090553921178187, + "learning_rate": 4.6805699799967744e-09, + "loss": 1.3436, + "step": 2639 + }, + { + "epoch": 1.974569932685116, + "grad_norm": 2.059478590007305, + "learning_rate": 4.41696853515916e-09, + "loss": 1.503, + "step": 2640 + }, + { + "epoch": 1.975317875841436, + "grad_norm": 2.3002126360483675, + "learning_rate": 4.161003277085574e-09, + "loss": 1.4643, + "step": 2641 + }, + { + "epoch": 1.976065818997756, + "grad_norm": 1.9282562858752212, + "learning_rate": 3.912674597041921e-09, + "loss": 1.5647, + "step": 2642 + }, + { + "epoch": 1.9768137621540762, + "grad_norm": 2.009269025709866, + "learning_rate": 3.671982874619562e-09, + "loss": 1.4914, + "step": 2643 + }, + { + "epoch": 1.9775617053103964, + "grad_norm": 2.3374949141269807, + "learning_rate": 3.4389284777375242e-09, + "loss": 1.3917, + "step": 2644 + }, + { + "epoch": 1.9783096484667166, + "grad_norm": 2.0455675175404893, + "learning_rate": 3.213511762639732e-09, + "loss": 1.4019, + "step": 2645 + }, + { + "epoch": 1.9790575916230366, + "grad_norm": 2.041083120305976, + "learning_rate": 2.995733073895557e-09, + "loss": 1.4383, + "step": 2646 + }, + { + "epoch": 1.9798055347793566, + "grad_norm": 1.8917304987640313, + "learning_rate": 2.785592744398713e-09, + "loss": 1.5047, + "step": 2647 + }, + { + "epoch": 1.9805534779356768, + "grad_norm": 2.5207939740767507, + "learning_rate": 2.5830910953666966e-09, + "loss": 1.4325, + "step": 2648 + }, + { + "epoch": 1.981301421091997, + "grad_norm": 2.1694635794336734, + "learning_rate": 2.3882284363413444e-09, + "loss": 1.4762, + "step": 2649 + }, + { + "epoch": 1.9820493642483172, + "grad_norm": 1.9062021670742357, + "learning_rate": 2.2010050651871673e-09, + "loss": 1.3125, + "step": 2650 + }, + { + "epoch": 1.9827973074046372, + "grad_norm": 2.1501170648770334, + "learning_rate": 2.0214212680919056e-09, + "loss": 1.6387, + "step": 2651 + }, + { + "epoch": 1.9835452505609572, + "grad_norm": 2.153597501336623, + "learning_rate": 1.849477319564863e-09, + "loss": 1.7669, + "step": 2652 + }, + { + "epoch": 1.9842931937172774, + "grad_norm": 2.2916079628693953, + "learning_rate": 1.6851734824380184e-09, + "loss": 1.7193, + "step": 2653 + }, + { + "epoch": 1.9850411368735976, + "grad_norm": 2.1651654883928644, + "learning_rate": 1.5285100078632487e-09, + "loss": 1.4158, + "step": 2654 + }, + { + "epoch": 1.9857890800299178, + "grad_norm": 2.1141747838921283, + "learning_rate": 1.37948713531455e-09, + "loss": 1.6102, + "step": 2655 + }, + { + "epoch": 1.9865370231862378, + "grad_norm": 2.237128920703023, + "learning_rate": 1.2381050925869276e-09, + "loss": 1.4536, + "step": 2656 + }, + { + "epoch": 1.9872849663425578, + "grad_norm": 1.8641813737236854, + "learning_rate": 1.104364095795285e-09, + "loss": 1.5274, + "step": 2657 + }, + { + "epoch": 1.988032909498878, + "grad_norm": 2.167453560592637, + "learning_rate": 9.782643493744248e-10, + "loss": 1.7216, + "step": 2658 + }, + { + "epoch": 1.9887808526551982, + "grad_norm": 1.9508324252156821, + "learning_rate": 8.598060460779378e-10, + "loss": 1.4567, + "step": 2659 + }, + { + "epoch": 1.9895287958115184, + "grad_norm": 2.455415546261433, + "learning_rate": 7.48989366980979e-10, + "loss": 1.2707, + "step": 2660 + }, + { + "epoch": 1.9902767389678384, + "grad_norm": 2.475744960442436, + "learning_rate": 6.45814481475826e-10, + "loss": 1.5457, + "step": 2661 + }, + { + "epoch": 1.9910246821241586, + "grad_norm": 2.179209098774429, + "learning_rate": 5.50281547275211e-10, + "loss": 1.4134, + "step": 2662 + }, + { + "epoch": 1.9917726252804786, + "grad_norm": 2.0326746339819, + "learning_rate": 4.623907104084335e-10, + "loss": 1.6649, + "step": 2663 + }, + { + "epoch": 1.9925205684367988, + "grad_norm": 2.3436316589718955, + "learning_rate": 3.821421052246921e-10, + "loss": 1.5728, + "step": 2664 + }, + { + "epoch": 1.993268511593119, + "grad_norm": 1.89255873893591, + "learning_rate": 3.095358543914184e-10, + "loss": 1.4851, + "step": 2665 + }, + { + "epoch": 1.994016454749439, + "grad_norm": 34.155944453561524, + "learning_rate": 2.4457206889316744e-10, + "loss": 1.3396, + "step": 2666 + }, + { + "epoch": 1.9947643979057592, + "grad_norm": 1.9375710937375825, + "learning_rate": 1.872508480332824e-10, + "loss": 1.8091, + "step": 2667 + }, + { + "epoch": 1.9955123410620792, + "grad_norm": 2.589537580001781, + "learning_rate": 1.3757227943167472e-10, + "loss": 1.4284, + "step": 2668 + }, + { + "epoch": 1.9962602842183994, + "grad_norm": 1.87745403020336, + "learning_rate": 9.553643902704412e-11, + "loss": 1.64, + "step": 2669 + }, + { + "epoch": 1.9970082273747196, + "grad_norm": 1.94411372096112, + "learning_rate": 6.114339107465838e-11, + "loss": 1.403, + "step": 2670 + }, + { + "epoch": 1.9977561705310396, + "grad_norm": 1.9943352713849216, + "learning_rate": 3.43931881469084e-11, + "loss": 1.5464, + "step": 2671 + }, + { + "epoch": 1.9985041136873598, + "grad_norm": 1.9829695217515748, + "learning_rate": 1.5285871134418372e-11, + "loss": 1.5093, + "step": 2672 + }, + { + "epoch": 1.9992520568436798, + "grad_norm": 1.8004598327761268, + "learning_rate": 3.8214692438254266e-12, + "loss": 1.459, + "step": 2673 + }, + { + "epoch": 2.0, + "grad_norm": 2.012020684857708, + "learning_rate": 0.0, + "loss": 1.424, + "step": 2674 + } + ], + "logging_steps": 1, + "max_steps": 2674, + "num_input_tokens_seen": 0, + "num_train_epochs": 2, + "save_steps": 335, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 872122637352960.0, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}