| { | |
| "best_metric": 0.4462641775608063, | |
| "best_model_checkpoint": "cbb-3b/checkpoint-366", | |
| "epoch": 0.9993174061433447, | |
| "eval_steps": 500, | |
| "global_step": 366, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0027303754266211604, | |
| "grad_norm": 0.7549694776535034, | |
| "learning_rate": 1.360544217687075e-06, | |
| "loss": 1.2225, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005460750853242321, | |
| "grad_norm": 0.7538214325904846, | |
| "learning_rate": 2.72108843537415e-06, | |
| "loss": 1.2103, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.008191126279863481, | |
| "grad_norm": 0.7328954935073853, | |
| "learning_rate": 4.081632653061224e-06, | |
| "loss": 1.1858, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.010921501706484642, | |
| "grad_norm": 0.7359272837638855, | |
| "learning_rate": 5.4421768707483e-06, | |
| "loss": 1.1885, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.013651877133105802, | |
| "grad_norm": 0.740386426448822, | |
| "learning_rate": 6.802721088435375e-06, | |
| "loss": 1.1781, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.016382252559726963, | |
| "grad_norm": 0.6984951496124268, | |
| "learning_rate": 8.163265306122448e-06, | |
| "loss": 1.1395, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.01911262798634812, | |
| "grad_norm": 0.6689624786376953, | |
| "learning_rate": 9.523809523809523e-06, | |
| "loss": 1.137, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.021843003412969283, | |
| "grad_norm": 0.6134174466133118, | |
| "learning_rate": 1.08843537414966e-05, | |
| "loss": 1.1531, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.024573378839590442, | |
| "grad_norm": 0.5647606253623962, | |
| "learning_rate": 1.2244897959183674e-05, | |
| "loss": 1.1201, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.027303754266211604, | |
| "grad_norm": 0.541833221912384, | |
| "learning_rate": 1.360544217687075e-05, | |
| "loss": 1.0989, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.030034129692832763, | |
| "grad_norm": 0.4785626232624054, | |
| "learning_rate": 1.4965986394557824e-05, | |
| "loss": 1.0664, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.032764505119453925, | |
| "grad_norm": 0.42421552538871765, | |
| "learning_rate": 1.6326530612244897e-05, | |
| "loss": 1.057, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.03549488054607509, | |
| "grad_norm": 0.384870707988739, | |
| "learning_rate": 1.7687074829931973e-05, | |
| "loss": 0.9794, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.03822525597269624, | |
| "grad_norm": 0.31449463963508606, | |
| "learning_rate": 1.9047619047619046e-05, | |
| "loss": 0.9485, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.040955631399317405, | |
| "grad_norm": 0.29094135761260986, | |
| "learning_rate": 2.0408163265306123e-05, | |
| "loss": 0.9581, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04368600682593857, | |
| "grad_norm": 0.2500893771648407, | |
| "learning_rate": 2.17687074829932e-05, | |
| "loss": 0.9363, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04641638225255973, | |
| "grad_norm": 0.2445881962776184, | |
| "learning_rate": 2.3129251700680275e-05, | |
| "loss": 0.9186, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.049146757679180884, | |
| "grad_norm": 0.2477860301733017, | |
| "learning_rate": 2.448979591836735e-05, | |
| "loss": 0.9099, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.05187713310580205, | |
| "grad_norm": 0.24853268265724182, | |
| "learning_rate": 2.5850340136054425e-05, | |
| "loss": 0.912, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.05460750853242321, | |
| "grad_norm": 0.22501873970031738, | |
| "learning_rate": 2.72108843537415e-05, | |
| "loss": 0.8836, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05733788395904437, | |
| "grad_norm": 0.21223071217536926, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 0.8651, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.060068259385665526, | |
| "grad_norm": 0.20172430574893951, | |
| "learning_rate": 2.9931972789115647e-05, | |
| "loss": 0.8393, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.06279863481228669, | |
| "grad_norm": 0.17902718484401703, | |
| "learning_rate": 3.1292517006802724e-05, | |
| "loss": 0.8033, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.06552901023890785, | |
| "grad_norm": 0.1813097447156906, | |
| "learning_rate": 3.265306122448979e-05, | |
| "loss": 0.8152, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06825938566552901, | |
| "grad_norm": 0.19280143082141876, | |
| "learning_rate": 3.401360544217687e-05, | |
| "loss": 0.8051, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.07098976109215017, | |
| "grad_norm": 0.17157189548015594, | |
| "learning_rate": 3.5374149659863946e-05, | |
| "loss": 0.794, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.07372013651877134, | |
| "grad_norm": 0.1467738002538681, | |
| "learning_rate": 3.673469387755102e-05, | |
| "loss": 0.7874, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.07645051194539249, | |
| "grad_norm": 0.13913457095623016, | |
| "learning_rate": 3.809523809523809e-05, | |
| "loss": 0.7519, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.07918088737201365, | |
| "grad_norm": 0.13179022073745728, | |
| "learning_rate": 3.945578231292517e-05, | |
| "loss": 0.76, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.08191126279863481, | |
| "grad_norm": 0.1376553773880005, | |
| "learning_rate": 4.0816326530612245e-05, | |
| "loss": 0.7369, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08464163822525597, | |
| "grad_norm": 0.14040575921535492, | |
| "learning_rate": 4.217687074829932e-05, | |
| "loss": 0.7463, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.08737201365187713, | |
| "grad_norm": 0.13217338919639587, | |
| "learning_rate": 4.35374149659864e-05, | |
| "loss": 0.7298, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0901023890784983, | |
| "grad_norm": 0.11285194754600525, | |
| "learning_rate": 4.4897959183673474e-05, | |
| "loss": 0.7134, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.09283276450511946, | |
| "grad_norm": 0.10098642110824585, | |
| "learning_rate": 4.625850340136055e-05, | |
| "loss": 0.7238, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.09556313993174062, | |
| "grad_norm": 0.10341370850801468, | |
| "learning_rate": 4.761904761904762e-05, | |
| "loss": 0.6908, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.09829351535836177, | |
| "grad_norm": 0.09662918746471405, | |
| "learning_rate": 4.89795918367347e-05, | |
| "loss": 0.7, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.10102389078498293, | |
| "grad_norm": 0.09548471122980118, | |
| "learning_rate": 5.034013605442177e-05, | |
| "loss": 0.7207, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.1037542662116041, | |
| "grad_norm": 0.09512269496917725, | |
| "learning_rate": 5.170068027210885e-05, | |
| "loss": 0.7016, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.10648464163822526, | |
| "grad_norm": 0.0912129282951355, | |
| "learning_rate": 5.3061224489795926e-05, | |
| "loss": 0.6891, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.10921501706484642, | |
| "grad_norm": 0.08661182224750519, | |
| "learning_rate": 5.4421768707483e-05, | |
| "loss": 0.6982, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.11194539249146758, | |
| "grad_norm": 0.09124922007322311, | |
| "learning_rate": 5.5782312925170065e-05, | |
| "loss": 0.7051, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.11467576791808874, | |
| "grad_norm": 0.09174500405788422, | |
| "learning_rate": 5.714285714285714e-05, | |
| "loss": 0.6978, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.1174061433447099, | |
| "grad_norm": 0.0679943636059761, | |
| "learning_rate": 5.850340136054422e-05, | |
| "loss": 0.6889, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.12013651877133105, | |
| "grad_norm": 0.07204238325357437, | |
| "learning_rate": 5.9863945578231295e-05, | |
| "loss": 0.704, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.12286689419795221, | |
| "grad_norm": 0.08089234679937363, | |
| "learning_rate": 6.122448979591838e-05, | |
| "loss": 0.6838, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.12559726962457338, | |
| "grad_norm": 0.09053023904561996, | |
| "learning_rate": 6.258503401360545e-05, | |
| "loss": 0.6754, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.12832764505119454, | |
| "grad_norm": 0.07513958215713501, | |
| "learning_rate": 6.394557823129253e-05, | |
| "loss": 0.6894, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1310580204778157, | |
| "grad_norm": 0.07480401545763016, | |
| "learning_rate": 6.530612244897959e-05, | |
| "loss": 0.6809, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.13378839590443686, | |
| "grad_norm": 0.07617643475532532, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 0.697, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.13651877133105803, | |
| "grad_norm": 0.06744271516799927, | |
| "learning_rate": 6.802721088435374e-05, | |
| "loss": 0.6921, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1392491467576792, | |
| "grad_norm": 0.07185206562280655, | |
| "learning_rate": 6.938775510204082e-05, | |
| "loss": 0.6536, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.14197952218430035, | |
| "grad_norm": 0.07255382090806961, | |
| "learning_rate": 7.074829931972789e-05, | |
| "loss": 0.653, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.1447098976109215, | |
| "grad_norm": 0.07474930584430695, | |
| "learning_rate": 7.210884353741498e-05, | |
| "loss": 0.6888, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.14744027303754267, | |
| "grad_norm": 0.0754467323422432, | |
| "learning_rate": 7.346938775510205e-05, | |
| "loss": 0.6818, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.15017064846416384, | |
| "grad_norm": 0.07726683467626572, | |
| "learning_rate": 7.482993197278913e-05, | |
| "loss": 0.6835, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.15290102389078497, | |
| "grad_norm": 0.07462974637746811, | |
| "learning_rate": 7.619047619047618e-05, | |
| "loss": 0.667, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.15563139931740613, | |
| "grad_norm": 0.06939647346735, | |
| "learning_rate": 7.755102040816327e-05, | |
| "loss": 0.6668, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.1583617747440273, | |
| "grad_norm": 0.08218149840831757, | |
| "learning_rate": 7.891156462585034e-05, | |
| "loss": 0.6762, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.16109215017064846, | |
| "grad_norm": 0.0838819146156311, | |
| "learning_rate": 8.027210884353742e-05, | |
| "loss": 0.6685, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.16382252559726962, | |
| "grad_norm": 0.07441603392362595, | |
| "learning_rate": 8.163265306122449e-05, | |
| "loss": 0.6573, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.16655290102389078, | |
| "grad_norm": 0.0746053010225296, | |
| "learning_rate": 8.299319727891157e-05, | |
| "loss": 0.6582, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.16928327645051194, | |
| "grad_norm": 0.08602144569158554, | |
| "learning_rate": 8.435374149659864e-05, | |
| "loss": 0.6547, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.1720136518771331, | |
| "grad_norm": 0.08236663043498993, | |
| "learning_rate": 8.571428571428571e-05, | |
| "loss": 0.6081, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.17474402730375427, | |
| "grad_norm": 0.08744888752698898, | |
| "learning_rate": 8.70748299319728e-05, | |
| "loss": 0.6576, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.17747440273037543, | |
| "grad_norm": 0.08321461081504822, | |
| "learning_rate": 8.843537414965987e-05, | |
| "loss": 0.6137, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.1802047781569966, | |
| "grad_norm": 0.08639347553253174, | |
| "learning_rate": 8.979591836734695e-05, | |
| "loss": 0.6579, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.18293515358361775, | |
| "grad_norm": 0.09154847264289856, | |
| "learning_rate": 9.115646258503402e-05, | |
| "loss": 0.6391, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.18566552901023892, | |
| "grad_norm": 0.1094379723072052, | |
| "learning_rate": 9.25170068027211e-05, | |
| "loss": 0.61, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.18839590443686008, | |
| "grad_norm": 0.11089900881052017, | |
| "learning_rate": 9.387755102040817e-05, | |
| "loss": 0.6452, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.19112627986348124, | |
| "grad_norm": 0.11615785956382751, | |
| "learning_rate": 9.523809523809524e-05, | |
| "loss": 0.6463, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.19385665529010238, | |
| "grad_norm": 0.08359086513519287, | |
| "learning_rate": 9.659863945578231e-05, | |
| "loss": 0.6364, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.19658703071672354, | |
| "grad_norm": 0.0885363295674324, | |
| "learning_rate": 9.79591836734694e-05, | |
| "loss": 0.6092, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.1993174061433447, | |
| "grad_norm": 0.09258115291595459, | |
| "learning_rate": 9.931972789115646e-05, | |
| "loss": 0.6229, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.20204778156996586, | |
| "grad_norm": 0.08969170600175858, | |
| "learning_rate": 0.00010068027210884355, | |
| "loss": 0.6173, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.20477815699658702, | |
| "grad_norm": 0.10124260932207108, | |
| "learning_rate": 0.00010204081632653062, | |
| "loss": 0.6414, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2075085324232082, | |
| "grad_norm": 0.08671349287033081, | |
| "learning_rate": 0.0001034013605442177, | |
| "loss": 0.6145, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.21023890784982935, | |
| "grad_norm": 0.09684890508651733, | |
| "learning_rate": 0.00010476190476190477, | |
| "loss": 0.6262, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.2129692832764505, | |
| "grad_norm": 0.08690830320119858, | |
| "learning_rate": 0.00010612244897959185, | |
| "loss": 0.6316, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.21569965870307167, | |
| "grad_norm": 0.10457205027341843, | |
| "learning_rate": 0.00010748299319727892, | |
| "loss": 0.639, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.21843003412969283, | |
| "grad_norm": 0.10080841183662415, | |
| "learning_rate": 0.000108843537414966, | |
| "loss": 0.592, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.221160409556314, | |
| "grad_norm": 0.08858262002468109, | |
| "learning_rate": 0.00011020408163265306, | |
| "loss": 0.6471, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.22389078498293516, | |
| "grad_norm": 0.08708172291517258, | |
| "learning_rate": 0.00011156462585034013, | |
| "loss": 0.6222, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.22662116040955632, | |
| "grad_norm": 0.1075206995010376, | |
| "learning_rate": 0.00011292517006802721, | |
| "loss": 0.5961, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.22935153583617748, | |
| "grad_norm": 0.11788732558488846, | |
| "learning_rate": 0.00011428571428571428, | |
| "loss": 0.609, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.23208191126279865, | |
| "grad_norm": 0.0956830084323883, | |
| "learning_rate": 0.00011564625850340137, | |
| "loss": 0.6042, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.2348122866894198, | |
| "grad_norm": 0.09799174964427948, | |
| "learning_rate": 0.00011700680272108844, | |
| "loss": 0.6045, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.23754266211604094, | |
| "grad_norm": 0.09177012741565704, | |
| "learning_rate": 0.00011836734693877552, | |
| "loss": 0.6068, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.2402730375426621, | |
| "grad_norm": 0.10407502949237823, | |
| "learning_rate": 0.00011972789115646259, | |
| "loss": 0.5993, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.24300341296928327, | |
| "grad_norm": 0.1047271341085434, | |
| "learning_rate": 0.00012108843537414967, | |
| "loss": 0.6144, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.24573378839590443, | |
| "grad_norm": 0.0866198018193245, | |
| "learning_rate": 0.00012244897959183676, | |
| "loss": 0.6203, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2484641638225256, | |
| "grad_norm": 0.09400323033332825, | |
| "learning_rate": 0.0001238095238095238, | |
| "loss": 0.6056, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.25119453924914675, | |
| "grad_norm": 0.0817628726363182, | |
| "learning_rate": 0.0001251700680272109, | |
| "loss": 0.5853, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.25392491467576794, | |
| "grad_norm": 0.09105788916349411, | |
| "learning_rate": 0.00012653061224489798, | |
| "loss": 0.5952, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.2566552901023891, | |
| "grad_norm": 0.09889201074838638, | |
| "learning_rate": 0.00012789115646258506, | |
| "loss": 0.5994, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2593856655290102, | |
| "grad_norm": 0.09481444954872131, | |
| "learning_rate": 0.00012925170068027212, | |
| "loss": 0.5918, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2621160409556314, | |
| "grad_norm": 0.11730329692363739, | |
| "learning_rate": 0.00013061224489795917, | |
| "loss": 0.592, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.26484641638225254, | |
| "grad_norm": 0.15733356773853302, | |
| "learning_rate": 0.00013197278911564626, | |
| "loss": 0.5636, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.2675767918088737, | |
| "grad_norm": 0.20819880068302155, | |
| "learning_rate": 0.00013333333333333334, | |
| "loss": 0.6101, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.27030716723549486, | |
| "grad_norm": 0.18305541574954987, | |
| "learning_rate": 0.0001346938775510204, | |
| "loss": 0.5814, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.27303754266211605, | |
| "grad_norm": 0.10316050797700882, | |
| "learning_rate": 0.00013605442176870748, | |
| "loss": 0.5871, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2757679180887372, | |
| "grad_norm": 0.13305549323558807, | |
| "learning_rate": 0.00013741496598639456, | |
| "loss": 0.5846, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.2784982935153584, | |
| "grad_norm": 0.0950811356306076, | |
| "learning_rate": 0.00013877551020408165, | |
| "loss": 0.5711, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.2812286689419795, | |
| "grad_norm": 0.1198628693819046, | |
| "learning_rate": 0.0001401360544217687, | |
| "loss": 0.5914, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.2839590443686007, | |
| "grad_norm": 0.08809541165828705, | |
| "learning_rate": 0.00014149659863945578, | |
| "loss": 0.5872, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.28668941979522183, | |
| "grad_norm": 0.09801067411899567, | |
| "learning_rate": 0.00014285714285714287, | |
| "loss": 0.566, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.289419795221843, | |
| "grad_norm": 0.08766568452119827, | |
| "learning_rate": 0.00014421768707482995, | |
| "loss": 0.5808, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.29215017064846416, | |
| "grad_norm": 0.09133429825305939, | |
| "learning_rate": 0.000145578231292517, | |
| "loss": 0.6037, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.29488054607508535, | |
| "grad_norm": 0.09074072539806366, | |
| "learning_rate": 0.0001469387755102041, | |
| "loss": 0.5897, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.2976109215017065, | |
| "grad_norm": 0.08934789896011353, | |
| "learning_rate": 0.00014829931972789117, | |
| "loss": 0.5998, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.3003412969283277, | |
| "grad_norm": 0.08707176148891449, | |
| "learning_rate": 0.00014965986394557826, | |
| "loss": 0.5762, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3030716723549488, | |
| "grad_norm": 0.0948200449347496, | |
| "learning_rate": 0.0001510204081632653, | |
| "loss": 0.5734, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.30580204778156994, | |
| "grad_norm": 0.08889783173799515, | |
| "learning_rate": 0.00015238095238095237, | |
| "loss": 0.5867, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.30853242320819113, | |
| "grad_norm": 0.08152323961257935, | |
| "learning_rate": 0.00015374149659863945, | |
| "loss": 0.5527, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.31126279863481227, | |
| "grad_norm": 0.09019389748573303, | |
| "learning_rate": 0.00015510204081632654, | |
| "loss": 0.6007, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.31399317406143346, | |
| "grad_norm": 0.08257456868886948, | |
| "learning_rate": 0.00015646258503401362, | |
| "loss": 0.5569, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.3167235494880546, | |
| "grad_norm": 0.08834348618984222, | |
| "learning_rate": 0.00015782312925170067, | |
| "loss": 0.6026, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.3194539249146758, | |
| "grad_norm": 0.08634665608406067, | |
| "learning_rate": 0.00015918367346938776, | |
| "loss": 0.5926, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3221843003412969, | |
| "grad_norm": 0.07867719978094101, | |
| "learning_rate": 0.00016054421768707484, | |
| "loss": 0.5707, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.3249146757679181, | |
| "grad_norm": 0.09690061956644058, | |
| "learning_rate": 0.00016190476190476192, | |
| "loss": 0.5793, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.32764505119453924, | |
| "grad_norm": 0.08276376128196716, | |
| "learning_rate": 0.00016326530612244898, | |
| "loss": 0.5459, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.33037542662116043, | |
| "grad_norm": 0.09276240319013596, | |
| "learning_rate": 0.00016462585034013606, | |
| "loss": 0.5732, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.33310580204778156, | |
| "grad_norm": 0.0819844901561737, | |
| "learning_rate": 0.00016598639455782315, | |
| "loss": 0.5349, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.33583617747440275, | |
| "grad_norm": 0.08146791905164719, | |
| "learning_rate": 0.00016734693877551023, | |
| "loss": 0.5656, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.3385665529010239, | |
| "grad_norm": 0.0879024788737297, | |
| "learning_rate": 0.00016870748299319729, | |
| "loss": 0.5758, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.3412969283276451, | |
| "grad_norm": 0.07890356332063675, | |
| "learning_rate": 0.00017006802721088434, | |
| "loss": 0.5332, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.3440273037542662, | |
| "grad_norm": 0.10049955546855927, | |
| "learning_rate": 0.00017142857142857143, | |
| "loss": 0.5671, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.34675767918088735, | |
| "grad_norm": 0.09643971920013428, | |
| "learning_rate": 0.0001727891156462585, | |
| "loss": 0.5812, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.34948805460750854, | |
| "grad_norm": 0.08666185289621353, | |
| "learning_rate": 0.0001741496598639456, | |
| "loss": 0.5487, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.35221843003412967, | |
| "grad_norm": 0.1031438484787941, | |
| "learning_rate": 0.00017551020408163265, | |
| "loss": 0.5558, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.35494880546075086, | |
| "grad_norm": 0.09404855966567993, | |
| "learning_rate": 0.00017687074829931973, | |
| "loss": 0.5615, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.357679180887372, | |
| "grad_norm": 0.09127198159694672, | |
| "learning_rate": 0.00017823129251700681, | |
| "loss": 0.5656, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.3604095563139932, | |
| "grad_norm": 0.08694130182266235, | |
| "learning_rate": 0.0001795918367346939, | |
| "loss": 0.5379, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.3631399317406143, | |
| "grad_norm": 0.09511597454547882, | |
| "learning_rate": 0.00018095238095238095, | |
| "loss": 0.5535, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.3658703071672355, | |
| "grad_norm": 0.09129739552736282, | |
| "learning_rate": 0.00018231292517006804, | |
| "loss": 0.5678, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.36860068259385664, | |
| "grad_norm": 0.09248334169387817, | |
| "learning_rate": 0.00018367346938775512, | |
| "loss": 0.5574, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.37133105802047783, | |
| "grad_norm": 0.09906318038702011, | |
| "learning_rate": 0.0001850340136054422, | |
| "loss": 0.5499, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.37406143344709897, | |
| "grad_norm": 0.09928654134273529, | |
| "learning_rate": 0.00018639455782312926, | |
| "loss": 0.5413, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.37679180887372016, | |
| "grad_norm": 0.07559472322463989, | |
| "learning_rate": 0.00018775510204081634, | |
| "loss": 0.5475, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.3795221843003413, | |
| "grad_norm": 0.08408834040164948, | |
| "learning_rate": 0.00018911564625850343, | |
| "loss": 0.5432, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3822525597269625, | |
| "grad_norm": 0.08800789713859558, | |
| "learning_rate": 0.00019047619047619048, | |
| "loss": 0.5587, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3849829351535836, | |
| "grad_norm": 0.09994784742593765, | |
| "learning_rate": 0.00019183673469387756, | |
| "loss": 0.555, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.38771331058020475, | |
| "grad_norm": 0.07616768032312393, | |
| "learning_rate": 0.00019319727891156462, | |
| "loss": 0.5621, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.39044368600682594, | |
| "grad_norm": 0.10337202996015549, | |
| "learning_rate": 0.0001945578231292517, | |
| "loss": 0.5282, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.3931740614334471, | |
| "grad_norm": 0.08526328206062317, | |
| "learning_rate": 0.0001959183673469388, | |
| "loss": 0.5439, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.39590443686006827, | |
| "grad_norm": 0.10538353770971298, | |
| "learning_rate": 0.00019727891156462587, | |
| "loss": 0.5481, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.3986348122866894, | |
| "grad_norm": 0.07550521194934845, | |
| "learning_rate": 0.00019863945578231293, | |
| "loss": 0.5414, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.4013651877133106, | |
| "grad_norm": 0.10045620799064636, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5382, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.4040955631399317, | |
| "grad_norm": 0.08987366408109665, | |
| "learning_rate": 0.00019999971548969982, | |
| "loss": 0.5417, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.4068259385665529, | |
| "grad_norm": 0.0801815390586853, | |
| "learning_rate": 0.0001999988619604182, | |
| "loss": 0.5275, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.40955631399317405, | |
| "grad_norm": 0.08214934170246124, | |
| "learning_rate": 0.00019999743941701188, | |
| "loss": 0.543, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.41228668941979524, | |
| "grad_norm": 0.08146006613969803, | |
| "learning_rate": 0.00019999544786757545, | |
| "loss": 0.5409, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.4150170648464164, | |
| "grad_norm": 0.08081945031881332, | |
| "learning_rate": 0.00019999288732344122, | |
| "loss": 0.5509, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.41774744027303756, | |
| "grad_norm": 0.09135357290506363, | |
| "learning_rate": 0.0001999897577991792, | |
| "loss": 0.518, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.4204778156996587, | |
| "grad_norm": 0.09191333502531052, | |
| "learning_rate": 0.0001999860593125971, | |
| "loss": 0.5276, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.4232081911262799, | |
| "grad_norm": 0.08375995606184006, | |
| "learning_rate": 0.00019998179188473997, | |
| "loss": 0.5319, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.425938566552901, | |
| "grad_norm": 0.08481922000646591, | |
| "learning_rate": 0.00019997695553989042, | |
| "loss": 0.5437, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.4286689419795222, | |
| "grad_norm": 0.08768640458583832, | |
| "learning_rate": 0.00019997155030556822, | |
| "loss": 0.5445, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.43139931740614335, | |
| "grad_norm": 0.08787625283002853, | |
| "learning_rate": 0.00019996557621253027, | |
| "loss": 0.5479, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.4341296928327645, | |
| "grad_norm": 0.09505843371152878, | |
| "learning_rate": 0.0001999590332947704, | |
| "loss": 0.5263, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.43686006825938567, | |
| "grad_norm": 0.10003377497196198, | |
| "learning_rate": 0.00019995192158951919, | |
| "loss": 0.5228, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.4395904436860068, | |
| "grad_norm": 0.0675501748919487, | |
| "learning_rate": 0.00019994424113724363, | |
| "loss": 0.4977, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.442320819112628, | |
| "grad_norm": 0.09747067093849182, | |
| "learning_rate": 0.00019993599198164715, | |
| "loss": 0.5161, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.44505119453924913, | |
| "grad_norm": 0.0837995857000351, | |
| "learning_rate": 0.0001999271741696691, | |
| "loss": 0.5243, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.4477815699658703, | |
| "grad_norm": 0.0793512687087059, | |
| "learning_rate": 0.00019991778775148465, | |
| "loss": 0.5141, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.45051194539249145, | |
| "grad_norm": 0.07802822440862656, | |
| "learning_rate": 0.00019990783278050448, | |
| "loss": 0.515, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.45324232081911264, | |
| "grad_norm": 0.08355724066495895, | |
| "learning_rate": 0.0001998973093133744, | |
| "loss": 0.5176, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.4559726962457338, | |
| "grad_norm": 0.08045308291912079, | |
| "learning_rate": 0.00019988621740997512, | |
| "loss": 0.5151, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.45870307167235497, | |
| "grad_norm": 0.07589907944202423, | |
| "learning_rate": 0.00019987455713342187, | |
| "loss": 0.5249, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.4614334470989761, | |
| "grad_norm": 0.08553771674633026, | |
| "learning_rate": 0.000199862328550064, | |
| "loss": 0.5485, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.4641638225255973, | |
| "grad_norm": 0.08599649369716644, | |
| "learning_rate": 0.00019984953172948465, | |
| "loss": 0.53, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.4668941979522184, | |
| "grad_norm": 0.06906479597091675, | |
| "learning_rate": 0.0001998361667445004, | |
| "loss": 0.5336, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.4696245733788396, | |
| "grad_norm": 0.07526392489671707, | |
| "learning_rate": 0.00019982223367116076, | |
| "loss": 0.5013, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.47235494880546075, | |
| "grad_norm": 0.0722610279917717, | |
| "learning_rate": 0.00019980773258874778, | |
| "loss": 0.5217, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.4750853242320819, | |
| "grad_norm": 0.0773632749915123, | |
| "learning_rate": 0.00019979266357977564, | |
| "loss": 0.5184, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.4778156996587031, | |
| "grad_norm": 0.07160216569900513, | |
| "learning_rate": 0.00019977702672999007, | |
| "loss": 0.5009, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.4805460750853242, | |
| "grad_norm": 0.0764177069067955, | |
| "learning_rate": 0.00019976082212836793, | |
| "loss": 0.5126, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.4832764505119454, | |
| "grad_norm": 0.07116773724555969, | |
| "learning_rate": 0.0001997440498671168, | |
| "loss": 0.514, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.48600682593856653, | |
| "grad_norm": 0.08402683585882187, | |
| "learning_rate": 0.00019972671004167433, | |
| "loss": 0.5133, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.4887372013651877, | |
| "grad_norm": 0.07286666333675385, | |
| "learning_rate": 0.00019970880275070762, | |
| "loss": 0.5221, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.49146757679180886, | |
| "grad_norm": 0.08641263097524643, | |
| "learning_rate": 0.00019969032809611287, | |
| "loss": 0.4959, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.49419795221843005, | |
| "grad_norm": 0.08849737048149109, | |
| "learning_rate": 0.0001996712861830147, | |
| "loss": 0.4952, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.4969283276450512, | |
| "grad_norm": 0.08661802858114243, | |
| "learning_rate": 0.00019965167711976552, | |
| "loss": 0.5023, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.49965870307167237, | |
| "grad_norm": 0.08355259150266647, | |
| "learning_rate": 0.0001996315010179449, | |
| "loss": 0.5235, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.5023890784982935, | |
| "grad_norm": 0.07524804770946503, | |
| "learning_rate": 0.00019961075799235903, | |
| "loss": 0.5143, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.5051194539249146, | |
| "grad_norm": 0.08126044273376465, | |
| "learning_rate": 0.00019958944816104, | |
| "loss": 0.496, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.5078498293515359, | |
| "grad_norm": 0.08320248872041702, | |
| "learning_rate": 0.00019956757164524516, | |
| "loss": 0.5106, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.510580204778157, | |
| "grad_norm": 0.07375509291887283, | |
| "learning_rate": 0.00019954512856945632, | |
| "loss": 0.4811, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.5133105802047782, | |
| "grad_norm": 0.07187776267528534, | |
| "learning_rate": 0.00019952211906137932, | |
| "loss": 0.5104, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.5160409556313993, | |
| "grad_norm": 0.07441398501396179, | |
| "learning_rate": 0.00019949854325194294, | |
| "loss": 0.5304, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.5187713310580204, | |
| "grad_norm": 0.07976701855659485, | |
| "learning_rate": 0.00019947440127529836, | |
| "loss": 0.4945, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.5215017064846417, | |
| "grad_norm": 0.07280328124761581, | |
| "learning_rate": 0.00019944969326881845, | |
| "loss": 0.4848, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.5242320819112628, | |
| "grad_norm": 0.07618428766727448, | |
| "learning_rate": 0.00019942441937309684, | |
| "loss": 0.4858, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.5269624573378839, | |
| "grad_norm": 0.0665225088596344, | |
| "learning_rate": 0.00019939857973194717, | |
| "loss": 0.4955, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.5296928327645051, | |
| "grad_norm": 0.08379194140434265, | |
| "learning_rate": 0.0001993721744924024, | |
| "loss": 0.5067, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.5324232081911263, | |
| "grad_norm": 0.07564423978328705, | |
| "learning_rate": 0.00019934520380471372, | |
| "loss": 0.5159, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.5351535836177475, | |
| "grad_norm": 0.07225633412599564, | |
| "learning_rate": 0.0001993176678223499, | |
| "loss": 0.5144, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.5378839590443686, | |
| "grad_norm": 0.07224252074956894, | |
| "learning_rate": 0.0001992895667019964, | |
| "loss": 0.4859, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.5406143344709897, | |
| "grad_norm": 0.079926997423172, | |
| "learning_rate": 0.0001992609006035543, | |
| "loss": 0.4872, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.543344709897611, | |
| "grad_norm": 0.08545151352882385, | |
| "learning_rate": 0.0001992316696901397, | |
| "loss": 0.5105, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.5460750853242321, | |
| "grad_norm": 0.08008193224668503, | |
| "learning_rate": 0.00019920187412808248, | |
| "loss": 0.4903, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5488054607508532, | |
| "grad_norm": 0.06717066466808319, | |
| "learning_rate": 0.0001991715140869255, | |
| "loss": 0.5037, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.5515358361774744, | |
| "grad_norm": 0.08613338321447372, | |
| "learning_rate": 0.00019914058973942368, | |
| "loss": 0.4999, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.5542662116040956, | |
| "grad_norm": 0.07288234680891037, | |
| "learning_rate": 0.00019910910126154293, | |
| "loss": 0.5019, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.5569965870307167, | |
| "grad_norm": 0.07831370085477829, | |
| "learning_rate": 0.00019907704883245916, | |
| "loss": 0.4595, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.5597269624573379, | |
| "grad_norm": 0.0916525200009346, | |
| "learning_rate": 0.00019904443263455728, | |
| "loss": 0.4994, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.562457337883959, | |
| "grad_norm": 0.07431495934724808, | |
| "learning_rate": 0.00019901125285343022, | |
| "loss": 0.5059, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.5651877133105802, | |
| "grad_norm": 0.07864730060100555, | |
| "learning_rate": 0.0001989775096778777, | |
| "loss": 0.4824, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.5679180887372014, | |
| "grad_norm": 0.06928006559610367, | |
| "learning_rate": 0.0001989432032999054, | |
| "loss": 0.4887, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.5706484641638225, | |
| "grad_norm": 0.07330948859453201, | |
| "learning_rate": 0.0001989083339147237, | |
| "loss": 0.4804, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.5733788395904437, | |
| "grad_norm": 0.07905860990285873, | |
| "learning_rate": 0.0001988729017207465, | |
| "loss": 0.5126, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5761092150170648, | |
| "grad_norm": 0.07062509655952454, | |
| "learning_rate": 0.00019883690691959035, | |
| "loss": 0.5063, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.578839590443686, | |
| "grad_norm": 0.071404367685318, | |
| "learning_rate": 0.00019880034971607308, | |
| "loss": 0.495, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.5815699658703072, | |
| "grad_norm": 0.0727284774184227, | |
| "learning_rate": 0.00019876323031821266, | |
| "loss": 0.4994, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.5843003412969283, | |
| "grad_norm": 0.07198608666658401, | |
| "learning_rate": 0.00019872554893722618, | |
| "loss": 0.4903, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.5870307167235495, | |
| "grad_norm": 0.07637451589107513, | |
| "learning_rate": 0.0001986873057875284, | |
| "loss": 0.5057, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.5897610921501707, | |
| "grad_norm": 0.06596951186656952, | |
| "learning_rate": 0.00019864850108673073, | |
| "loss": 0.4932, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.5924914675767918, | |
| "grad_norm": 0.06999579071998596, | |
| "learning_rate": 0.0001986091350556399, | |
| "loss": 0.4887, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.595221843003413, | |
| "grad_norm": 0.06687980890274048, | |
| "learning_rate": 0.00019856920791825683, | |
| "loss": 0.472, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.5979522184300341, | |
| "grad_norm": 0.07001427561044693, | |
| "learning_rate": 0.00019852871990177503, | |
| "loss": 0.4692, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.6006825938566553, | |
| "grad_norm": 0.06714101880788803, | |
| "learning_rate": 0.00019848767123657976, | |
| "loss": 0.4813, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.6034129692832765, | |
| "grad_norm": 0.07292049378156662, | |
| "learning_rate": 0.0001984460621562463, | |
| "loss": 0.4885, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.6061433447098976, | |
| "grad_norm": 0.06814104318618774, | |
| "learning_rate": 0.00019840389289753896, | |
| "loss": 0.4938, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.6088737201365187, | |
| "grad_norm": 0.06866355985403061, | |
| "learning_rate": 0.00019836116370040944, | |
| "loss": 0.4776, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.6116040955631399, | |
| "grad_norm": 0.07145702093839645, | |
| "learning_rate": 0.00019831787480799568, | |
| "loss": 0.4883, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.6143344709897611, | |
| "grad_norm": 0.06319977343082428, | |
| "learning_rate": 0.00019827402646662047, | |
| "loss": 0.4882, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.6170648464163823, | |
| "grad_norm": 0.08186688274145126, | |
| "learning_rate": 0.0001982296189257898, | |
| "loss": 0.4917, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.6197952218430034, | |
| "grad_norm": 0.06892900168895721, | |
| "learning_rate": 0.00019818465243819184, | |
| "loss": 0.4808, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.6225255972696245, | |
| "grad_norm": 0.0752168744802475, | |
| "learning_rate": 0.00019813912725969509, | |
| "loss": 0.4858, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.6252559726962458, | |
| "grad_norm": 0.08079662919044495, | |
| "learning_rate": 0.0001980930436493472, | |
| "loss": 0.5101, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.6279863481228669, | |
| "grad_norm": 0.0717153325676918, | |
| "learning_rate": 0.00019804640186937343, | |
| "loss": 0.4799, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.630716723549488, | |
| "grad_norm": 0.08962002396583557, | |
| "learning_rate": 0.0001979992021851751, | |
| "loss": 0.5067, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.6334470989761092, | |
| "grad_norm": 0.08904211223125458, | |
| "learning_rate": 0.00019795144486532814, | |
| "loss": 0.4725, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.6361774744027304, | |
| "grad_norm": 0.06842932850122452, | |
| "learning_rate": 0.00019790313018158156, | |
| "loss": 0.4996, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.6389078498293516, | |
| "grad_norm": 0.08361311256885529, | |
| "learning_rate": 0.0001978542584088558, | |
| "loss": 0.4945, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.6416382252559727, | |
| "grad_norm": 0.07219431549310684, | |
| "learning_rate": 0.00019780482982524142, | |
| "loss": 0.4488, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.6443686006825938, | |
| "grad_norm": 0.07717226445674896, | |
| "learning_rate": 0.00019775484471199715, | |
| "loss": 0.4814, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.647098976109215, | |
| "grad_norm": 0.07770105451345444, | |
| "learning_rate": 0.0001977043033535486, | |
| "loss": 0.4731, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.6498293515358362, | |
| "grad_norm": 0.06878919899463654, | |
| "learning_rate": 0.00019765320603748655, | |
| "loss": 0.4833, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.6525597269624573, | |
| "grad_norm": 0.07085343450307846, | |
| "learning_rate": 0.0001976015530545652, | |
| "loss": 0.4907, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.6552901023890785, | |
| "grad_norm": 0.07935165613889694, | |
| "learning_rate": 0.0001975493446987007, | |
| "loss": 0.4794, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.6580204778156996, | |
| "grad_norm": 0.06543820351362228, | |
| "learning_rate": 0.00019749658126696934, | |
| "loss": 0.4906, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.6607508532423209, | |
| "grad_norm": 0.07727054506540298, | |
| "learning_rate": 0.00019744326305960595, | |
| "loss": 0.4868, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.663481228668942, | |
| "grad_norm": 0.06668544560670853, | |
| "learning_rate": 0.00019738939038000205, | |
| "loss": 0.475, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.6662116040955631, | |
| "grad_norm": 0.07048569619655609, | |
| "learning_rate": 0.00019733496353470433, | |
| "loss": 0.4878, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.6689419795221843, | |
| "grad_norm": 0.07110477238893509, | |
| "learning_rate": 0.00019727998283341274, | |
| "loss": 0.4663, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.6716723549488055, | |
| "grad_norm": 0.07245586067438126, | |
| "learning_rate": 0.00019722444858897878, | |
| "loss": 0.4899, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.6744027303754266, | |
| "grad_norm": 0.07484875619411469, | |
| "learning_rate": 0.00019716836111740378, | |
| "loss": 0.4831, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.6771331058020478, | |
| "grad_norm": 0.07812648266553879, | |
| "learning_rate": 0.00019711172073783696, | |
| "loss": 0.4654, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.6798634812286689, | |
| "grad_norm": 0.060632165521383286, | |
| "learning_rate": 0.00019705452777257377, | |
| "loss": 0.4706, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.6825938566552902, | |
| "grad_norm": 0.07092992216348648, | |
| "learning_rate": 0.000196996782547054, | |
| "loss": 0.4792, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6853242320819113, | |
| "grad_norm": 0.06629595905542374, | |
| "learning_rate": 0.00019693848538985983, | |
| "loss": 0.4791, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.6880546075085324, | |
| "grad_norm": 0.06915664672851562, | |
| "learning_rate": 0.00019687963663271409, | |
| "loss": 0.4623, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.6907849829351536, | |
| "grad_norm": 0.0694665014743805, | |
| "learning_rate": 0.00019682023661047836, | |
| "loss": 0.48, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.6935153583617747, | |
| "grad_norm": 0.06899196654558182, | |
| "learning_rate": 0.00019676028566115102, | |
| "loss": 0.4855, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.6962457337883959, | |
| "grad_norm": 0.0740811675786972, | |
| "learning_rate": 0.00019669978412586528, | |
| "loss": 0.4833, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.6989761092150171, | |
| "grad_norm": 0.06517481803894043, | |
| "learning_rate": 0.00019663873234888733, | |
| "loss": 0.4523, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.7017064846416382, | |
| "grad_norm": 0.06481153517961502, | |
| "learning_rate": 0.0001965771306776144, | |
| "loss": 0.4689, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.7044368600682593, | |
| "grad_norm": 0.06042364612221718, | |
| "learning_rate": 0.00019651497946257266, | |
| "loss": 0.4757, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.7071672354948806, | |
| "grad_norm": 0.0717868059873581, | |
| "learning_rate": 0.00019645227905741534, | |
| "loss": 0.4773, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.7098976109215017, | |
| "grad_norm": 0.06427443772554398, | |
| "learning_rate": 0.00019638902981892068, | |
| "loss": 0.4875, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.7126279863481229, | |
| "grad_norm": 0.07786547392606735, | |
| "learning_rate": 0.00019632523210698987, | |
| "loss": 0.4758, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.715358361774744, | |
| "grad_norm": 0.07115910202264786, | |
| "learning_rate": 0.00019626088628464498, | |
| "loss": 0.4651, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.7180887372013652, | |
| "grad_norm": 0.06626811623573303, | |
| "learning_rate": 0.00019619599271802706, | |
| "loss": 0.4873, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.7208191126279864, | |
| "grad_norm": 0.07854583859443665, | |
| "learning_rate": 0.00019613055177639384, | |
| "loss": 0.4945, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.7235494880546075, | |
| "grad_norm": 0.0847892239689827, | |
| "learning_rate": 0.00019606456383211777, | |
| "loss": 0.4671, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.7262798634812286, | |
| "grad_norm": 0.06735772639513016, | |
| "learning_rate": 0.00019599802926068384, | |
| "loss": 0.4767, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.7290102389078499, | |
| "grad_norm": 0.07502768933773041, | |
| "learning_rate": 0.00019593094844068748, | |
| "loss": 0.462, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.731740614334471, | |
| "grad_norm": 0.07276903837919235, | |
| "learning_rate": 0.00019586332175383238, | |
| "loss": 0.4754, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.7344709897610922, | |
| "grad_norm": 0.07755447924137115, | |
| "learning_rate": 0.00019579514958492826, | |
| "loss": 0.492, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.7372013651877133, | |
| "grad_norm": 0.07876396179199219, | |
| "learning_rate": 0.0001957264323218889, | |
| "loss": 0.4737, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.7399317406143344, | |
| "grad_norm": 0.07997962832450867, | |
| "learning_rate": 0.0001956571703557296, | |
| "loss": 0.4592, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.7426621160409557, | |
| "grad_norm": 0.08079583197832108, | |
| "learning_rate": 0.00019558736408056525, | |
| "loss": 0.473, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.7453924914675768, | |
| "grad_norm": 0.0736604854464531, | |
| "learning_rate": 0.00019551701389360795, | |
| "loss": 0.4741, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.7481228668941979, | |
| "grad_norm": 0.0741550549864769, | |
| "learning_rate": 0.00019544612019516472, | |
| "loss": 0.4611, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.7508532423208191, | |
| "grad_norm": 0.06802786141633987, | |
| "learning_rate": 0.00019537468338863537, | |
| "loss": 0.4621, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.7535836177474403, | |
| "grad_norm": 0.06499720364809036, | |
| "learning_rate": 0.00019530270388050998, | |
| "loss": 0.4676, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.7563139931740614, | |
| "grad_norm": 0.06809037923812866, | |
| "learning_rate": 0.00019523018208036677, | |
| "loss": 0.475, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.7590443686006826, | |
| "grad_norm": 0.06455886363983154, | |
| "learning_rate": 0.0001951571184008698, | |
| "loss": 0.4807, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.7617747440273037, | |
| "grad_norm": 0.06833679229021072, | |
| "learning_rate": 0.00019508351325776642, | |
| "loss": 0.4751, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.764505119453925, | |
| "grad_norm": 0.07593976706266403, | |
| "learning_rate": 0.00019500936706988502, | |
| "loss": 0.4714, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.7672354948805461, | |
| "grad_norm": 0.0687364712357521, | |
| "learning_rate": 0.00019493468025913276, | |
| "loss": 0.4575, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.7699658703071672, | |
| "grad_norm": 0.07183225452899933, | |
| "learning_rate": 0.00019485945325049288, | |
| "loss": 0.4815, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.7726962457337884, | |
| "grad_norm": 0.06775309145450592, | |
| "learning_rate": 0.00019478368647202264, | |
| "loss": 0.4543, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.7754266211604095, | |
| "grad_norm": 0.06261654198169708, | |
| "learning_rate": 0.00019470738035485058, | |
| "loss": 0.4724, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.7781569965870307, | |
| "grad_norm": 0.06674676388502121, | |
| "learning_rate": 0.00019463053533317425, | |
| "loss": 0.4667, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.7808873720136519, | |
| "grad_norm": 0.06266098469495773, | |
| "learning_rate": 0.0001945531518442576, | |
| "loss": 0.4614, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.783617747440273, | |
| "grad_norm": 0.06769178062677383, | |
| "learning_rate": 0.0001944752303284287, | |
| "loss": 0.4609, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.7863481228668942, | |
| "grad_norm": 0.07618339359760284, | |
| "learning_rate": 0.00019439677122907697, | |
| "loss": 0.4822, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.7890784982935154, | |
| "grad_norm": 0.06216439977288246, | |
| "learning_rate": 0.00019431777499265087, | |
| "loss": 0.4573, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.7918088737201365, | |
| "grad_norm": 0.06998062878847122, | |
| "learning_rate": 0.00019423824206865527, | |
| "loss": 0.4683, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7945392491467577, | |
| "grad_norm": 0.06178448721766472, | |
| "learning_rate": 0.00019415817290964883, | |
| "loss": 0.4643, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.7972696245733788, | |
| "grad_norm": 0.06611185520887375, | |
| "learning_rate": 0.00019407756797124164, | |
| "loss": 0.4712, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.06682468205690384, | |
| "learning_rate": 0.00019399642771209238, | |
| "loss": 0.474, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.8027303754266212, | |
| "grad_norm": 0.0632803738117218, | |
| "learning_rate": 0.00019391475259390584, | |
| "loss": 0.4776, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.8054607508532423, | |
| "grad_norm": 0.06498962640762329, | |
| "learning_rate": 0.0001938325430814302, | |
| "loss": 0.4735, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.8081911262798634, | |
| "grad_norm": 0.06621643900871277, | |
| "learning_rate": 0.00019374979964245463, | |
| "loss": 0.4785, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.8109215017064847, | |
| "grad_norm": 0.05847141519188881, | |
| "learning_rate": 0.00019366652274780628, | |
| "loss": 0.4702, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.8136518771331058, | |
| "grad_norm": 0.06962229311466217, | |
| "learning_rate": 0.00019358271287134784, | |
| "loss": 0.4612, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.816382252559727, | |
| "grad_norm": 0.06132384389638901, | |
| "learning_rate": 0.00019349837048997478, | |
| "loss": 0.4453, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.8191126279863481, | |
| "grad_norm": 0.06574399024248123, | |
| "learning_rate": 0.00019341349608361267, | |
| "loss": 0.4545, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.8218430034129692, | |
| "grad_norm": 0.06561442464590073, | |
| "learning_rate": 0.00019332809013521428, | |
| "loss": 0.4619, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.8245733788395905, | |
| "grad_norm": 0.06309875100851059, | |
| "learning_rate": 0.00019324215313075706, | |
| "loss": 0.465, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.8273037542662116, | |
| "grad_norm": 0.06544878333806992, | |
| "learning_rate": 0.00019315568555924035, | |
| "loss": 0.4571, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.8300341296928327, | |
| "grad_norm": 0.07011238485574722, | |
| "learning_rate": 0.0001930686879126824, | |
| "loss": 0.4579, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.8327645051194539, | |
| "grad_norm": 0.06445574760437012, | |
| "learning_rate": 0.0001929811606861177, | |
| "loss": 0.4695, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.8354948805460751, | |
| "grad_norm": 0.061930734664201736, | |
| "learning_rate": 0.00019289310437759427, | |
| "loss": 0.4449, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.8382252559726963, | |
| "grad_norm": 0.0658838227391243, | |
| "learning_rate": 0.00019280451948817059, | |
| "loss": 0.4726, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.8409556313993174, | |
| "grad_norm": 0.06302706897258759, | |
| "learning_rate": 0.00019271540652191296, | |
| "loss": 0.447, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.8436860068259385, | |
| "grad_norm": 0.08308806270360947, | |
| "learning_rate": 0.0001926257659858925, | |
| "loss": 0.4605, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.8464163822525598, | |
| "grad_norm": 0.06508838385343552, | |
| "learning_rate": 0.00019253559839018235, | |
| "loss": 0.4778, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.8491467576791809, | |
| "grad_norm": 0.07429094612598419, | |
| "learning_rate": 0.00019244490424785468, | |
| "loss": 0.4659, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.851877133105802, | |
| "grad_norm": 0.07138285785913467, | |
| "learning_rate": 0.00019235368407497788, | |
| "loss": 0.4564, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.8546075085324232, | |
| "grad_norm": 0.07202211022377014, | |
| "learning_rate": 0.00019226193839061347, | |
| "loss": 0.4377, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.8573378839590444, | |
| "grad_norm": 0.0779070258140564, | |
| "learning_rate": 0.0001921696677168133, | |
| "loss": 0.4532, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.8600682593856656, | |
| "grad_norm": 0.07717596739530563, | |
| "learning_rate": 0.00019207687257861655, | |
| "loss": 0.4654, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.8627986348122867, | |
| "grad_norm": 0.0708346962928772, | |
| "learning_rate": 0.00019198355350404667, | |
| "loss": 0.4584, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.8655290102389078, | |
| "grad_norm": 0.0656716600060463, | |
| "learning_rate": 0.00019188971102410837, | |
| "loss": 0.4504, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.868259385665529, | |
| "grad_norm": 0.06869971752166748, | |
| "learning_rate": 0.00019179534567278475, | |
| "loss": 0.4592, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.8709897610921502, | |
| "grad_norm": 0.06358928978443146, | |
| "learning_rate": 0.00019170045798703406, | |
| "loss": 0.4376, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.8737201365187713, | |
| "grad_norm": 0.06602993607521057, | |
| "learning_rate": 0.0001916050485067868, | |
| "loss": 0.4692, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.8764505119453925, | |
| "grad_norm": 0.06115058436989784, | |
| "learning_rate": 0.00019150911777494258, | |
| "loss": 0.462, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.8791808873720136, | |
| "grad_norm": 0.06374403834342957, | |
| "learning_rate": 0.00019141266633736697, | |
| "loss": 0.4325, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.8819112627986349, | |
| "grad_norm": 0.06459895521402359, | |
| "learning_rate": 0.0001913156947428886, | |
| "loss": 0.4605, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.884641638225256, | |
| "grad_norm": 0.06160016357898712, | |
| "learning_rate": 0.00019121820354329577, | |
| "loss": 0.4604, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.8873720136518771, | |
| "grad_norm": 0.06345291435718536, | |
| "learning_rate": 0.00019112019329333346, | |
| "loss": 0.4565, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.8901023890784983, | |
| "grad_norm": 0.06534894555807114, | |
| "learning_rate": 0.00019102166455070024, | |
| "loss": 0.4619, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.8928327645051195, | |
| "grad_norm": 0.06186550110578537, | |
| "learning_rate": 0.00019092261787604492, | |
| "loss": 0.4477, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.8955631399317406, | |
| "grad_norm": 0.058699868619441986, | |
| "learning_rate": 0.00019082305383296352, | |
| "loss": 0.4484, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.8982935153583618, | |
| "grad_norm": 0.05798410624265671, | |
| "learning_rate": 0.00019072297298799589, | |
| "loss": 0.4605, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.9010238907849829, | |
| "grad_norm": 0.06147664040327072, | |
| "learning_rate": 0.00019062237591062272, | |
| "loss": 0.4489, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.903754266211604, | |
| "grad_norm": 0.06032559648156166, | |
| "learning_rate": 0.00019052126317326207, | |
| "loss": 0.4412, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.9064846416382253, | |
| "grad_norm": 0.06326504051685333, | |
| "learning_rate": 0.00019041963535126625, | |
| "loss": 0.4547, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.9092150170648464, | |
| "grad_norm": 0.06808637827634811, | |
| "learning_rate": 0.0001903174930229185, | |
| "loss": 0.4513, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.9119453924914676, | |
| "grad_norm": 0.06384904682636261, | |
| "learning_rate": 0.00019021483676942973, | |
| "loss": 0.4542, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.9146757679180887, | |
| "grad_norm": 0.07148803770542145, | |
| "learning_rate": 0.00019011166717493517, | |
| "loss": 0.4569, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.9174061433447099, | |
| "grad_norm": 0.06942867487668991, | |
| "learning_rate": 0.000190007984826491, | |
| "loss": 0.4496, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.9201365187713311, | |
| "grad_norm": 0.06153569370508194, | |
| "learning_rate": 0.00018990379031407124, | |
| "loss": 0.464, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.9228668941979522, | |
| "grad_norm": 0.07417679578065872, | |
| "learning_rate": 0.00018979908423056408, | |
| "loss": 0.4396, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.9255972696245733, | |
| "grad_norm": 0.06745341420173645, | |
| "learning_rate": 0.0001896938671717687, | |
| "loss": 0.4584, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.9283276450511946, | |
| "grad_norm": 0.060262780636548996, | |
| "learning_rate": 0.00018958813973639184, | |
| "loss": 0.4363, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.9310580204778157, | |
| "grad_norm": 0.06427337974309921, | |
| "learning_rate": 0.0001894819025260444, | |
| "loss": 0.4352, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.9337883959044369, | |
| "grad_norm": 0.06150776520371437, | |
| "learning_rate": 0.00018937515614523797, | |
| "loss": 0.4644, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.936518771331058, | |
| "grad_norm": 0.06864424049854279, | |
| "learning_rate": 0.0001892679012013815, | |
| "loss": 0.4608, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.9392491467576792, | |
| "grad_norm": 0.06174071133136749, | |
| "learning_rate": 0.00018916013830477766, | |
| "loss": 0.4402, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.9419795221843004, | |
| "grad_norm": 0.0684589147567749, | |
| "learning_rate": 0.00018905186806861957, | |
| "loss": 0.4569, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.9447098976109215, | |
| "grad_norm": 0.05750627443194389, | |
| "learning_rate": 0.00018894309110898712, | |
| "loss": 0.4522, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.9474402730375426, | |
| "grad_norm": 0.0697883740067482, | |
| "learning_rate": 0.00018883380804484367, | |
| "loss": 0.4594, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.9501706484641638, | |
| "grad_norm": 0.06613462418317795, | |
| "learning_rate": 0.00018872401949803237, | |
| "loss": 0.4459, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.952901023890785, | |
| "grad_norm": 0.06346327811479568, | |
| "learning_rate": 0.00018861372609327263, | |
| "loss": 0.4316, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.9556313993174061, | |
| "grad_norm": 0.06382953375577927, | |
| "learning_rate": 0.00018850292845815672, | |
| "loss": 0.4358, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.9583617747440273, | |
| "grad_norm": 0.07121171057224274, | |
| "learning_rate": 0.0001883916272231459, | |
| "loss": 0.465, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.9610921501706484, | |
| "grad_norm": 0.06311832368373871, | |
| "learning_rate": 0.0001882798230215672, | |
| "loss": 0.4478, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.9638225255972697, | |
| "grad_norm": 0.06858519464731216, | |
| "learning_rate": 0.00018816751648960956, | |
| "loss": 0.4402, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.9665529010238908, | |
| "grad_norm": 0.06063356623053551, | |
| "learning_rate": 0.00018805470826632024, | |
| "loss": 0.4373, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.9692832764505119, | |
| "grad_norm": 0.06550437211990356, | |
| "learning_rate": 0.0001879413989936013, | |
| "loss": 0.4448, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.9720136518771331, | |
| "grad_norm": 0.06248946860432625, | |
| "learning_rate": 0.00018782758931620584, | |
| "loss": 0.4576, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.9747440273037543, | |
| "grad_norm": 0.07067371159791946, | |
| "learning_rate": 0.00018771327988173435, | |
| "loss": 0.4644, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.9774744027303754, | |
| "grad_norm": 0.06225898116827011, | |
| "learning_rate": 0.00018759847134063108, | |
| "loss": 0.4617, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.9802047781569966, | |
| "grad_norm": 0.061437107622623444, | |
| "learning_rate": 0.0001874831643461803, | |
| "loss": 0.4339, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.9829351535836177, | |
| "grad_norm": 0.059149857610464096, | |
| "learning_rate": 0.00018736735955450251, | |
| "loss": 0.4238, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.985665529010239, | |
| "grad_norm": 0.06511219590902328, | |
| "learning_rate": 0.0001872510576245509, | |
| "loss": 0.4394, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.9883959044368601, | |
| "grad_norm": 0.06580841541290283, | |
| "learning_rate": 0.00018713425921810733, | |
| "loss": 0.4218, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.9911262798634812, | |
| "grad_norm": 0.07789267599582672, | |
| "learning_rate": 0.00018701696499977884, | |
| "loss": 0.4524, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.9938566552901024, | |
| "grad_norm": 0.06430528312921524, | |
| "learning_rate": 0.0001868991756369937, | |
| "loss": 0.4503, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.9965870307167235, | |
| "grad_norm": 0.06355779618024826, | |
| "learning_rate": 0.00018678089179999762, | |
| "loss": 0.4556, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.9993174061433447, | |
| "grad_norm": 0.06800378113985062, | |
| "learning_rate": 0.00018666211416184999, | |
| "loss": 0.44, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.9993174061433447, | |
| "eval_loss": 0.4462641775608063, | |
| "eval_runtime": 311.1378, | |
| "eval_samples_per_second": 8.369, | |
| "eval_steps_per_second": 1.048, | |
| "step": 366 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1464, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.186219049536717e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |