| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 504, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.003968253968253968, | |
| "grad_norm": 234.72491455078125, | |
| "learning_rate": 3.846153846153847e-07, | |
| "loss": 1.767, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007936507936507936, | |
| "grad_norm": 242.27430725097656, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 1.8234, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.011904761904761904, | |
| "grad_norm": 265.2084655761719, | |
| "learning_rate": 1.153846153846154e-06, | |
| "loss": 1.8467, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.015873015873015872, | |
| "grad_norm": 219.6301727294922, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 1.8489, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01984126984126984, | |
| "grad_norm": 204.40042114257812, | |
| "learning_rate": 1.9230769230769234e-06, | |
| "loss": 1.7277, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.023809523809523808, | |
| "grad_norm": 185.3464813232422, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 1.7701, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.027777777777777776, | |
| "grad_norm": 174.37451171875, | |
| "learning_rate": 2.6923076923076923e-06, | |
| "loss": 1.7454, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.031746031746031744, | |
| "grad_norm": 53.370826721191406, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.589, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03571428571428571, | |
| "grad_norm": 65.17623138427734, | |
| "learning_rate": 3.4615384615384617e-06, | |
| "loss": 1.5916, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03968253968253968, | |
| "grad_norm": 131.7885284423828, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 1.5824, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04365079365079365, | |
| "grad_norm": 29.81640625, | |
| "learning_rate": 4.230769230769231e-06, | |
| "loss": 1.4016, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.047619047619047616, | |
| "grad_norm": 19.819799423217773, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 1.3137, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.051587301587301584, | |
| "grad_norm": 15.250455856323242, | |
| "learning_rate": 5e-06, | |
| "loss": 1.238, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.05555555555555555, | |
| "grad_norm": 20.343124389648438, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 1.2187, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.05952380952380952, | |
| "grad_norm": 16.577085494995117, | |
| "learning_rate": 5.769230769230769e-06, | |
| "loss": 1.0833, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06349206349206349, | |
| "grad_norm": 14.074089050292969, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9363, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.06746031746031746, | |
| "grad_norm": 12.567553520202637, | |
| "learning_rate": 6.538461538461539e-06, | |
| "loss": 0.9097, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 7.722719192504883, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 0.8343, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.07539682539682539, | |
| "grad_norm": 6.8982768058776855, | |
| "learning_rate": 7.307692307692308e-06, | |
| "loss": 0.8282, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.07936507936507936, | |
| "grad_norm": 18.259958267211914, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.8209, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08333333333333333, | |
| "grad_norm": 39.2940788269043, | |
| "learning_rate": 8.076923076923077e-06, | |
| "loss": 0.7472, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0873015873015873, | |
| "grad_norm": 22.495046615600586, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 1.0387, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.09126984126984126, | |
| "grad_norm": 6.657768726348877, | |
| "learning_rate": 8.846153846153847e-06, | |
| "loss": 1.2742, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.09523809523809523, | |
| "grad_norm": 5.994826793670654, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 1.176, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0992063492063492, | |
| "grad_norm": 5.183099746704102, | |
| "learning_rate": 9.615384615384616e-06, | |
| "loss": 1.0675, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.10317460317460317, | |
| "grad_norm": 4.059016227722168, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7689, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.10714285714285714, | |
| "grad_norm": 3.379298448562622, | |
| "learning_rate": 9.999892010284378e-06, | |
| "loss": 0.7588, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 2.759723663330078, | |
| "learning_rate": 9.999568045802216e-06, | |
| "loss": 0.699, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.11507936507936507, | |
| "grad_norm": 2.853120803833008, | |
| "learning_rate": 9.999028120547456e-06, | |
| "loss": 0.6595, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.11904761904761904, | |
| "grad_norm": 2.8823561668395996, | |
| "learning_rate": 9.99827225784264e-06, | |
| "loss": 0.7276, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12301587301587301, | |
| "grad_norm": 2.4436709880828857, | |
| "learning_rate": 9.99730049033793e-06, | |
| "loss": 0.6525, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.12698412698412698, | |
| "grad_norm": 2.590996265411377, | |
| "learning_rate": 9.996112860009689e-06, | |
| "loss": 0.706, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.13095238095238096, | |
| "grad_norm": 2.5507211685180664, | |
| "learning_rate": 9.994709418158652e-06, | |
| "loss": 0.6868, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.1349206349206349, | |
| "grad_norm": 2.1269373893737793, | |
| "learning_rate": 9.993090225407743e-06, | |
| "loss": 0.5903, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1388888888888889, | |
| "grad_norm": 2.204864263534546, | |
| "learning_rate": 9.991255351699422e-06, | |
| "loss": 0.6045, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 2.2101891040802, | |
| "learning_rate": 9.98920487629269e-06, | |
| "loss": 0.6233, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.14682539682539683, | |
| "grad_norm": 2.092217206954956, | |
| "learning_rate": 9.986938887759643e-06, | |
| "loss": 0.5969, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.15079365079365079, | |
| "grad_norm": 2.083162546157837, | |
| "learning_rate": 9.98445748398167e-06, | |
| "loss": 0.542, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.15476190476190477, | |
| "grad_norm": 2.029996871948242, | |
| "learning_rate": 9.981760772145201e-06, | |
| "loss": 0.5057, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.15873015873015872, | |
| "grad_norm": 2.119372606277466, | |
| "learning_rate": 9.978848868737099e-06, | |
| "loss": 0.5209, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1626984126984127, | |
| "grad_norm": 2.2384204864501953, | |
| "learning_rate": 9.975721899539607e-06, | |
| "loss": 0.5371, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.16666666666666666, | |
| "grad_norm": 2.2897753715515137, | |
| "learning_rate": 9.972379999624935e-06, | |
| "loss": 0.536, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.17063492063492064, | |
| "grad_norm": 2.219625234603882, | |
| "learning_rate": 9.968823313349412e-06, | |
| "loss": 0.4613, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.1746031746031746, | |
| "grad_norm": 2.27551007270813, | |
| "learning_rate": 9.96505199434725e-06, | |
| "loss": 0.4786, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.17857142857142858, | |
| "grad_norm": 2.279512405395508, | |
| "learning_rate": 9.961066205523917e-06, | |
| "loss": 0.4847, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.18253968253968253, | |
| "grad_norm": 2.447103500366211, | |
| "learning_rate": 9.956866119049095e-06, | |
| "loss": 0.5569, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1865079365079365, | |
| "grad_norm": 2.9808733463287354, | |
| "learning_rate": 9.952451916349242e-06, | |
| "loss": 0.6556, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.19047619047619047, | |
| "grad_norm": 2.7629613876342773, | |
| "learning_rate": 9.947823788099754e-06, | |
| "loss": 0.6049, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.19444444444444445, | |
| "grad_norm": 2.7128512859344482, | |
| "learning_rate": 9.942981934216731e-06, | |
| "loss": 0.5914, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.1984126984126984, | |
| "grad_norm": 3.130530595779419, | |
| "learning_rate": 9.937926563848345e-06, | |
| "loss": 0.6209, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.20238095238095238, | |
| "grad_norm": 3.015249490737915, | |
| "learning_rate": 9.9326578953658e-06, | |
| "loss": 0.7189, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.20634920634920634, | |
| "grad_norm": 2.8060615062713623, | |
| "learning_rate": 9.9271761563539e-06, | |
| "loss": 0.5967, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.21031746031746032, | |
| "grad_norm": 2.6807661056518555, | |
| "learning_rate": 9.921481583601218e-06, | |
| "loss": 0.5586, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 2.776455879211426, | |
| "learning_rate": 9.915574423089872e-06, | |
| "loss": 0.5792, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.21825396825396826, | |
| "grad_norm": 2.8880770206451416, | |
| "learning_rate": 9.909454929984894e-06, | |
| "loss": 0.556, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 3.1384806632995605, | |
| "learning_rate": 9.903123368623216e-06, | |
| "loss": 0.5644, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.2261904761904762, | |
| "grad_norm": 3.1627347469329834, | |
| "learning_rate": 9.896580012502238e-06, | |
| "loss": 0.5096, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.23015873015873015, | |
| "grad_norm": 3.3309388160705566, | |
| "learning_rate": 9.889825144268029e-06, | |
| "loss": 0.5594, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.23412698412698413, | |
| "grad_norm": 3.1470706462860107, | |
| "learning_rate": 9.882859055703109e-06, | |
| "loss": 0.537, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.23809523809523808, | |
| "grad_norm": 3.1385302543640137, | |
| "learning_rate": 9.875682047713847e-06, | |
| "loss": 0.5098, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.24206349206349206, | |
| "grad_norm": 3.0547971725463867, | |
| "learning_rate": 9.868294430317464e-06, | |
| "loss": 0.4684, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.24603174603174602, | |
| "grad_norm": 3.0865085124969482, | |
| "learning_rate": 9.860696522628638e-06, | |
| "loss": 0.4754, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 3.402650833129883, | |
| "learning_rate": 9.852888652845729e-06, | |
| "loss": 0.5196, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.25396825396825395, | |
| "grad_norm": 3.3936548233032227, | |
| "learning_rate": 9.84487115823659e-06, | |
| "loss": 0.4685, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.25793650793650796, | |
| "grad_norm": 3.30293869972229, | |
| "learning_rate": 9.836644385124006e-06, | |
| "loss": 0.4312, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.2619047619047619, | |
| "grad_norm": 3.3731415271759033, | |
| "learning_rate": 9.828208688870736e-06, | |
| "loss": 0.4573, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.26587301587301587, | |
| "grad_norm": 3.361337661743164, | |
| "learning_rate": 9.81956443386415e-06, | |
| "loss": 0.4268, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.2698412698412698, | |
| "grad_norm": 3.448617458343506, | |
| "learning_rate": 9.810711993500506e-06, | |
| "loss": 0.4354, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.27380952380952384, | |
| "grad_norm": 3.015842914581299, | |
| "learning_rate": 9.801651750168815e-06, | |
| "loss": 0.362, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2777777777777778, | |
| "grad_norm": 3.302525043487549, | |
| "learning_rate": 9.792384095234312e-06, | |
| "loss": 0.397, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.28174603174603174, | |
| "grad_norm": 3.5341339111328125, | |
| "learning_rate": 9.782909429021568e-06, | |
| "loss": 0.4057, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 3.903911828994751, | |
| "learning_rate": 9.773228160797187e-06, | |
| "loss": 0.4724, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.2896825396825397, | |
| "grad_norm": 3.7740511894226074, | |
| "learning_rate": 9.76334070875213e-06, | |
| "loss": 0.5131, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.29365079365079366, | |
| "grad_norm": 3.9347620010375977, | |
| "learning_rate": 9.753247499983649e-06, | |
| "loss": 0.4401, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.2976190476190476, | |
| "grad_norm": 4.2350358963012695, | |
| "learning_rate": 9.742948970476845e-06, | |
| "loss": 0.4372, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.30158730158730157, | |
| "grad_norm": 3.9649524688720703, | |
| "learning_rate": 9.732445565085823e-06, | |
| "loss": 0.4933, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.3055555555555556, | |
| "grad_norm": 4.056861400604248, | |
| "learning_rate": 9.721737737514492e-06, | |
| "loss": 0.4634, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.30952380952380953, | |
| "grad_norm": 4.233841419219971, | |
| "learning_rate": 9.71082595029695e-06, | |
| "loss": 0.4573, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.3134920634920635, | |
| "grad_norm": 4.390049934387207, | |
| "learning_rate": 9.699710674777519e-06, | |
| "loss": 0.461, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.31746031746031744, | |
| "grad_norm": 4.465388774871826, | |
| "learning_rate": 9.688392391090374e-06, | |
| "loss": 0.4396, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.32142857142857145, | |
| "grad_norm": 4.317083835601807, | |
| "learning_rate": 9.676871588138812e-06, | |
| "loss": 0.4186, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.3253968253968254, | |
| "grad_norm": 4.572980880737305, | |
| "learning_rate": 9.665148763574123e-06, | |
| "loss": 0.3889, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.32936507936507936, | |
| "grad_norm": 4.19005823135376, | |
| "learning_rate": 9.653224423774107e-06, | |
| "loss": 0.3759, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 4.183736801147461, | |
| "learning_rate": 9.64109908382119e-06, | |
| "loss": 0.3515, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.3373015873015873, | |
| "grad_norm": 4.017164707183838, | |
| "learning_rate": 9.628773267480177e-06, | |
| "loss": 0.3651, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.3412698412698413, | |
| "grad_norm": 4.175520420074463, | |
| "learning_rate": 9.616247507175624e-06, | |
| "loss": 0.4147, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.34523809523809523, | |
| "grad_norm": 4.339958667755127, | |
| "learning_rate": 9.603522343968852e-06, | |
| "loss": 0.3841, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3492063492063492, | |
| "grad_norm": 3.773083448410034, | |
| "learning_rate": 9.590598327534563e-06, | |
| "loss": 0.309, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.3531746031746032, | |
| "grad_norm": 3.596532106399536, | |
| "learning_rate": 9.577476016137105e-06, | |
| "loss": 0.293, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 3.574747323989868, | |
| "learning_rate": 9.56415597660634e-06, | |
| "loss": 0.2856, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3611111111111111, | |
| "grad_norm": 3.69950008392334, | |
| "learning_rate": 9.550638784313187e-06, | |
| "loss": 0.294, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.36507936507936506, | |
| "grad_norm": 3.8686187267303467, | |
| "learning_rate": 9.536925023144742e-06, | |
| "loss": 0.2925, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.36904761904761907, | |
| "grad_norm": 3.1374640464782715, | |
| "learning_rate": 9.523015285479076e-06, | |
| "loss": 0.2522, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.373015873015873, | |
| "grad_norm": 3.073716640472412, | |
| "learning_rate": 9.508910172159635e-06, | |
| "loss": 0.2562, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.376984126984127, | |
| "grad_norm": 3.2597389221191406, | |
| "learning_rate": 9.494610292469287e-06, | |
| "loss": 0.2603, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.38095238095238093, | |
| "grad_norm": 3.1903927326202393, | |
| "learning_rate": 9.48011626410401e-06, | |
| "loss": 0.2337, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.38492063492063494, | |
| "grad_norm": 3.3626108169555664, | |
| "learning_rate": 9.465428713146206e-06, | |
| "loss": 0.3231, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3888888888888889, | |
| "grad_norm": 3.21762752532959, | |
| "learning_rate": 9.450548274037652e-06, | |
| "loss": 0.336, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.39285714285714285, | |
| "grad_norm": 3.2634785175323486, | |
| "learning_rate": 9.435475589552107e-06, | |
| "loss": 0.3261, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.3968253968253968, | |
| "grad_norm": 3.6221210956573486, | |
| "learning_rate": 9.420211310767534e-06, | |
| "loss": 0.3475, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4007936507936508, | |
| "grad_norm": 3.105833053588867, | |
| "learning_rate": 9.40475609703798e-06, | |
| "loss": 0.3846, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.40476190476190477, | |
| "grad_norm": 3.1111133098602295, | |
| "learning_rate": 9.389110615965102e-06, | |
| "loss": 0.3029, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.4087301587301587, | |
| "grad_norm": 3.2716522216796875, | |
| "learning_rate": 9.37327554336932e-06, | |
| "loss": 0.3478, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.4126984126984127, | |
| "grad_norm": 3.0953073501586914, | |
| "learning_rate": 9.35725156326063e-06, | |
| "loss": 0.2974, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.4166666666666667, | |
| "grad_norm": 2.904229164123535, | |
| "learning_rate": 9.341039367809056e-06, | |
| "loss": 0.2466, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.42063492063492064, | |
| "grad_norm": 3.0948331356048584, | |
| "learning_rate": 9.324639657314742e-06, | |
| "loss": 0.2547, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.4246031746031746, | |
| "grad_norm": 2.9093923568725586, | |
| "learning_rate": 9.308053140177722e-06, | |
| "loss": 0.2844, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 2.973367691040039, | |
| "learning_rate": 9.291280532867301e-06, | |
| "loss": 0.2226, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.43253968253968256, | |
| "grad_norm": 2.961211681365967, | |
| "learning_rate": 9.27432255989112e-06, | |
| "loss": 0.2088, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.4365079365079365, | |
| "grad_norm": 2.6171209812164307, | |
| "learning_rate": 9.257179953763846e-06, | |
| "loss": 0.2068, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.44047619047619047, | |
| "grad_norm": 2.5742547512054443, | |
| "learning_rate": 9.239853454975548e-06, | |
| "loss": 0.2371, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 2.2769546508789062, | |
| "learning_rate": 9.222343811959694e-06, | |
| "loss": 0.2152, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.44841269841269843, | |
| "grad_norm": 2.1215806007385254, | |
| "learning_rate": 9.204651781060832e-06, | |
| "loss": 0.179, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.4523809523809524, | |
| "grad_norm": 1.9973154067993164, | |
| "learning_rate": 9.186778126501916e-06, | |
| "loss": 0.1877, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.45634920634920634, | |
| "grad_norm": 1.805450201034546, | |
| "learning_rate": 9.168723620351298e-06, | |
| "loss": 0.197, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.4603174603174603, | |
| "grad_norm": 1.7883168458938599, | |
| "learning_rate": 9.150489042489368e-06, | |
| "loss": 0.1725, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.4642857142857143, | |
| "grad_norm": 1.5096471309661865, | |
| "learning_rate": 9.13207518057488e-06, | |
| "loss": 0.1594, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.46825396825396826, | |
| "grad_norm": 1.4532548189163208, | |
| "learning_rate": 9.113482830010918e-06, | |
| "loss": 0.1641, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.4722222222222222, | |
| "grad_norm": 1.3101332187652588, | |
| "learning_rate": 9.094712793910541e-06, | |
| "loss": 0.1508, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.47619047619047616, | |
| "grad_norm": 1.2906259298324585, | |
| "learning_rate": 9.075765883062093e-06, | |
| "loss": 0.1624, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4801587301587302, | |
| "grad_norm": 1.3046653270721436, | |
| "learning_rate": 9.056642915894182e-06, | |
| "loss": 0.1947, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.48412698412698413, | |
| "grad_norm": 1.424941062927246, | |
| "learning_rate": 9.037344718440321e-06, | |
| "loss": 0.2381, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4880952380952381, | |
| "grad_norm": 1.2889655828475952, | |
| "learning_rate": 9.017872124303255e-06, | |
| "loss": 0.2395, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.49206349206349204, | |
| "grad_norm": 1.39620041847229, | |
| "learning_rate": 8.99822597461894e-06, | |
| "loss": 0.1924, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.49603174603174605, | |
| "grad_norm": 1.2884349822998047, | |
| "learning_rate": 8.978407118020226e-06, | |
| "loss": 0.1972, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.5466859340667725, | |
| "learning_rate": 8.958416410600188e-06, | |
| "loss": 0.3101, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.503968253968254, | |
| "grad_norm": 1.4077582359313965, | |
| "learning_rate": 8.938254715875152e-06, | |
| "loss": 0.2774, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.5079365079365079, | |
| "grad_norm": 1.34942626953125, | |
| "learning_rate": 8.917922904747385e-06, | |
| "loss": 0.2537, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.5119047619047619, | |
| "grad_norm": 1.1995731592178345, | |
| "learning_rate": 8.897421855467491e-06, | |
| "loss": 0.2633, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.5158730158730159, | |
| "grad_norm": 1.0786157846450806, | |
| "learning_rate": 8.876752453596462e-06, | |
| "loss": 0.2352, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5198412698412699, | |
| "grad_norm": 1.173683762550354, | |
| "learning_rate": 8.85591559196743e-06, | |
| "loss": 0.2088, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.5238095238095238, | |
| "grad_norm": 1.1212071180343628, | |
| "learning_rate": 8.834912170647102e-06, | |
| "loss": 0.1927, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.5277777777777778, | |
| "grad_norm": 1.0188961029052734, | |
| "learning_rate": 8.813743096896872e-06, | |
| "loss": 0.1701, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.5317460317460317, | |
| "grad_norm": 1.1430011987686157, | |
| "learning_rate": 8.792409285133644e-06, | |
| "loss": 0.2245, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.5357142857142857, | |
| "grad_norm": 0.9560830593109131, | |
| "learning_rate": 8.770911656890325e-06, | |
| "loss": 0.1723, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5396825396825397, | |
| "grad_norm": 1.0917646884918213, | |
| "learning_rate": 8.749251140776016e-06, | |
| "loss": 0.1941, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.5436507936507936, | |
| "grad_norm": 1.1524584293365479, | |
| "learning_rate": 8.727428672435911e-06, | |
| "loss": 0.1797, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.5476190476190477, | |
| "grad_norm": 1.0847519636154175, | |
| "learning_rate": 8.705445194510868e-06, | |
| "loss": 0.1807, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.5515873015873016, | |
| "grad_norm": 0.9003533720970154, | |
| "learning_rate": 8.6833016565967e-06, | |
| "loss": 0.1512, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 1.016819953918457, | |
| "learning_rate": 8.660999015203152e-06, | |
| "loss": 0.1559, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5595238095238095, | |
| "grad_norm": 1.136342167854309, | |
| "learning_rate": 8.638538233712581e-06, | |
| "loss": 0.1931, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5634920634920635, | |
| "grad_norm": 0.9974120259284973, | |
| "learning_rate": 8.615920282338355e-06, | |
| "loss": 0.1535, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5674603174603174, | |
| "grad_norm": 1.1280145645141602, | |
| "learning_rate": 8.593146138082925e-06, | |
| "loss": 0.176, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 1.0223793983459473, | |
| "learning_rate": 8.570216784695637e-06, | |
| "loss": 0.1499, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5753968253968254, | |
| "grad_norm": 0.9792421460151672, | |
| "learning_rate": 8.54713321263023e-06, | |
| "loss": 0.1267, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5793650793650794, | |
| "grad_norm": 1.0992575883865356, | |
| "learning_rate": 8.52389641900206e-06, | |
| "loss": 0.1431, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.5833333333333334, | |
| "grad_norm": 1.6400508880615234, | |
| "learning_rate": 8.50050740754502e-06, | |
| "loss": 0.272, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5873015873015873, | |
| "grad_norm": 1.57867431640625, | |
| "learning_rate": 8.476967188568187e-06, | |
| "loss": 0.2249, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5912698412698413, | |
| "grad_norm": 1.2956342697143555, | |
| "learning_rate": 8.453276778912186e-06, | |
| "loss": 0.2267, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5952380952380952, | |
| "grad_norm": 1.2199386358261108, | |
| "learning_rate": 8.429437201905254e-06, | |
| "loss": 0.1851, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5992063492063492, | |
| "grad_norm": 1.5200244188308716, | |
| "learning_rate": 8.405449487319049e-06, | |
| "loss": 0.3301, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.6031746031746031, | |
| "grad_norm": 1.3942667245864868, | |
| "learning_rate": 8.38131467132416e-06, | |
| "loss": 0.2049, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.6071428571428571, | |
| "grad_norm": 1.3229930400848389, | |
| "learning_rate": 8.357033796445356e-06, | |
| "loss": 0.2388, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.6111111111111112, | |
| "grad_norm": 1.3120101690292358, | |
| "learning_rate": 8.332607911516545e-06, | |
| "loss": 0.1985, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.6150793650793651, | |
| "grad_norm": 1.3684817552566528, | |
| "learning_rate": 8.308038071635475e-06, | |
| "loss": 0.1831, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.6190476190476191, | |
| "grad_norm": 1.4442847967147827, | |
| "learning_rate": 8.283325338118154e-06, | |
| "loss": 0.2086, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.623015873015873, | |
| "grad_norm": 1.4207266569137573, | |
| "learning_rate": 8.258470778453005e-06, | |
| "loss": 0.1743, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.626984126984127, | |
| "grad_norm": 1.4148527383804321, | |
| "learning_rate": 8.233475466254766e-06, | |
| "loss": 0.1797, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.6309523809523809, | |
| "grad_norm": 1.449039101600647, | |
| "learning_rate": 8.208340481218094e-06, | |
| "loss": 0.1672, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.6349206349206349, | |
| "grad_norm": 1.5011506080627441, | |
| "learning_rate": 8.183066909070946e-06, | |
| "loss": 0.1616, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6388888888888888, | |
| "grad_norm": 1.4110753536224365, | |
| "learning_rate": 8.15765584152767e-06, | |
| "loss": 0.1715, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 1.4286863803863525, | |
| "learning_rate": 8.132108376241849e-06, | |
| "loss": 0.19, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.6468253968253969, | |
| "grad_norm": 1.3429644107818604, | |
| "learning_rate": 8.106425616758886e-06, | |
| "loss": 0.1569, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.6507936507936508, | |
| "grad_norm": 1.2959730625152588, | |
| "learning_rate": 8.08060867246834e-06, | |
| "loss": 0.1689, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.6547619047619048, | |
| "grad_norm": 1.3238943815231323, | |
| "learning_rate": 8.054658658555998e-06, | |
| "loss": 0.1392, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6587301587301587, | |
| "grad_norm": 1.4001777172088623, | |
| "learning_rate": 8.028576695955711e-06, | |
| "loss": 0.1717, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.6626984126984127, | |
| "grad_norm": 1.26632821559906, | |
| "learning_rate": 8.002363911300966e-06, | |
| "loss": 0.1607, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 1.1917613744735718, | |
| "learning_rate": 7.976021436876232e-06, | |
| "loss": 0.1198, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6706349206349206, | |
| "grad_norm": 1.226851224899292, | |
| "learning_rate": 7.949550410568033e-06, | |
| "loss": 0.1263, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6746031746031746, | |
| "grad_norm": 1.2502480745315552, | |
| "learning_rate": 7.92295197581581e-06, | |
| "loss": 0.115, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6785714285714286, | |
| "grad_norm": 1.3546652793884277, | |
| "learning_rate": 7.89622728156253e-06, | |
| "loss": 0.1472, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6825396825396826, | |
| "grad_norm": 1.7175713777542114, | |
| "learning_rate": 7.869377482205042e-06, | |
| "loss": 0.2404, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.6865079365079365, | |
| "grad_norm": 1.5571590662002563, | |
| "learning_rate": 7.842403737544226e-06, | |
| "loss": 0.1928, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.6904761904761905, | |
| "grad_norm": 1.344901442527771, | |
| "learning_rate": 7.815307212734888e-06, | |
| "loss": 0.1551, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.6944444444444444, | |
| "grad_norm": 1.4249378442764282, | |
| "learning_rate": 7.788089078235432e-06, | |
| "loss": 0.1925, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6984126984126984, | |
| "grad_norm": 1.7427445650100708, | |
| "learning_rate": 7.7607505097573e-06, | |
| "loss": 0.2586, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.7023809523809523, | |
| "grad_norm": 1.585457444190979, | |
| "learning_rate": 7.733292688214182e-06, | |
| "loss": 0.2072, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.7063492063492064, | |
| "grad_norm": 1.5098254680633545, | |
| "learning_rate": 7.705716799671019e-06, | |
| "loss": 0.2175, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.7103174603174603, | |
| "grad_norm": 1.3991138935089111, | |
| "learning_rate": 7.678024035292757e-06, | |
| "loss": 0.2066, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 1.357462763786316, | |
| "learning_rate": 7.650215591292888e-06, | |
| "loss": 0.1824, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7182539682539683, | |
| "grad_norm": 1.4721425771713257, | |
| "learning_rate": 7.622292668881805e-06, | |
| "loss": 0.1863, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.7222222222222222, | |
| "grad_norm": 1.399949550628662, | |
| "learning_rate": 7.594256474214883e-06, | |
| "loss": 0.1471, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.7261904761904762, | |
| "grad_norm": 1.3106352090835571, | |
| "learning_rate": 7.566108218340399e-06, | |
| "loss": 0.142, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.7301587301587301, | |
| "grad_norm": 1.3351104259490967, | |
| "learning_rate": 7.537849117147212e-06, | |
| "loss": 0.1183, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.7341269841269841, | |
| "grad_norm": 1.3894598484039307, | |
| "learning_rate": 7.509480391312243e-06, | |
| "loss": 0.1601, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.7380952380952381, | |
| "grad_norm": 1.2343592643737793, | |
| "learning_rate": 7.481003266247745e-06, | |
| "loss": 0.1422, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.7420634920634921, | |
| "grad_norm": 1.1929510831832886, | |
| "learning_rate": 7.452418972048372e-06, | |
| "loss": 0.1443, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.746031746031746, | |
| "grad_norm": 1.0833501815795898, | |
| "learning_rate": 7.4237287434380485e-06, | |
| "loss": 0.1174, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.089729905128479, | |
| "learning_rate": 7.394933819716625e-06, | |
| "loss": 0.1145, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.753968253968254, | |
| "grad_norm": 1.0947588682174683, | |
| "learning_rate": 7.366035444706346e-06, | |
| "loss": 0.1134, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.7579365079365079, | |
| "grad_norm": 1.125807285308838, | |
| "learning_rate": 7.337034866698138e-06, | |
| "loss": 0.115, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.7619047619047619, | |
| "grad_norm": 1.0649607181549072, | |
| "learning_rate": 7.307933338397667e-06, | |
| "loss": 0.121, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.7658730158730159, | |
| "grad_norm": 1.12957763671875, | |
| "learning_rate": 7.278732116871239e-06, | |
| "loss": 0.1215, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.7698412698412699, | |
| "grad_norm": 1.0527234077453613, | |
| "learning_rate": 7.249432463491498e-06, | |
| "loss": 0.1171, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.7738095238095238, | |
| "grad_norm": 0.9681161046028137, | |
| "learning_rate": 7.220035643882938e-06, | |
| "loss": 0.1023, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 0.9377241134643555, | |
| "learning_rate": 7.190542927867234e-06, | |
| "loss": 0.093, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7817460317460317, | |
| "grad_norm": 1.9532684087753296, | |
| "learning_rate": 7.160955589408395e-06, | |
| "loss": 0.241, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 1.528989553451538, | |
| "learning_rate": 7.131274906557725e-06, | |
| "loss": 0.2079, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.7896825396825397, | |
| "grad_norm": 2.8748903274536133, | |
| "learning_rate": 7.101502161398626e-06, | |
| "loss": 0.1578, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.7936507936507936, | |
| "grad_norm": 1.079186201095581, | |
| "learning_rate": 7.0716386399912075e-06, | |
| "loss": 0.1487, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7976190476190477, | |
| "grad_norm": 1.2633140087127686, | |
| "learning_rate": 7.041685632316748e-06, | |
| "loss": 0.1666, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.8015873015873016, | |
| "grad_norm": 1.2365785837173462, | |
| "learning_rate": 7.0116444322219575e-06, | |
| "loss": 0.1655, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.8055555555555556, | |
| "grad_norm": 1.3451050519943237, | |
| "learning_rate": 6.981516337363099e-06, | |
| "loss": 0.2007, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.8095238095238095, | |
| "grad_norm": 1.210461139678955, | |
| "learning_rate": 6.95130264914993e-06, | |
| "loss": 0.1565, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.8134920634920635, | |
| "grad_norm": 1.1013389825820923, | |
| "learning_rate": 6.9210046726894885e-06, | |
| "loss": 0.1658, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.8174603174603174, | |
| "grad_norm": 1.211225152015686, | |
| "learning_rate": 6.890623716729724e-06, | |
| "loss": 0.1427, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.8214285714285714, | |
| "grad_norm": 1.0477981567382812, | |
| "learning_rate": 6.860161093602949e-06, | |
| "loss": 0.1355, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.8253968253968254, | |
| "grad_norm": 0.9986271262168884, | |
| "learning_rate": 6.829618119169169e-06, | |
| "loss": 0.121, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.8293650793650794, | |
| "grad_norm": 0.9779046177864075, | |
| "learning_rate": 6.798996112759233e-06, | |
| "loss": 0.0976, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 0.9704580307006836, | |
| "learning_rate": 6.768296397117848e-06, | |
| "loss": 0.0923, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.8373015873015873, | |
| "grad_norm": 1.0493831634521484, | |
| "learning_rate": 6.737520298346438e-06, | |
| "loss": 0.1463, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.8412698412698413, | |
| "grad_norm": 0.9381864070892334, | |
| "learning_rate": 6.706669145845863e-06, | |
| "loss": 0.123, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.8452380952380952, | |
| "grad_norm": 0.9599462151527405, | |
| "learning_rate": 6.6757442722590015e-06, | |
| "loss": 0.1263, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.8492063492063492, | |
| "grad_norm": 0.8259004354476929, | |
| "learning_rate": 6.6447470134131685e-06, | |
| "loss": 0.0963, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.8531746031746031, | |
| "grad_norm": 1.3779510259628296, | |
| "learning_rate": 6.613678708262439e-06, | |
| "loss": 0.1151, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 0.8121216297149658, | |
| "learning_rate": 6.5825406988297815e-06, | |
| "loss": 0.0882, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.8611111111111112, | |
| "grad_norm": 0.8545671105384827, | |
| "learning_rate": 6.551334330149114e-06, | |
| "loss": 0.1035, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.8650793650793651, | |
| "grad_norm": 0.9634934067726135, | |
| "learning_rate": 6.520060950207186e-06, | |
| "loss": 0.1139, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.8690476190476191, | |
| "grad_norm": 0.7680690884590149, | |
| "learning_rate": 6.488721909885359e-06, | |
| "loss": 0.0953, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.873015873015873, | |
| "grad_norm": 0.8121854066848755, | |
| "learning_rate": 6.457318562901257e-06, | |
| "loss": 0.1031, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.876984126984127, | |
| "grad_norm": 1.0451233386993408, | |
| "learning_rate": 6.425852265750282e-06, | |
| "loss": 0.1359, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.8809523809523809, | |
| "grad_norm": 1.2654521465301514, | |
| "learning_rate": 6.394324377647028e-06, | |
| "loss": 0.1926, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.8849206349206349, | |
| "grad_norm": 1.0600279569625854, | |
| "learning_rate": 6.362736260466561e-06, | |
| "loss": 0.1337, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 1.101129412651062, | |
| "learning_rate": 6.331089278685599e-06, | |
| "loss": 0.1979, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.8928571428571429, | |
| "grad_norm": 1.0881141424179077, | |
| "learning_rate": 6.299384799323568e-06, | |
| "loss": 0.1261, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.8968253968253969, | |
| "grad_norm": 1.2393383979797363, | |
| "learning_rate": 6.267624191883551e-06, | |
| "loss": 0.2384, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.9007936507936508, | |
| "grad_norm": 0.9808958172798157, | |
| "learning_rate": 6.235808828293135e-06, | |
| "loss": 0.1481, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.9047619047619048, | |
| "grad_norm": 1.0703552961349487, | |
| "learning_rate": 6.203940082845144e-06, | |
| "loss": 0.1392, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.9087301587301587, | |
| "grad_norm": 1.0855351686477661, | |
| "learning_rate": 6.172019332138285e-06, | |
| "loss": 0.1805, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.9126984126984127, | |
| "grad_norm": 1.043116807937622, | |
| "learning_rate": 6.140047955017672e-06, | |
| "loss": 0.1467, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.9166666666666666, | |
| "grad_norm": 1.0198463201522827, | |
| "learning_rate": 6.108027332515276e-06, | |
| "loss": 0.1494, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.9206349206349206, | |
| "grad_norm": 0.886110246181488, | |
| "learning_rate": 6.075958847790262e-06, | |
| "loss": 0.1226, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.9246031746031746, | |
| "grad_norm": 0.9238107800483704, | |
| "learning_rate": 6.043843886069251e-06, | |
| "loss": 0.1209, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 0.952556848526001, | |
| "learning_rate": 6.011683834586474e-06, | |
| "loss": 0.1165, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.9325396825396826, | |
| "grad_norm": 0.9139618873596191, | |
| "learning_rate": 5.979480082523858e-06, | |
| "loss": 0.1027, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.9365079365079365, | |
| "grad_norm": 0.9542132019996643, | |
| "learning_rate": 5.947234020951015e-06, | |
| "loss": 0.1337, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.9404761904761905, | |
| "grad_norm": 0.9156303405761719, | |
| "learning_rate": 5.914947042765149e-06, | |
| "loss": 0.1217, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.9444444444444444, | |
| "grad_norm": 0.8287809491157532, | |
| "learning_rate": 5.882620542630901e-06, | |
| "loss": 0.1131, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.9484126984126984, | |
| "grad_norm": 0.7955264449119568, | |
| "learning_rate": 5.850255916920093e-06, | |
| "loss": 0.1047, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.9523809523809523, | |
| "grad_norm": 0.9561585783958435, | |
| "learning_rate": 5.817854563651415e-06, | |
| "loss": 0.0907, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.9563492063492064, | |
| "grad_norm": 0.8837376832962036, | |
| "learning_rate": 5.785417882430035e-06, | |
| "loss": 0.1123, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.9603174603174603, | |
| "grad_norm": 0.7713195085525513, | |
| "learning_rate": 5.752947274387147e-06, | |
| "loss": 0.0926, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.9642857142857143, | |
| "grad_norm": 0.7587618231773376, | |
| "learning_rate": 5.720444142119445e-06, | |
| "loss": 0.0912, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.9682539682539683, | |
| "grad_norm": 0.7779955267906189, | |
| "learning_rate": 5.687909889628529e-06, | |
| "loss": 0.0763, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.9722222222222222, | |
| "grad_norm": 0.7983916997909546, | |
| "learning_rate": 5.6553459222602714e-06, | |
| "loss": 0.0932, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.9761904761904762, | |
| "grad_norm": 0.7773364186286926, | |
| "learning_rate": 5.622753646644102e-06, | |
| "loss": 0.0883, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.9801587301587301, | |
| "grad_norm": 5.570642471313477, | |
| "learning_rate": 5.59013447063225e-06, | |
| "loss": 0.1963, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.9841269841269841, | |
| "grad_norm": 1.3332061767578125, | |
| "learning_rate": 5.557489803238934e-06, | |
| "loss": 0.1468, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.9880952380952381, | |
| "grad_norm": 1.0517665147781372, | |
| "learning_rate": 5.524821054579491e-06, | |
| "loss": 0.1464, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.9920634920634921, | |
| "grad_norm": 0.9235040545463562, | |
| "learning_rate": 5.492129635809473e-06, | |
| "loss": 0.1276, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.996031746031746, | |
| "grad_norm": 0.9754127860069275, | |
| "learning_rate": 5.459416959063688e-06, | |
| "loss": 0.1326, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.0198100805282593, | |
| "learning_rate": 5.426684437395196e-06, | |
| "loss": 0.1658, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.003968253968254, | |
| "grad_norm": 1.043383002281189, | |
| "learning_rate": 5.393933484714284e-06, | |
| "loss": 0.1776, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.007936507936508, | |
| "grad_norm": 0.913267970085144, | |
| "learning_rate": 5.361165515727374e-06, | |
| "loss": 0.1503, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.0119047619047619, | |
| "grad_norm": 0.9391182065010071, | |
| "learning_rate": 5.328381945875933e-06, | |
| "loss": 0.133, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.0158730158730158, | |
| "grad_norm": 0.8706400394439697, | |
| "learning_rate": 5.295584191275308e-06, | |
| "loss": 0.1221, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.0198412698412698, | |
| "grad_norm": 0.8898656368255615, | |
| "learning_rate": 5.26277366865358e-06, | |
| "loss": 0.1204, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.0238095238095237, | |
| "grad_norm": 0.8045791983604431, | |
| "learning_rate": 5.229951795290353e-06, | |
| "loss": 0.0911, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.0277777777777777, | |
| "grad_norm": 0.7952799201011658, | |
| "learning_rate": 5.197119988955534e-06, | |
| "loss": 0.0929, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.0317460317460316, | |
| "grad_norm": 0.7498448491096497, | |
| "learning_rate": 5.164279667848094e-06, | |
| "loss": 0.0867, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.0357142857142858, | |
| "grad_norm": 0.8970192670822144, | |
| "learning_rate": 5.131432250534809e-06, | |
| "loss": 0.1042, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.0396825396825398, | |
| "grad_norm": 0.7270601391792297, | |
| "learning_rate": 5.0985791558889785e-06, | |
| "loss": 0.0738, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.0436507936507937, | |
| "grad_norm": 0.8380938172340393, | |
| "learning_rate": 5.065721803029146e-06, | |
| "loss": 0.1087, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.0476190476190477, | |
| "grad_norm": 0.8650514483451843, | |
| "learning_rate": 5.032861611257783e-06, | |
| "loss": 0.1023, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.0515873015873016, | |
| "grad_norm": 0.8322713971138, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0902, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.0555555555555556, | |
| "grad_norm": 0.7874230742454529, | |
| "learning_rate": 4.967138388742218e-06, | |
| "loss": 0.0766, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.0595238095238095, | |
| "grad_norm": 0.8071075677871704, | |
| "learning_rate": 4.934278196970857e-06, | |
| "loss": 0.0796, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.0634920634920635, | |
| "grad_norm": 0.9313676357269287, | |
| "learning_rate": 4.9014208441110215e-06, | |
| "loss": 0.0844, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.0674603174603174, | |
| "grad_norm": 0.7492650151252747, | |
| "learning_rate": 4.868567749465192e-06, | |
| "loss": 0.0728, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.0714285714285714, | |
| "grad_norm": 0.6248689889907837, | |
| "learning_rate": 4.835720332151907e-06, | |
| "loss": 0.0579, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.0753968253968254, | |
| "grad_norm": 0.813044011592865, | |
| "learning_rate": 4.802880011044467e-06, | |
| "loss": 0.0824, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.0793650793650793, | |
| "grad_norm": 0.7809593081474304, | |
| "learning_rate": 4.770048204709648e-06, | |
| "loss": 0.0647, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.0833333333333333, | |
| "grad_norm": 0.826143741607666, | |
| "learning_rate": 4.73722633134642e-06, | |
| "loss": 0.0729, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.0873015873015872, | |
| "grad_norm": 0.9692986011505127, | |
| "learning_rate": 4.7044158087246926e-06, | |
| "loss": 0.1141, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.0912698412698412, | |
| "grad_norm": 1.088172197341919, | |
| "learning_rate": 4.67161805412407e-06, | |
| "loss": 0.1238, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.0952380952380953, | |
| "grad_norm": 0.7897351384162903, | |
| "learning_rate": 4.6388344842726266e-06, | |
| "loss": 0.095, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.0992063492063493, | |
| "grad_norm": 0.7610841393470764, | |
| "learning_rate": 4.606066515285719e-06, | |
| "loss": 0.1081, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.1031746031746033, | |
| "grad_norm": 1.0061901807785034, | |
| "learning_rate": 4.573315562604804e-06, | |
| "loss": 0.15, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.1071428571428572, | |
| "grad_norm": 0.9851556420326233, | |
| "learning_rate": 4.540583040936313e-06, | |
| "loss": 0.102, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 0.9912914037704468, | |
| "learning_rate": 4.5078703641905275e-06, | |
| "loss": 0.0999, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.1150793650793651, | |
| "grad_norm": 0.8964777588844299, | |
| "learning_rate": 4.4751789454205105e-06, | |
| "loss": 0.0914, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.119047619047619, | |
| "grad_norm": 0.869835376739502, | |
| "learning_rate": 4.442510196761068e-06, | |
| "loss": 0.0898, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.123015873015873, | |
| "grad_norm": 0.842560887336731, | |
| "learning_rate": 4.409865529367751e-06, | |
| "loss": 0.095, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.126984126984127, | |
| "grad_norm": 0.8103805780410767, | |
| "learning_rate": 4.377246353355899e-06, | |
| "loss": 0.0942, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.130952380952381, | |
| "grad_norm": 0.7182227373123169, | |
| "learning_rate": 4.34465407773973e-06, | |
| "loss": 0.0801, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.1349206349206349, | |
| "grad_norm": 0.7724276781082153, | |
| "learning_rate": 4.312090110371473e-06, | |
| "loss": 0.0883, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.1388888888888888, | |
| "grad_norm": 0.7744220495223999, | |
| "learning_rate": 4.279555857880558e-06, | |
| "loss": 0.0751, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 0.7078888416290283, | |
| "learning_rate": 4.247052725612853e-06, | |
| "loss": 0.0792, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.1468253968253967, | |
| "grad_norm": 0.7455774545669556, | |
| "learning_rate": 4.214582117569966e-06, | |
| "loss": 0.0799, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.1507936507936507, | |
| "grad_norm": 0.7366694211959839, | |
| "learning_rate": 4.182145436348587e-06, | |
| "loss": 0.0726, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.1547619047619047, | |
| "grad_norm": 0.6610168814659119, | |
| "learning_rate": 4.1497440830799084e-06, | |
| "loss": 0.0691, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.1587301587301586, | |
| "grad_norm": 0.7766444683074951, | |
| "learning_rate": 4.1173794573691e-06, | |
| "loss": 0.0747, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.1626984126984128, | |
| "grad_norm": 0.765711784362793, | |
| "learning_rate": 4.0850529572348505e-06, | |
| "loss": 0.0724, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.1666666666666667, | |
| "grad_norm": 0.7582992911338806, | |
| "learning_rate": 4.052765979048986e-06, | |
| "loss": 0.0775, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.1706349206349207, | |
| "grad_norm": 0.7233676314353943, | |
| "learning_rate": 4.0205199174761435e-06, | |
| "loss": 0.0593, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.1746031746031746, | |
| "grad_norm": 0.7994574904441833, | |
| "learning_rate": 3.988316165413528e-06, | |
| "loss": 0.0728, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.1785714285714286, | |
| "grad_norm": 0.7465212941169739, | |
| "learning_rate": 3.956156113930752e-06, | |
| "loss": 0.0608, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.1825396825396826, | |
| "grad_norm": 0.7673329710960388, | |
| "learning_rate": 3.924041152209739e-06, | |
| "loss": 0.1052, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.1865079365079365, | |
| "grad_norm": 0.927262008190155, | |
| "learning_rate": 3.891972667484726e-06, | |
| "loss": 0.1092, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.1904761904761905, | |
| "grad_norm": 0.8392330408096313, | |
| "learning_rate": 3.859952044982329e-06, | |
| "loss": 0.1121, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.1944444444444444, | |
| "grad_norm": 0.8286350965499878, | |
| "learning_rate": 3.827980667861716e-06, | |
| "loss": 0.1141, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.1984126984126984, | |
| "grad_norm": 0.6824773550033569, | |
| "learning_rate": 3.7960599171548572e-06, | |
| "loss": 0.087, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.2023809523809523, | |
| "grad_norm": 1.077426791191101, | |
| "learning_rate": 3.764191171706867e-06, | |
| "loss": 0.159, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.2063492063492063, | |
| "grad_norm": 0.9567804932594299, | |
| "learning_rate": 3.732375808116451e-06, | |
| "loss": 0.1005, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.2103174603174602, | |
| "grad_norm": 0.8252041339874268, | |
| "learning_rate": 3.7006152006764336e-06, | |
| "loss": 0.0983, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.2142857142857142, | |
| "grad_norm": 0.9266362190246582, | |
| "learning_rate": 3.6689107213144025e-06, | |
| "loss": 0.1034, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.2182539682539684, | |
| "grad_norm": 0.7563996315002441, | |
| "learning_rate": 3.6372637395334416e-06, | |
| "loss": 0.087, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.2222222222222223, | |
| "grad_norm": 0.9262226223945618, | |
| "learning_rate": 3.6056756223529734e-06, | |
| "loss": 0.0929, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.2261904761904763, | |
| "grad_norm": 0.8781715035438538, | |
| "learning_rate": 3.574147734249719e-06, | |
| "loss": 0.0886, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.2301587301587302, | |
| "grad_norm": 0.7825202941894531, | |
| "learning_rate": 3.542681437098745e-06, | |
| "loss": 0.0726, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.2341269841269842, | |
| "grad_norm": 0.7199097275733948, | |
| "learning_rate": 3.5112780901146426e-06, | |
| "loss": 0.0662, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.2380952380952381, | |
| "grad_norm": 0.6553000211715698, | |
| "learning_rate": 3.479939049792817e-06, | |
| "loss": 0.0648, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.242063492063492, | |
| "grad_norm": 0.7398982644081116, | |
| "learning_rate": 3.448665669850888e-06, | |
| "loss": 0.0772, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.246031746031746, | |
| "grad_norm": 0.887241005897522, | |
| "learning_rate": 3.4174593011702197e-06, | |
| "loss": 0.1159, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.8772380948066711, | |
| "learning_rate": 3.386321291737563e-06, | |
| "loss": 0.1076, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.253968253968254, | |
| "grad_norm": 0.7155666947364807, | |
| "learning_rate": 3.3552529865868323e-06, | |
| "loss": 0.0788, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.257936507936508, | |
| "grad_norm": 0.6863790154457092, | |
| "learning_rate": 3.3242557277410015e-06, | |
| "loss": 0.0676, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.2619047619047619, | |
| "grad_norm": 0.6985348463058472, | |
| "learning_rate": 3.2933308541541365e-06, | |
| "loss": 0.0791, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.2658730158730158, | |
| "grad_norm": 0.7188520431518555, | |
| "learning_rate": 3.2624797016535626e-06, | |
| "loss": 0.0639, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.2698412698412698, | |
| "grad_norm": 0.6728897094726562, | |
| "learning_rate": 3.2317036028821523e-06, | |
| "loss": 0.0697, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.2738095238095237, | |
| "grad_norm": 0.7679276466369629, | |
| "learning_rate": 3.201003887240768e-06, | |
| "loss": 0.0753, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.2777777777777777, | |
| "grad_norm": 0.7593963146209717, | |
| "learning_rate": 3.1703818808308327e-06, | |
| "loss": 0.0621, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.2817460317460316, | |
| "grad_norm": 0.6922963261604309, | |
| "learning_rate": 3.1398389063970512e-06, | |
| "loss": 0.062, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 0.8319666385650635, | |
| "learning_rate": 3.1093762832702775e-06, | |
| "loss": 0.101, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.2896825396825398, | |
| "grad_norm": 0.9140247702598572, | |
| "learning_rate": 3.0789953273105123e-06, | |
| "loss": 0.1128, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.2936507936507937, | |
| "grad_norm": 0.8871818780899048, | |
| "learning_rate": 3.048697350850073e-06, | |
| "loss": 0.1334, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.2976190476190477, | |
| "grad_norm": 0.9085113406181335, | |
| "learning_rate": 3.0184836626369034e-06, | |
| "loss": 0.1145, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.3015873015873016, | |
| "grad_norm": 0.9396091103553772, | |
| "learning_rate": 2.988355567778043e-06, | |
| "loss": 0.1367, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.3055555555555556, | |
| "grad_norm": 0.9186078310012817, | |
| "learning_rate": 2.9583143676832526e-06, | |
| "loss": 0.1276, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.3095238095238095, | |
| "grad_norm": 0.940380871295929, | |
| "learning_rate": 2.9283613600087933e-06, | |
| "loss": 0.1178, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.3134920634920635, | |
| "grad_norm": 0.8242756128311157, | |
| "learning_rate": 2.8984978386013767e-06, | |
| "loss": 0.0972, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.3174603174603174, | |
| "grad_norm": 0.7357708811759949, | |
| "learning_rate": 2.8687250934422774e-06, | |
| "loss": 0.0873, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.3214285714285714, | |
| "grad_norm": 0.7667533755302429, | |
| "learning_rate": 2.839044410591606e-06, | |
| "loss": 0.0886, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.3253968253968254, | |
| "grad_norm": 0.85086590051651, | |
| "learning_rate": 2.809457072132766e-06, | |
| "loss": 0.0818, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.3293650793650793, | |
| "grad_norm": 0.6802100539207458, | |
| "learning_rate": 2.779964356117063e-06, | |
| "loss": 0.0681, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.8486253619194031, | |
| "learning_rate": 2.750567536508504e-06, | |
| "loss": 0.0751, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.3373015873015874, | |
| "grad_norm": 0.7886071801185608, | |
| "learning_rate": 2.7212678831287627e-06, | |
| "loss": 0.0856, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.3412698412698414, | |
| "grad_norm": 0.786064088344574, | |
| "learning_rate": 2.692066661602333e-06, | |
| "loss": 0.0777, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.3452380952380953, | |
| "grad_norm": 0.7525327801704407, | |
| "learning_rate": 2.662965133301862e-06, | |
| "loss": 0.0754, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.3492063492063493, | |
| "grad_norm": 0.6594064235687256, | |
| "learning_rate": 2.633964555293654e-06, | |
| "loss": 0.069, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.3531746031746033, | |
| "grad_norm": 0.6315282583236694, | |
| "learning_rate": 2.605066180283378e-06, | |
| "loss": 0.0611, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.3571428571428572, | |
| "grad_norm": 0.6245422959327698, | |
| "learning_rate": 2.576271256561953e-06, | |
| "loss": 0.0584, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.3611111111111112, | |
| "grad_norm": 0.8218834400177002, | |
| "learning_rate": 2.5475810279516287e-06, | |
| "loss": 0.0698, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.3650793650793651, | |
| "grad_norm": 0.6540297269821167, | |
| "learning_rate": 2.5189967337522574e-06, | |
| "loss": 0.0709, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.369047619047619, | |
| "grad_norm": 0.671125590801239, | |
| "learning_rate": 2.49051960868776e-06, | |
| "loss": 0.0599, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.373015873015873, | |
| "grad_norm": 0.6762871146202087, | |
| "learning_rate": 2.46215088285279e-06, | |
| "loss": 0.0612, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.376984126984127, | |
| "grad_norm": 0.7836350798606873, | |
| "learning_rate": 2.433891781659603e-06, | |
| "loss": 0.0691, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.380952380952381, | |
| "grad_norm": 0.8989992737770081, | |
| "learning_rate": 2.4057435257851173e-06, | |
| "loss": 0.122, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.3849206349206349, | |
| "grad_norm": 0.8189021944999695, | |
| "learning_rate": 2.377707331118196e-06, | |
| "loss": 0.1074, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.3888888888888888, | |
| "grad_norm": 0.6615262031555176, | |
| "learning_rate": 2.349784408707112e-06, | |
| "loss": 0.094, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.3928571428571428, | |
| "grad_norm": 0.7202811241149902, | |
| "learning_rate": 2.3219759647072467e-06, | |
| "loss": 0.0821, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.3968253968253967, | |
| "grad_norm": 0.7678143382072449, | |
| "learning_rate": 2.2942832003289823e-06, | |
| "loss": 0.0909, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.4007936507936507, | |
| "grad_norm": 0.8822202086448669, | |
| "learning_rate": 2.2667073117858185e-06, | |
| "loss": 0.1244, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.4047619047619047, | |
| "grad_norm": 0.8480525612831116, | |
| "learning_rate": 2.2392494902427027e-06, | |
| "loss": 0.1065, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.4087301587301586, | |
| "grad_norm": 0.842397928237915, | |
| "learning_rate": 2.2119109217645697e-06, | |
| "loss": 0.1057, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.4126984126984126, | |
| "grad_norm": 0.9880526065826416, | |
| "learning_rate": 2.1846927872651135e-06, | |
| "loss": 0.1402, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.4166666666666667, | |
| "grad_norm": 0.8727255463600159, | |
| "learning_rate": 2.1575962624557754e-06, | |
| "loss": 0.109, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.4206349206349207, | |
| "grad_norm": 0.9676539301872253, | |
| "learning_rate": 2.1306225177949584e-06, | |
| "loss": 0.1097, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.4246031746031746, | |
| "grad_norm": 0.7708781361579895, | |
| "learning_rate": 2.1037727184374705e-06, | |
| "loss": 0.0754, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.757243812084198, | |
| "learning_rate": 2.07704802418419e-06, | |
| "loss": 0.0776, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.4325396825396826, | |
| "grad_norm": 0.661636233329773, | |
| "learning_rate": 2.050449589431969e-06, | |
| "loss": 0.0597, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.4365079365079365, | |
| "grad_norm": 0.7882020473480225, | |
| "learning_rate": 2.023978563123771e-06, | |
| "loss": 0.0836, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.4404761904761905, | |
| "grad_norm": 0.7708131074905396, | |
| "learning_rate": 1.997636088699035e-06, | |
| "loss": 0.0796, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.4444444444444444, | |
| "grad_norm": 0.9235148429870605, | |
| "learning_rate": 1.9714233040442915e-06, | |
| "loss": 0.106, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.4484126984126984, | |
| "grad_norm": 0.8021928668022156, | |
| "learning_rate": 1.9453413414440043e-06, | |
| "loss": 0.0896, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.4523809523809523, | |
| "grad_norm": 0.7290391325950623, | |
| "learning_rate": 1.919391327531663e-06, | |
| "loss": 0.0754, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.4563492063492063, | |
| "grad_norm": 0.6653832197189331, | |
| "learning_rate": 1.8935743832411163e-06, | |
| "loss": 0.0705, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.4603174603174602, | |
| "grad_norm": 0.7387292385101318, | |
| "learning_rate": 1.8678916237581524e-06, | |
| "loss": 0.056, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.4642857142857144, | |
| "grad_norm": 0.6850692629814148, | |
| "learning_rate": 1.8423441584723312e-06, | |
| "loss": 0.0621, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.4682539682539684, | |
| "grad_norm": 0.8518023490905762, | |
| "learning_rate": 1.816933090929055e-06, | |
| "loss": 0.0686, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.4722222222222223, | |
| "grad_norm": 0.7155593037605286, | |
| "learning_rate": 1.791659518781908e-06, | |
| "loss": 0.0655, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.4761904761904763, | |
| "grad_norm": 0.628253698348999, | |
| "learning_rate": 1.7665245337452368e-06, | |
| "loss": 0.0548, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.4801587301587302, | |
| "grad_norm": 0.6528667211532593, | |
| "learning_rate": 1.7415292215469948e-06, | |
| "loss": 0.062, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.4841269841269842, | |
| "grad_norm": 0.7322276830673218, | |
| "learning_rate": 1.716674661881848e-06, | |
| "loss": 0.084, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.4880952380952381, | |
| "grad_norm": 0.9926005601882935, | |
| "learning_rate": 1.6919619283645262e-06, | |
| "loss": 0.1097, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.492063492063492, | |
| "grad_norm": 0.7675796747207642, | |
| "learning_rate": 1.667392088483456e-06, | |
| "loss": 0.0914, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.496031746031746, | |
| "grad_norm": 0.8206343054771423, | |
| "learning_rate": 1.6429662035546451e-06, | |
| "loss": 0.0865, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.840177059173584, | |
| "learning_rate": 1.6186853286758397e-06, | |
| "loss": 0.1106, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.503968253968254, | |
| "grad_norm": 0.9032308459281921, | |
| "learning_rate": 1.5945505126809524e-06, | |
| "loss": 0.1251, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.507936507936508, | |
| "grad_norm": 0.913895308971405, | |
| "learning_rate": 1.570562798094747e-06, | |
| "loss": 0.1124, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.5119047619047619, | |
| "grad_norm": 1.0361695289611816, | |
| "learning_rate": 1.5467232210878153e-06, | |
| "loss": 0.0855, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.5158730158730158, | |
| "grad_norm": 0.7832853198051453, | |
| "learning_rate": 1.5230328114318127e-06, | |
| "loss": 0.084, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.5198412698412698, | |
| "grad_norm": 0.8517787456512451, | |
| "learning_rate": 1.4994925924549797e-06, | |
| "loss": 0.0844, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.5238095238095237, | |
| "grad_norm": 0.6555785536766052, | |
| "learning_rate": 1.4761035809979395e-06, | |
| "loss": 0.0587, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.5277777777777777, | |
| "grad_norm": 0.839399516582489, | |
| "learning_rate": 1.452866787369771e-06, | |
| "loss": 0.0922, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.5317460317460316, | |
| "grad_norm": 0.7285085916519165, | |
| "learning_rate": 1.4297832153043657e-06, | |
| "loss": 0.066, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.5357142857142856, | |
| "grad_norm": 0.7503807544708252, | |
| "learning_rate": 1.4068538619170763e-06, | |
| "loss": 0.067, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.5396825396825395, | |
| "grad_norm": 0.791999876499176, | |
| "learning_rate": 1.3840797176616467e-06, | |
| "loss": 0.0873, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.5436507936507935, | |
| "grad_norm": 0.8322513699531555, | |
| "learning_rate": 1.3614617662874197e-06, | |
| "loss": 0.0763, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.5476190476190477, | |
| "grad_norm": 0.808028519153595, | |
| "learning_rate": 1.3390009847968505e-06, | |
| "loss": 0.0668, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.5515873015873016, | |
| "grad_norm": 0.5791826248168945, | |
| "learning_rate": 1.316698343403302e-06, | |
| "loss": 0.0519, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 0.6464975476264954, | |
| "learning_rate": 1.2945548054891322e-06, | |
| "loss": 0.0574, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.5595238095238095, | |
| "grad_norm": 0.7663049697875977, | |
| "learning_rate": 1.27257132756409e-06, | |
| "loss": 0.0596, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.5634920634920635, | |
| "grad_norm": 0.7549077868461609, | |
| "learning_rate": 1.2507488592239848e-06, | |
| "loss": 0.0739, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.5674603174603174, | |
| "grad_norm": 0.6650795936584473, | |
| "learning_rate": 1.2290883431096778e-06, | |
| "loss": 0.0546, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 0.6088014245033264, | |
| "learning_rate": 1.2075907148663579e-06, | |
| "loss": 0.0506, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.5753968253968254, | |
| "grad_norm": 0.6363900899887085, | |
| "learning_rate": 1.186256903103129e-06, | |
| "loss": 0.0494, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.5793650793650795, | |
| "grad_norm": 0.7967106103897095, | |
| "learning_rate": 1.1650878293528994e-06, | |
| "loss": 0.0814, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.5833333333333335, | |
| "grad_norm": 0.7806827425956726, | |
| "learning_rate": 1.1440844080325703e-06, | |
| "loss": 0.1009, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.5873015873015874, | |
| "grad_norm": 0.745284378528595, | |
| "learning_rate": 1.1232475464035386e-06, | |
| "loss": 0.0947, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.5912698412698414, | |
| "grad_norm": 0.6619460582733154, | |
| "learning_rate": 1.10257814453251e-06, | |
| "loss": 0.0675, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.5952380952380953, | |
| "grad_norm": 0.7792571187019348, | |
| "learning_rate": 1.0820770952526155e-06, | |
| "loss": 0.0977, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.5992063492063493, | |
| "grad_norm": 0.9165710806846619, | |
| "learning_rate": 1.0617452841248494e-06, | |
| "loss": 0.1407, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.6031746031746033, | |
| "grad_norm": 0.8994587659835815, | |
| "learning_rate": 1.0415835893998116e-06, | |
| "loss": 0.1076, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.6071428571428572, | |
| "grad_norm": 0.8812888264656067, | |
| "learning_rate": 1.0215928819797744e-06, | |
| "loss": 0.0916, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.6111111111111112, | |
| "grad_norm": 1.0087072849273682, | |
| "learning_rate": 1.0017740253810608e-06, | |
| "loss": 0.1402, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.6150793650793651, | |
| "grad_norm": 0.7847492694854736, | |
| "learning_rate": 9.821278756967467e-07, | |
| "loss": 0.0843, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.619047619047619, | |
| "grad_norm": 0.8777037262916565, | |
| "learning_rate": 9.62655281559679e-07, | |
| "loss": 0.1122, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.623015873015873, | |
| "grad_norm": 0.7136726379394531, | |
| "learning_rate": 9.433570841058187e-07, | |
| "loss": 0.0841, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.626984126984127, | |
| "grad_norm": 0.7613129615783691, | |
| "learning_rate": 9.242341169379077e-07, | |
| "loss": 0.0816, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.630952380952381, | |
| "grad_norm": 0.7026035785675049, | |
| "learning_rate": 9.052872060894613e-07, | |
| "loss": 0.0702, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.6349206349206349, | |
| "grad_norm": 0.7371145486831665, | |
| "learning_rate": 8.865171699890835e-07, | |
| "loss": 0.0677, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.6388888888888888, | |
| "grad_norm": 0.6632136702537537, | |
| "learning_rate": 8.679248194251211e-07, | |
| "loss": 0.062, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.6428571428571428, | |
| "grad_norm": 0.7706975340843201, | |
| "learning_rate": 8.495109575106331e-07, | |
| "loss": 0.0746, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.6468253968253967, | |
| "grad_norm": 0.9097256064414978, | |
| "learning_rate": 8.312763796487038e-07, | |
| "loss": 0.0869, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.6507936507936507, | |
| "grad_norm": 0.7003320455551147, | |
| "learning_rate": 8.132218734980852e-07, | |
| "loss": 0.0565, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.6547619047619047, | |
| "grad_norm": 0.6501448750495911, | |
| "learning_rate": 7.953482189391687e-07, | |
| "loss": 0.0582, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.6587301587301586, | |
| "grad_norm": 0.6474438309669495, | |
| "learning_rate": 7.776561880403072e-07, | |
| "loss": 0.0533, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.6626984126984126, | |
| "grad_norm": 0.7110742926597595, | |
| "learning_rate": 7.601465450244528e-07, | |
| "loss": 0.0672, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.6978676915168762, | |
| "learning_rate": 7.42820046236154e-07, | |
| "loss": 0.0716, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.6706349206349205, | |
| "grad_norm": 0.6054847836494446, | |
| "learning_rate": 7.256774401088817e-07, | |
| "loss": 0.0534, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.6746031746031746, | |
| "grad_norm": 0.7495798468589783, | |
| "learning_rate": 7.087194671326986e-07, | |
| "loss": 0.0634, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.6785714285714286, | |
| "grad_norm": 0.7438201308250427, | |
| "learning_rate": 6.91946859822279e-07, | |
| "loss": 0.0547, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.6825396825396826, | |
| "grad_norm": 0.9728838205337524, | |
| "learning_rate": 6.753603426852589e-07, | |
| "loss": 0.1373, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.6865079365079365, | |
| "grad_norm": 0.7290263772010803, | |
| "learning_rate": 6.589606321909464e-07, | |
| "loss": 0.0903, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.6904761904761905, | |
| "grad_norm": 0.7993935346603394, | |
| "learning_rate": 6.427484367393699e-07, | |
| "loss": 0.1113, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.6944444444444444, | |
| "grad_norm": 0.7429379820823669, | |
| "learning_rate": 6.267244566306801e-07, | |
| "loss": 0.0848, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.6984126984126984, | |
| "grad_norm": 1.0221623182296753, | |
| "learning_rate": 6.108893840348995e-07, | |
| "loss": 0.182, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.7023809523809523, | |
| "grad_norm": 0.9371434450149536, | |
| "learning_rate": 5.952439029620222e-07, | |
| "loss": 0.1308, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.7063492063492065, | |
| "grad_norm": 0.7311164736747742, | |
| "learning_rate": 5.797886892324695e-07, | |
| "loss": 0.083, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.7103174603174605, | |
| "grad_norm": 0.8895715475082397, | |
| "learning_rate": 5.645244104478947e-07, | |
| "loss": 0.1012, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 0.6606709361076355, | |
| "learning_rate": 5.494517259623478e-07, | |
| "loss": 0.0674, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.7182539682539684, | |
| "grad_norm": 0.8366999626159668, | |
| "learning_rate": 5.34571286853795e-07, | |
| "loss": 0.0976, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.7222222222222223, | |
| "grad_norm": 0.8094767928123474, | |
| "learning_rate": 5.198837358959901e-07, | |
| "loss": 0.0846, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.7261904761904763, | |
| "grad_norm": 0.6366130709648132, | |
| "learning_rate": 5.05389707530714e-07, | |
| "loss": 0.058, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.7301587301587302, | |
| "grad_norm": 0.76398766040802, | |
| "learning_rate": 4.91089827840367e-07, | |
| "loss": 0.0699, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.7341269841269842, | |
| "grad_norm": 0.6415200233459473, | |
| "learning_rate": 4.769847145209244e-07, | |
| "loss": 0.0541, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.7380952380952381, | |
| "grad_norm": 0.6345378756523132, | |
| "learning_rate": 4.6307497685525894e-07, | |
| "loss": 0.0586, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.742063492063492, | |
| "grad_norm": 0.8252426385879517, | |
| "learning_rate": 4.4936121568681546e-07, | |
| "loss": 0.0859, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.746031746031746, | |
| "grad_norm": 0.6913403868675232, | |
| "learning_rate": 4.3584402339366174e-07, | |
| "loss": 0.0674, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.7470701336860657, | |
| "learning_rate": 4.225239838628981e-07, | |
| "loss": 0.0798, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.753968253968254, | |
| "grad_norm": 0.6094357371330261, | |
| "learning_rate": 4.0940167246543595e-07, | |
| "loss": 0.0525, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.757936507936508, | |
| "grad_norm": 0.6257034540176392, | |
| "learning_rate": 3.964776560311484e-07, | |
| "loss": 0.0572, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.7619047619047619, | |
| "grad_norm": 0.6773633360862732, | |
| "learning_rate": 3.8375249282437743e-07, | |
| "loss": 0.0678, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.7658730158730158, | |
| "grad_norm": 0.7246127724647522, | |
| "learning_rate": 3.71226732519826e-07, | |
| "loss": 0.0642, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.7698412698412698, | |
| "grad_norm": 0.7381252646446228, | |
| "learning_rate": 3.589009161788104e-07, | |
| "loss": 0.0709, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.7738095238095237, | |
| "grad_norm": 0.5760139226913452, | |
| "learning_rate": 3.4677557622589175e-07, | |
| "loss": 0.0492, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.7324284911155701, | |
| "learning_rate": 3.3485123642587657e-07, | |
| "loss": 0.0566, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.7817460317460316, | |
| "grad_norm": 0.732704758644104, | |
| "learning_rate": 3.2312841186118937e-07, | |
| "loss": 0.0723, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "grad_norm": 0.809144914150238, | |
| "learning_rate": 3.116076089096265e-07, | |
| "loss": 0.0946, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.7896825396825395, | |
| "grad_norm": 0.6922594904899597, | |
| "learning_rate": 3.0028932522248256e-07, | |
| "loss": 0.087, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.7936507936507935, | |
| "grad_norm": 0.7200435400009155, | |
| "learning_rate": 2.8917404970305096e-07, | |
| "loss": 0.0778, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.7976190476190477, | |
| "grad_norm": 0.8525399565696716, | |
| "learning_rate": 2.782622624855097e-07, | |
| "loss": 0.1184, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.8015873015873016, | |
| "grad_norm": 1.011326551437378, | |
| "learning_rate": 2.6755443491417786e-07, | |
| "loss": 0.1315, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.8055555555555556, | |
| "grad_norm": 0.8276491761207581, | |
| "learning_rate": 2.570510295231571e-07, | |
| "loss": 0.0983, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.8095238095238095, | |
| "grad_norm": 0.8850067257881165, | |
| "learning_rate": 2.467525000163523e-07, | |
| "loss": 0.1102, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.8134920634920635, | |
| "grad_norm": 0.8555384278297424, | |
| "learning_rate": 2.36659291247871e-07, | |
| "loss": 0.105, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.8174603174603174, | |
| "grad_norm": 0.9438866376876831, | |
| "learning_rate": 2.2677183920281342e-07, | |
| "loss": 0.1173, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.8214285714285714, | |
| "grad_norm": 0.7671685814857483, | |
| "learning_rate": 2.1709057097843266e-07, | |
| "loss": 0.0834, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.8253968253968254, | |
| "grad_norm": 0.7308467626571655, | |
| "learning_rate": 2.0761590476568893e-07, | |
| "loss": 0.0605, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.8293650793650795, | |
| "grad_norm": 0.7060202360153198, | |
| "learning_rate": 1.9834824983118673e-07, | |
| "loss": 0.0623, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.8333333333333335, | |
| "grad_norm": 0.7693857550621033, | |
| "learning_rate": 1.892880064994934e-07, | |
| "loss": 0.0715, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.8373015873015874, | |
| "grad_norm": 0.67180997133255, | |
| "learning_rate": 1.8043556613585143e-07, | |
| "loss": 0.0491, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.8412698412698414, | |
| "grad_norm": 0.8116313815116882, | |
| "learning_rate": 1.7179131112926628e-07, | |
| "loss": 0.0835, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.8452380952380953, | |
| "grad_norm": 0.7374025583267212, | |
| "learning_rate": 1.6335561487599406e-07, | |
| "loss": 0.0739, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.8492063492063493, | |
| "grad_norm": 0.6444379091262817, | |
| "learning_rate": 1.551288417634106e-07, | |
| "loss": 0.0591, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.8531746031746033, | |
| "grad_norm": 0.5900991559028625, | |
| "learning_rate": 1.471113471542712e-07, | |
| "loss": 0.0504, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 0.6888481378555298, | |
| "learning_rate": 1.3930347737136195e-07, | |
| "loss": 0.0572, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.8611111111111112, | |
| "grad_norm": 0.7114200592041016, | |
| "learning_rate": 1.3170556968253756e-07, | |
| "loss": 0.0616, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.8650793650793651, | |
| "grad_norm": 0.5715036392211914, | |
| "learning_rate": 1.2431795228615372e-07, | |
| "loss": 0.049, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.869047619047619, | |
| "grad_norm": 0.6382952928543091, | |
| "learning_rate": 1.1714094429689127e-07, | |
| "loss": 0.0532, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.873015873015873, | |
| "grad_norm": 0.6427430510520935, | |
| "learning_rate": 1.1017485573197151e-07, | |
| "loss": 0.0431, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.876984126984127, | |
| "grad_norm": 0.6394485235214233, | |
| "learning_rate": 1.0341998749776316e-07, | |
| "loss": 0.0569, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.880952380952381, | |
| "grad_norm": 0.85865318775177, | |
| "learning_rate": 9.687663137678605e-08, | |
| "loss": 0.0979, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.8849206349206349, | |
| "grad_norm": 0.9239011406898499, | |
| "learning_rate": 9.054507001510727e-08, | |
| "loss": 0.1225, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.8888888888888888, | |
| "grad_norm": 0.6483728885650635, | |
| "learning_rate": 8.442557691013042e-08, | |
| "loss": 0.0858, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.8928571428571428, | |
| "grad_norm": 0.8771274089813232, | |
| "learning_rate": 7.851841639878399e-08, | |
| "loss": 0.0886, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.8968253968253967, | |
| "grad_norm": 0.9564152359962463, | |
| "learning_rate": 7.282384364610207e-08, | |
| "loss": 0.1339, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.9007936507936507, | |
| "grad_norm": 0.844241738319397, | |
| "learning_rate": 6.734210463420099e-08, | |
| "loss": 0.0913, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.9047619047619047, | |
| "grad_norm": 0.7983336448669434, | |
| "learning_rate": 6.207343615165562e-08, | |
| "loss": 0.0798, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.9087301587301586, | |
| "grad_norm": 0.8046018481254578, | |
| "learning_rate": 5.701806578327029e-08, | |
| "loss": 0.0886, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.9126984126984126, | |
| "grad_norm": 0.7213545441627502, | |
| "learning_rate": 5.21762119002478e-08, | |
| "loss": 0.0772, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.9166666666666665, | |
| "grad_norm": 0.7173414826393127, | |
| "learning_rate": 4.7548083650759134e-08, | |
| "loss": 0.0704, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.9206349206349205, | |
| "grad_norm": 0.7590553164482117, | |
| "learning_rate": 4.31338809509052e-08, | |
| "loss": 0.079, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.9246031746031746, | |
| "grad_norm": 0.6267789006233215, | |
| "learning_rate": 3.8933794476083143e-08, | |
| "loss": 0.0564, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.9285714285714286, | |
| "grad_norm": 0.7686551213264465, | |
| "learning_rate": 3.494800565275125e-08, | |
| "loss": 0.0543, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.9325396825396826, | |
| "grad_norm": 0.7288287281990051, | |
| "learning_rate": 3.1176686650589147e-08, | |
| "loss": 0.0645, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.9365079365079365, | |
| "grad_norm": 0.7328774333000183, | |
| "learning_rate": 2.7620000375064848e-08, | |
| "loss": 0.0726, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.9404761904761905, | |
| "grad_norm": 0.7171541452407837, | |
| "learning_rate": 2.4278100460393138e-08, | |
| "loss": 0.0578, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.9444444444444444, | |
| "grad_norm": 0.6567690968513489, | |
| "learning_rate": 2.115113126290258e-08, | |
| "loss": 0.0566, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.9484126984126984, | |
| "grad_norm": 0.707410991191864, | |
| "learning_rate": 1.8239227854799368e-08, | |
| "loss": 0.0596, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.9523809523809523, | |
| "grad_norm": 0.6027669906616211, | |
| "learning_rate": 1.554251601833201e-08, | |
| "loss": 0.0482, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.9563492063492065, | |
| "grad_norm": 0.6256717443466187, | |
| "learning_rate": 1.3061112240357887e-08, | |
| "loss": 0.0558, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 1.9603174603174605, | |
| "grad_norm": 0.5645001530647278, | |
| "learning_rate": 1.0795123707312283e-08, | |
| "loss": 0.0493, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.9642857142857144, | |
| "grad_norm": 0.6371707916259766, | |
| "learning_rate": 8.744648300578196e-09, | |
| "loss": 0.0547, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.9682539682539684, | |
| "grad_norm": 0.6413885354995728, | |
| "learning_rate": 6.9097745922580564e-09, | |
| "loss": 0.0572, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.9722222222222223, | |
| "grad_norm": 0.5743963122367859, | |
| "learning_rate": 5.2905818413478975e-09, | |
| "loss": 0.0529, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.9761904761904763, | |
| "grad_norm": 0.7577771544456482, | |
| "learning_rate": 3.887139990313427e-09, | |
| "loss": 0.0839, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.9801587301587302, | |
| "grad_norm": 0.8245152831077576, | |
| "learning_rate": 2.699509662069666e-09, | |
| "loss": 0.1112, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 1.9841269841269842, | |
| "grad_norm": 0.613452672958374, | |
| "learning_rate": 1.7277421573608234e-09, | |
| "loss": 0.0778, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.9880952380952381, | |
| "grad_norm": 0.6996891498565674, | |
| "learning_rate": 9.71879452545399e-10, | |
| "loss": 0.0881, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 1.992063492063492, | |
| "grad_norm": 0.8825079798698425, | |
| "learning_rate": 4.3195419778319095e-10, | |
| "loss": 0.1147, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.996031746031746, | |
| "grad_norm": 0.6807881593704224, | |
| "learning_rate": 1.0798971562364647e-10, | |
| "loss": 0.0602, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.7227473258972168, | |
| "learning_rate": 0.0, | |
| "loss": 0.0786, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 504, | |
| "total_flos": 78280651210752.0, | |
| "train_loss": 0.2315316873026036, | |
| "train_runtime": 2979.2646, | |
| "train_samples_per_second": 2.701, | |
| "train_steps_per_second": 0.169 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 504, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 78280651210752.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |