| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.16704, | |
| "eval_steps": 500, | |
| "global_step": 261, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00064, | |
| "grad_norm": 2.618795845193889, | |
| "learning_rate": 0.0, | |
| "loss": 0.5764, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00128, | |
| "grad_norm": 2.6347789365042167, | |
| "learning_rate": 6.329113924050633e-07, | |
| "loss": 0.6154, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00192, | |
| "grad_norm": 2.7588406940768695, | |
| "learning_rate": 1.2658227848101265e-06, | |
| "loss": 0.5763, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.00256, | |
| "grad_norm": 3.0430509991625927, | |
| "learning_rate": 1.8987341772151901e-06, | |
| "loss": 0.5897, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0032, | |
| "grad_norm": 3.0181149898389754, | |
| "learning_rate": 2.531645569620253e-06, | |
| "loss": 0.5849, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.00384, | |
| "grad_norm": 2.708156370622355, | |
| "learning_rate": 3.1645569620253167e-06, | |
| "loss": 0.6007, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.00448, | |
| "grad_norm": 2.2121700715533077, | |
| "learning_rate": 3.7974683544303802e-06, | |
| "loss": 0.5589, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.00512, | |
| "grad_norm": 2.5547904662956786, | |
| "learning_rate": 4.430379746835443e-06, | |
| "loss": 0.5868, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.00576, | |
| "grad_norm": 15.741913008515068, | |
| "learning_rate": 5.063291139240506e-06, | |
| "loss": 0.5496, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0064, | |
| "grad_norm": 1.4025681078646888, | |
| "learning_rate": 5.69620253164557e-06, | |
| "loss": 0.5335, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.00704, | |
| "grad_norm": 2.296825978922617, | |
| "learning_rate": 6.329113924050633e-06, | |
| "loss": 0.5614, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.00768, | |
| "grad_norm": 1.3790308520522885, | |
| "learning_rate": 6.9620253164556965e-06, | |
| "loss": 0.5155, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.00832, | |
| "grad_norm": 1.5578214066440281, | |
| "learning_rate": 7.5949367088607605e-06, | |
| "loss": 0.4714, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.00896, | |
| "grad_norm": 1.6513983171037458, | |
| "learning_rate": 8.227848101265822e-06, | |
| "loss": 0.4844, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0096, | |
| "grad_norm": 1.0659029951583614, | |
| "learning_rate": 8.860759493670886e-06, | |
| "loss": 0.4445, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.01024, | |
| "grad_norm": 1.5553885236728338, | |
| "learning_rate": 9.49367088607595e-06, | |
| "loss": 0.4296, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.01088, | |
| "grad_norm": 1.6423185898553931, | |
| "learning_rate": 1.0126582278481012e-05, | |
| "loss": 0.4558, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.01152, | |
| "grad_norm": 1.337112701272778, | |
| "learning_rate": 1.0759493670886076e-05, | |
| "loss": 0.4572, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.01216, | |
| "grad_norm": 1.091077579656484, | |
| "learning_rate": 1.139240506329114e-05, | |
| "loss": 0.4194, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 0.8415436037519902, | |
| "learning_rate": 1.2025316455696203e-05, | |
| "loss": 0.43, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01344, | |
| "grad_norm": 1.0942576082541409, | |
| "learning_rate": 1.2658227848101267e-05, | |
| "loss": 0.4342, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.01408, | |
| "grad_norm": 0.9266622560738933, | |
| "learning_rate": 1.3291139240506329e-05, | |
| "loss": 0.3692, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.01472, | |
| "grad_norm": 0.730124403683561, | |
| "learning_rate": 1.3924050632911393e-05, | |
| "loss": 0.3842, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.01536, | |
| "grad_norm": 1.0621346265160687, | |
| "learning_rate": 1.4556962025316457e-05, | |
| "loss": 0.3548, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.016, | |
| "grad_norm": 0.9227953803036282, | |
| "learning_rate": 1.5189873417721521e-05, | |
| "loss": 0.3948, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.01664, | |
| "grad_norm": 0.736727893263626, | |
| "learning_rate": 1.5822784810126583e-05, | |
| "loss": 0.375, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.01728, | |
| "grad_norm": 2.053133845104068, | |
| "learning_rate": 1.6455696202531644e-05, | |
| "loss": 0.3719, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.01792, | |
| "grad_norm": 0.6875688796188869, | |
| "learning_rate": 1.7088607594936708e-05, | |
| "loss": 0.3723, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.01856, | |
| "grad_norm": 0.6623745034402684, | |
| "learning_rate": 1.7721518987341772e-05, | |
| "loss": 0.3588, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 0.7263864736482054, | |
| "learning_rate": 1.8354430379746836e-05, | |
| "loss": 0.3851, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.01984, | |
| "grad_norm": 0.5571337797316878, | |
| "learning_rate": 1.89873417721519e-05, | |
| "loss": 0.3792, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.02048, | |
| "grad_norm": 0.9258693296754104, | |
| "learning_rate": 1.962025316455696e-05, | |
| "loss": 0.3852, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.02112, | |
| "grad_norm": 0.8990784229406635, | |
| "learning_rate": 2.0253164556962025e-05, | |
| "loss": 0.3505, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.02176, | |
| "grad_norm": 2.3509423854017837, | |
| "learning_rate": 2.088607594936709e-05, | |
| "loss": 0.364, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0224, | |
| "grad_norm": 2.6014985443321215, | |
| "learning_rate": 2.1518987341772153e-05, | |
| "loss": 0.3637, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.02304, | |
| "grad_norm": 0.7225091993923429, | |
| "learning_rate": 2.2151898734177217e-05, | |
| "loss": 0.3794, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.02368, | |
| "grad_norm": 0.6641449499251059, | |
| "learning_rate": 2.278481012658228e-05, | |
| "loss": 0.3699, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.02432, | |
| "grad_norm": 0.8028481090301588, | |
| "learning_rate": 2.341772151898734e-05, | |
| "loss": 0.3647, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.02496, | |
| "grad_norm": 0.6565265160500543, | |
| "learning_rate": 2.4050632911392405e-05, | |
| "loss": 0.3579, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 0.7322064983471163, | |
| "learning_rate": 2.468354430379747e-05, | |
| "loss": 0.3484, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.02624, | |
| "grad_norm": 0.5039022880384163, | |
| "learning_rate": 2.5316455696202533e-05, | |
| "loss": 0.3096, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.02688, | |
| "grad_norm": 0.5483302917388222, | |
| "learning_rate": 2.5949367088607597e-05, | |
| "loss": 0.3371, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.02752, | |
| "grad_norm": 1.3783780747410168, | |
| "learning_rate": 2.6582278481012658e-05, | |
| "loss": 0.3263, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.02816, | |
| "grad_norm": 0.5158426665517652, | |
| "learning_rate": 2.7215189873417722e-05, | |
| "loss": 0.3449, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.0288, | |
| "grad_norm": 0.5751703137103499, | |
| "learning_rate": 2.7848101265822786e-05, | |
| "loss": 0.3561, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.02944, | |
| "grad_norm": 0.6590394546695514, | |
| "learning_rate": 2.848101265822785e-05, | |
| "loss": 0.3478, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.03008, | |
| "grad_norm": 0.5214718720462116, | |
| "learning_rate": 2.9113924050632914e-05, | |
| "loss": 0.3408, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.03072, | |
| "grad_norm": 0.9091430743600112, | |
| "learning_rate": 2.9746835443037974e-05, | |
| "loss": 0.3526, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.03136, | |
| "grad_norm": 1.1052640742837447, | |
| "learning_rate": 3.0379746835443042e-05, | |
| "loss": 0.3346, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 0.48507046472864723, | |
| "learning_rate": 3.10126582278481e-05, | |
| "loss": 0.3443, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03264, | |
| "grad_norm": 0.5388534097509862, | |
| "learning_rate": 3.1645569620253167e-05, | |
| "loss": 0.3312, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.03328, | |
| "grad_norm": 0.5894091981022138, | |
| "learning_rate": 3.227848101265823e-05, | |
| "loss": 0.3317, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.03392, | |
| "grad_norm": 0.4751863643628404, | |
| "learning_rate": 3.291139240506329e-05, | |
| "loss": 0.2952, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.03456, | |
| "grad_norm": 0.5790137817312099, | |
| "learning_rate": 3.354430379746836e-05, | |
| "loss": 0.3392, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.0352, | |
| "grad_norm": 1.5087120760365582, | |
| "learning_rate": 3.4177215189873416e-05, | |
| "loss": 0.3339, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.03584, | |
| "grad_norm": 0.5745952365097137, | |
| "learning_rate": 3.4810126582278487e-05, | |
| "loss": 0.3347, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.03648, | |
| "grad_norm": 0.7517935525029015, | |
| "learning_rate": 3.5443037974683544e-05, | |
| "loss": 0.3324, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.03712, | |
| "grad_norm": 0.5324903887679082, | |
| "learning_rate": 3.607594936708861e-05, | |
| "loss": 0.3522, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.03776, | |
| "grad_norm": 0.5095112678612309, | |
| "learning_rate": 3.670886075949367e-05, | |
| "loss": 0.3394, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 0.5442974811657504, | |
| "learning_rate": 3.7341772151898736e-05, | |
| "loss": 0.3322, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03904, | |
| "grad_norm": 0.5228282028496956, | |
| "learning_rate": 3.79746835443038e-05, | |
| "loss": 0.34, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.03968, | |
| "grad_norm": 0.5234005807075949, | |
| "learning_rate": 3.8607594936708864e-05, | |
| "loss": 0.3205, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.04032, | |
| "grad_norm": 0.5456894801995082, | |
| "learning_rate": 3.924050632911392e-05, | |
| "loss": 0.3264, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.04096, | |
| "grad_norm": 0.6064859304760205, | |
| "learning_rate": 3.987341772151899e-05, | |
| "loss": 0.3603, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.0416, | |
| "grad_norm": 0.4817896106905655, | |
| "learning_rate": 4.050632911392405e-05, | |
| "loss": 0.3159, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.04224, | |
| "grad_norm": 0.5788597012081875, | |
| "learning_rate": 4.113924050632912e-05, | |
| "loss": 0.3509, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.04288, | |
| "grad_norm": 0.5412218771204261, | |
| "learning_rate": 4.177215189873418e-05, | |
| "loss": 0.3208, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.04352, | |
| "grad_norm": 1.3974653981931218, | |
| "learning_rate": 4.240506329113924e-05, | |
| "loss": 0.3319, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.04416, | |
| "grad_norm": 0.5446574915363599, | |
| "learning_rate": 4.3037974683544305e-05, | |
| "loss": 0.2997, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.0448, | |
| "grad_norm": 0.4568551671818489, | |
| "learning_rate": 4.367088607594937e-05, | |
| "loss": 0.3361, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.04544, | |
| "grad_norm": 0.5949709296262449, | |
| "learning_rate": 4.430379746835443e-05, | |
| "loss": 0.3211, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.04608, | |
| "grad_norm": 0.5482381035251598, | |
| "learning_rate": 4.49367088607595e-05, | |
| "loss": 0.3292, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.04672, | |
| "grad_norm": 0.4997143028932898, | |
| "learning_rate": 4.556962025316456e-05, | |
| "loss": 0.318, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.04736, | |
| "grad_norm": 0.5841404500165603, | |
| "learning_rate": 4.6202531645569625e-05, | |
| "loss": 0.3138, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 0.5028458152612025, | |
| "learning_rate": 4.683544303797468e-05, | |
| "loss": 0.3029, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.04864, | |
| "grad_norm": 0.48112579785481197, | |
| "learning_rate": 4.7468354430379746e-05, | |
| "loss": 0.3243, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.04928, | |
| "grad_norm": 1.0013986736975447, | |
| "learning_rate": 4.810126582278481e-05, | |
| "loss": 0.3308, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.04992, | |
| "grad_norm": 0.4247544379828167, | |
| "learning_rate": 4.8734177215189874e-05, | |
| "loss": 0.2958, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.05056, | |
| "grad_norm": 0.5421182493231368, | |
| "learning_rate": 4.936708860759494e-05, | |
| "loss": 0.3297, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 1.2844161669496286, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3198, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.05184, | |
| "grad_norm": 0.5627454560158847, | |
| "learning_rate": 4.999994390459423e-05, | |
| "loss": 0.3311, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.05248, | |
| "grad_norm": 0.4505373699775162, | |
| "learning_rate": 4.999977561862863e-05, | |
| "loss": 0.3227, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.05312, | |
| "grad_norm": 0.5349576256311448, | |
| "learning_rate": 4.999949514285842e-05, | |
| "loss": 0.323, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.05376, | |
| "grad_norm": 0.5591650619223139, | |
| "learning_rate": 4.999910247854227e-05, | |
| "loss": 0.336, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.0544, | |
| "grad_norm": 0.6827279250952998, | |
| "learning_rate": 4.99985976274423e-05, | |
| "loss": 0.3319, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.05504, | |
| "grad_norm": 0.7102827110283111, | |
| "learning_rate": 4.9997980591824126e-05, | |
| "loss": 0.3137, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.05568, | |
| "grad_norm": 0.5941384980560921, | |
| "learning_rate": 4.999725137445674e-05, | |
| "loss": 0.3242, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.05632, | |
| "grad_norm": 0.5293400414276046, | |
| "learning_rate": 4.999640997861264e-05, | |
| "loss": 0.3266, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.05696, | |
| "grad_norm": 0.6396725464059344, | |
| "learning_rate": 4.999545640806766e-05, | |
| "loss": 0.3297, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 0.47264233800825245, | |
| "learning_rate": 4.999439066710111e-05, | |
| "loss": 0.347, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.05824, | |
| "grad_norm": 0.5476181291945234, | |
| "learning_rate": 4.999321276049563e-05, | |
| "loss": 0.3222, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.05888, | |
| "grad_norm": 0.5706613009258329, | |
| "learning_rate": 4.9991922693537226e-05, | |
| "loss": 0.3122, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.05952, | |
| "grad_norm": 0.48706789610757445, | |
| "learning_rate": 4.9990520472015246e-05, | |
| "loss": 0.2993, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.06016, | |
| "grad_norm": 0.4115404611758198, | |
| "learning_rate": 4.998900610222236e-05, | |
| "loss": 0.3099, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.0608, | |
| "grad_norm": 0.532905037021598, | |
| "learning_rate": 4.998737959095449e-05, | |
| "loss": 0.2827, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.06144, | |
| "grad_norm": 0.4643257281999667, | |
| "learning_rate": 4.998564094551081e-05, | |
| "loss": 0.3205, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.06208, | |
| "grad_norm": 0.5191147113172441, | |
| "learning_rate": 4.998379017369375e-05, | |
| "loss": 0.3144, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.06272, | |
| "grad_norm": 0.5348016711569842, | |
| "learning_rate": 4.9981827283808855e-05, | |
| "loss": 0.3128, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.06336, | |
| "grad_norm": 0.5010085634276595, | |
| "learning_rate": 4.997975228466489e-05, | |
| "loss": 0.3404, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 0.4889491939114262, | |
| "learning_rate": 4.997756518557368e-05, | |
| "loss": 0.2828, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06464, | |
| "grad_norm": 0.46482892338474413, | |
| "learning_rate": 4.997526599635011e-05, | |
| "loss": 0.2892, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.06528, | |
| "grad_norm": 0.5287297970963423, | |
| "learning_rate": 4.997285472731211e-05, | |
| "loss": 0.3397, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.06592, | |
| "grad_norm": 0.46498628733863295, | |
| "learning_rate": 4.997033138928056e-05, | |
| "loss": 0.3187, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.06656, | |
| "grad_norm": 0.5394842738162251, | |
| "learning_rate": 4.996769599357927e-05, | |
| "loss": 0.3136, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.0672, | |
| "grad_norm": 0.5304962827503229, | |
| "learning_rate": 4.996494855203493e-05, | |
| "loss": 0.3286, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.06784, | |
| "grad_norm": 0.5626109543739455, | |
| "learning_rate": 4.996208907697706e-05, | |
| "loss": 0.3256, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.06848, | |
| "grad_norm": 0.47113589346903195, | |
| "learning_rate": 4.995911758123791e-05, | |
| "loss": 0.287, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.06912, | |
| "grad_norm": 0.5351822398424532, | |
| "learning_rate": 4.995603407815249e-05, | |
| "loss": 0.2996, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.06976, | |
| "grad_norm": 0.46338886806198026, | |
| "learning_rate": 4.9952838581558406e-05, | |
| "loss": 0.3056, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.0704, | |
| "grad_norm": 0.5172304018785423, | |
| "learning_rate": 4.994953110579588e-05, | |
| "loss": 0.3192, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.07104, | |
| "grad_norm": 0.510154274412123, | |
| "learning_rate": 4.9946111665707645e-05, | |
| "loss": 0.3307, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.07168, | |
| "grad_norm": 0.3838546254703525, | |
| "learning_rate": 4.99425802766389e-05, | |
| "loss": 0.2929, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.07232, | |
| "grad_norm": 0.5449189029345786, | |
| "learning_rate": 4.993893695443721e-05, | |
| "loss": 0.3069, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.07296, | |
| "grad_norm": 0.44577057887896443, | |
| "learning_rate": 4.993518171545247e-05, | |
| "loss": 0.2966, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.0736, | |
| "grad_norm": 0.4515440453670323, | |
| "learning_rate": 4.993131457653682e-05, | |
| "loss": 0.3061, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.07424, | |
| "grad_norm": 0.44391881310796194, | |
| "learning_rate": 4.992733555504454e-05, | |
| "loss": 0.3215, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.07488, | |
| "grad_norm": 0.42638885395980747, | |
| "learning_rate": 4.9923244668832025e-05, | |
| "loss": 0.3111, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.07552, | |
| "grad_norm": 0.44150879225613077, | |
| "learning_rate": 4.991904193625768e-05, | |
| "loss": 0.316, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.07616, | |
| "grad_norm": 0.4431237901584868, | |
| "learning_rate": 4.991472737618181e-05, | |
| "loss": 0.2776, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 0.42121400061187486, | |
| "learning_rate": 4.991030100796659e-05, | |
| "loss": 0.3064, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.07744, | |
| "grad_norm": 0.392309441081599, | |
| "learning_rate": 4.9905762851475915e-05, | |
| "loss": 0.3209, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.07808, | |
| "grad_norm": 0.45305290693167705, | |
| "learning_rate": 4.990111292707537e-05, | |
| "loss": 0.3183, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.07872, | |
| "grad_norm": 0.4293788826264019, | |
| "learning_rate": 4.989635125563211e-05, | |
| "loss": 0.2924, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.07936, | |
| "grad_norm": 0.4272809824845029, | |
| "learning_rate": 4.989147785851477e-05, | |
| "loss": 0.3137, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.417177718336002, | |
| "learning_rate": 4.988649275759335e-05, | |
| "loss": 0.323, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.08064, | |
| "grad_norm": 0.6352795142091163, | |
| "learning_rate": 4.988139597523916e-05, | |
| "loss": 0.2968, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.08128, | |
| "grad_norm": 0.4311613332706907, | |
| "learning_rate": 4.98761875343247e-05, | |
| "loss": 0.2864, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.08192, | |
| "grad_norm": 0.407677725063968, | |
| "learning_rate": 4.987086745822352e-05, | |
| "loss": 0.2887, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.08256, | |
| "grad_norm": 0.43620593917226175, | |
| "learning_rate": 4.986543577081017e-05, | |
| "loss": 0.326, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.0832, | |
| "grad_norm": 0.3577110758670567, | |
| "learning_rate": 4.985989249646007e-05, | |
| "loss": 0.3049, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.08384, | |
| "grad_norm": 0.40530215341984516, | |
| "learning_rate": 4.985423766004939e-05, | |
| "loss": 0.3139, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.08448, | |
| "grad_norm": 0.46328257103904175, | |
| "learning_rate": 4.9848471286954965e-05, | |
| "loss": 0.3306, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.08512, | |
| "grad_norm": 0.38436097877243486, | |
| "learning_rate": 4.984259340305416e-05, | |
| "loss": 0.2931, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.08576, | |
| "grad_norm": 0.6093078034059375, | |
| "learning_rate": 4.9836604034724753e-05, | |
| "loss": 0.2963, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.0864, | |
| "grad_norm": 0.46772762171913185, | |
| "learning_rate": 4.9830503208844835e-05, | |
| "loss": 0.2979, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.08704, | |
| "grad_norm": 0.48325714173598944, | |
| "learning_rate": 4.982429095279266e-05, | |
| "loss": 0.2984, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.08768, | |
| "grad_norm": 0.5477120347200936, | |
| "learning_rate": 4.981796729444655e-05, | |
| "loss": 0.3302, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.08832, | |
| "grad_norm": 0.534004632135256, | |
| "learning_rate": 4.981153226218477e-05, | |
| "loss": 0.3136, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.08896, | |
| "grad_norm": 0.4441445163984143, | |
| "learning_rate": 4.9804985884885366e-05, | |
| "loss": 0.2911, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 0.5818680485265652, | |
| "learning_rate": 4.979832819192608e-05, | |
| "loss": 0.3178, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.09024, | |
| "grad_norm": 0.5596729609258517, | |
| "learning_rate": 4.97915592131842e-05, | |
| "loss": 0.2946, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.09088, | |
| "grad_norm": 0.4173115723963595, | |
| "learning_rate": 4.97846789790364e-05, | |
| "loss": 0.312, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.09152, | |
| "grad_norm": 0.5797089756431767, | |
| "learning_rate": 4.977768752035865e-05, | |
| "loss": 0.317, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.09216, | |
| "grad_norm": 0.599992651207998, | |
| "learning_rate": 4.9770584868526046e-05, | |
| "loss": 0.3162, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.0928, | |
| "grad_norm": 0.5107814027610112, | |
| "learning_rate": 4.9763371055412685e-05, | |
| "loss": 0.313, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.09344, | |
| "grad_norm": 0.44298297975068884, | |
| "learning_rate": 4.9756046113391496e-05, | |
| "loss": 0.3351, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.09408, | |
| "grad_norm": 0.5327895632430579, | |
| "learning_rate": 4.974861007533413e-05, | |
| "loss": 0.2917, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.09472, | |
| "grad_norm": 0.43312248711811524, | |
| "learning_rate": 4.97410629746108e-05, | |
| "loss": 0.2964, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.09536, | |
| "grad_norm": 0.5304322484911422, | |
| "learning_rate": 4.9733404845090116e-05, | |
| "loss": 0.2898, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 0.3997434435275148, | |
| "learning_rate": 4.972563572113894e-05, | |
| "loss": 0.2941, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.09664, | |
| "grad_norm": 0.6731310579298981, | |
| "learning_rate": 4.971775563762226e-05, | |
| "loss": 0.2979, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.09728, | |
| "grad_norm": 0.44982760512871145, | |
| "learning_rate": 4.970976462990299e-05, | |
| "loss": 0.3181, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.09792, | |
| "grad_norm": 0.582429222704454, | |
| "learning_rate": 4.970166273384183e-05, | |
| "loss": 0.2944, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.09856, | |
| "grad_norm": 0.37285441290932575, | |
| "learning_rate": 4.969344998579712e-05, | |
| "loss": 0.2708, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.0992, | |
| "grad_norm": 0.5242244051320633, | |
| "learning_rate": 4.9685126422624644e-05, | |
| "loss": 0.322, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.09984, | |
| "grad_norm": 0.42399899648411177, | |
| "learning_rate": 4.96766920816775e-05, | |
| "loss": 0.3194, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.10048, | |
| "grad_norm": 0.4451557052439993, | |
| "learning_rate": 4.966814700080592e-05, | |
| "loss": 0.3242, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.10112, | |
| "grad_norm": 0.48601797793339585, | |
| "learning_rate": 4.965949121835707e-05, | |
| "loss": 0.3397, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.10176, | |
| "grad_norm": 0.44145425649965603, | |
| "learning_rate": 4.965072477317493e-05, | |
| "loss": 0.3146, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 0.4370780933539978, | |
| "learning_rate": 4.964184770460007e-05, | |
| "loss": 0.2978, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.10304, | |
| "grad_norm": 0.48836377923781304, | |
| "learning_rate": 4.963286005246953e-05, | |
| "loss": 0.2884, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.10368, | |
| "grad_norm": 0.41359620894086957, | |
| "learning_rate": 4.9623761857116594e-05, | |
| "loss": 0.3253, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.10432, | |
| "grad_norm": 0.44338475984615433, | |
| "learning_rate": 4.961455315937059e-05, | |
| "loss": 0.2871, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.10496, | |
| "grad_norm": 0.5079738340888645, | |
| "learning_rate": 4.960523400055679e-05, | |
| "loss": 0.3002, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.1056, | |
| "grad_norm": 0.427097830248176, | |
| "learning_rate": 4.959580442249615e-05, | |
| "loss": 0.3068, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.10624, | |
| "grad_norm": 0.3684291109883667, | |
| "learning_rate": 4.958626446750515e-05, | |
| "loss": 0.3256, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.10688, | |
| "grad_norm": 0.4683572393673182, | |
| "learning_rate": 4.9576614178395595e-05, | |
| "loss": 0.3054, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.10752, | |
| "grad_norm": 0.38807804876427243, | |
| "learning_rate": 4.956685359847445e-05, | |
| "loss": 0.3018, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.10816, | |
| "grad_norm": 0.387073605411047, | |
| "learning_rate": 4.95569827715436e-05, | |
| "loss": 0.2869, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.1088, | |
| "grad_norm": 0.4547710232799118, | |
| "learning_rate": 4.9547001741899693e-05, | |
| "loss": 0.3036, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.10944, | |
| "grad_norm": 0.39381908155851314, | |
| "learning_rate": 4.953691055433391e-05, | |
| "loss": 0.2966, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.11008, | |
| "grad_norm": 0.48725912333872035, | |
| "learning_rate": 4.952670925413182e-05, | |
| "loss": 0.3093, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.11072, | |
| "grad_norm": 0.3989019148895901, | |
| "learning_rate": 4.951639788707307e-05, | |
| "loss": 0.289, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.11136, | |
| "grad_norm": 0.3836802824526995, | |
| "learning_rate": 4.950597649943132e-05, | |
| "loss": 0.284, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.112, | |
| "grad_norm": 0.43375030098630124, | |
| "learning_rate": 4.9495445137973905e-05, | |
| "loss": 0.2993, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.11264, | |
| "grad_norm": 0.38353724590203925, | |
| "learning_rate": 4.948480384996172e-05, | |
| "loss": 0.2928, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.11328, | |
| "grad_norm": 0.41856497407127785, | |
| "learning_rate": 4.947405268314894e-05, | |
| "loss": 0.3058, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.11392, | |
| "grad_norm": 1.22965703788482, | |
| "learning_rate": 4.9463191685782865e-05, | |
| "loss": 0.3135, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.11456, | |
| "grad_norm": 0.3607079092136828, | |
| "learning_rate": 4.945222090660364e-05, | |
| "loss": 0.2911, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 0.42652787065228764, | |
| "learning_rate": 4.944114039484411e-05, | |
| "loss": 0.3105, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.11584, | |
| "grad_norm": 0.39223235556161457, | |
| "learning_rate": 4.942995020022954e-05, | |
| "loss": 0.3308, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.11648, | |
| "grad_norm": 0.4434271422328715, | |
| "learning_rate": 4.9418650372977384e-05, | |
| "loss": 0.2946, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.11712, | |
| "grad_norm": 0.3871233188728082, | |
| "learning_rate": 4.940724096379714e-05, | |
| "loss": 0.2908, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.11776, | |
| "grad_norm": 0.3504967706065276, | |
| "learning_rate": 4.9395722023890035e-05, | |
| "loss": 0.2869, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.1184, | |
| "grad_norm": 0.3800847472179798, | |
| "learning_rate": 4.938409360494883e-05, | |
| "loss": 0.2916, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.11904, | |
| "grad_norm": 0.35973303790543976, | |
| "learning_rate": 4.937235575915761e-05, | |
| "loss": 0.2994, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.11968, | |
| "grad_norm": 0.39040484393002967, | |
| "learning_rate": 4.93605085391915e-05, | |
| "loss": 0.2994, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.12032, | |
| "grad_norm": 0.45370992844530267, | |
| "learning_rate": 4.934855199821647e-05, | |
| "loss": 0.2888, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.12096, | |
| "grad_norm": 0.41032353843291086, | |
| "learning_rate": 4.93364861898891e-05, | |
| "loss": 0.2845, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.1216, | |
| "grad_norm": 0.4036630852460455, | |
| "learning_rate": 4.932431116835628e-05, | |
| "loss": 0.3052, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.12224, | |
| "grad_norm": 0.3484768322428675, | |
| "learning_rate": 4.931202698825503e-05, | |
| "loss": 0.2896, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.12288, | |
| "grad_norm": 0.34569927077349094, | |
| "learning_rate": 4.929963370471226e-05, | |
| "loss": 0.2953, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.12352, | |
| "grad_norm": 0.3321813143876421, | |
| "learning_rate": 4.928713137334445e-05, | |
| "loss": 0.2891, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.12416, | |
| "grad_norm": 0.3696723739246585, | |
| "learning_rate": 4.927452005025748e-05, | |
| "loss": 0.3091, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.1248, | |
| "grad_norm": 0.3525045846275956, | |
| "learning_rate": 4.926179979204633e-05, | |
| "loss": 0.3089, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.12544, | |
| "grad_norm": 0.4050534156570688, | |
| "learning_rate": 4.9248970655794836e-05, | |
| "loss": 0.2991, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.12608, | |
| "grad_norm": 0.37675234071249, | |
| "learning_rate": 4.923603269907545e-05, | |
| "loss": 0.2798, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.12672, | |
| "grad_norm": 0.34914495162279696, | |
| "learning_rate": 4.922298597994897e-05, | |
| "loss": 0.289, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.12736, | |
| "grad_norm": 0.3865497720329682, | |
| "learning_rate": 4.920983055696428e-05, | |
| "loss": 0.3002, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 0.446766451366359, | |
| "learning_rate": 4.919656648915807e-05, | |
| "loss": 0.2918, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.12864, | |
| "grad_norm": 0.473700352021563, | |
| "learning_rate": 4.918319383605462e-05, | |
| "loss": 0.2639, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.12928, | |
| "grad_norm": 0.4098673261200793, | |
| "learning_rate": 4.916971265766547e-05, | |
| "loss": 0.3025, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.12992, | |
| "grad_norm": 0.47472542511654414, | |
| "learning_rate": 4.915612301448919e-05, | |
| "loss": 0.3039, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.13056, | |
| "grad_norm": 0.4474522945723034, | |
| "learning_rate": 4.914242496751111e-05, | |
| "loss": 0.2962, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.1312, | |
| "grad_norm": 0.5376434925769887, | |
| "learning_rate": 4.912861857820303e-05, | |
| "loss": 0.3125, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.13184, | |
| "grad_norm": 0.415507945261084, | |
| "learning_rate": 4.911470390852295e-05, | |
| "loss": 0.3086, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.13248, | |
| "grad_norm": 0.4284511552336465, | |
| "learning_rate": 4.91006810209148e-05, | |
| "loss": 0.2952, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.13312, | |
| "grad_norm": 0.40052269427223414, | |
| "learning_rate": 4.9086549978308125e-05, | |
| "loss": 0.3062, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.13376, | |
| "grad_norm": 0.43743782442362256, | |
| "learning_rate": 4.9072310844117875e-05, | |
| "loss": 0.2794, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 0.4420765691985995, | |
| "learning_rate": 4.9057963682244045e-05, | |
| "loss": 0.298, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.13504, | |
| "grad_norm": 0.41724801069829737, | |
| "learning_rate": 4.904350855707142e-05, | |
| "loss": 0.2936, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.13568, | |
| "grad_norm": 0.3775542589408398, | |
| "learning_rate": 4.9028945533469276e-05, | |
| "loss": 0.2714, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.13632, | |
| "grad_norm": 0.4493944075287162, | |
| "learning_rate": 4.9014274676791134e-05, | |
| "loss": 0.3094, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.13696, | |
| "grad_norm": 0.4671949905514448, | |
| "learning_rate": 4.899949605287439e-05, | |
| "loss": 0.2883, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.1376, | |
| "grad_norm": 0.39187019356209574, | |
| "learning_rate": 4.898460972804009e-05, | |
| "loss": 0.3155, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.13824, | |
| "grad_norm": 0.3832421971309516, | |
| "learning_rate": 4.896961576909258e-05, | |
| "loss": 0.2891, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.13888, | |
| "grad_norm": 0.4701725265012741, | |
| "learning_rate": 4.895451424331924e-05, | |
| "loss": 0.2791, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.13952, | |
| "grad_norm": 0.3906675874850412, | |
| "learning_rate": 4.893930521849016e-05, | |
| "loss": 0.2739, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.14016, | |
| "grad_norm": 0.5048426982814626, | |
| "learning_rate": 4.892398876285785e-05, | |
| "loss": 0.2956, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 0.5550796911808695, | |
| "learning_rate": 4.890856494515696e-05, | |
| "loss": 0.2942, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.14144, | |
| "grad_norm": 0.4446478533142643, | |
| "learning_rate": 4.889303383460389e-05, | |
| "loss": 0.2876, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.14208, | |
| "grad_norm": 0.5139145715219713, | |
| "learning_rate": 4.887739550089656e-05, | |
| "loss": 0.2871, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.14272, | |
| "grad_norm": 0.433835044119607, | |
| "learning_rate": 4.886165001421408e-05, | |
| "loss": 0.3174, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.14336, | |
| "grad_norm": 0.4118433589806149, | |
| "learning_rate": 4.884579744521638e-05, | |
| "loss": 0.2915, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 0.36660272368355734, | |
| "learning_rate": 4.8829837865043984e-05, | |
| "loss": 0.2984, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.14464, | |
| "grad_norm": 0.37584693808307384, | |
| "learning_rate": 4.881377134531762e-05, | |
| "loss": 0.2738, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.14528, | |
| "grad_norm": 0.39658204298189026, | |
| "learning_rate": 4.8797597958137914e-05, | |
| "loss": 0.2868, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.14592, | |
| "grad_norm": 0.42820275509861444, | |
| "learning_rate": 4.878131777608509e-05, | |
| "loss": 0.2921, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.14656, | |
| "grad_norm": 0.4162784808158851, | |
| "learning_rate": 4.876493087221862e-05, | |
| "loss": 0.3034, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.1472, | |
| "grad_norm": 0.4050308180549546, | |
| "learning_rate": 4.87484373200769e-05, | |
| "loss": 0.2964, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.14784, | |
| "grad_norm": 0.396192734792577, | |
| "learning_rate": 4.8731837193676944e-05, | |
| "loss": 0.2879, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.14848, | |
| "grad_norm": 0.4022225227959102, | |
| "learning_rate": 4.871513056751401e-05, | |
| "loss": 0.3058, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.14912, | |
| "grad_norm": 0.4712047740803004, | |
| "learning_rate": 4.8698317516561295e-05, | |
| "loss": 0.307, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.14976, | |
| "grad_norm": 0.5020568589020615, | |
| "learning_rate": 4.8681398116269596e-05, | |
| "loss": 0.2937, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.1504, | |
| "grad_norm": 0.5384324202189518, | |
| "learning_rate": 4.8664372442566954e-05, | |
| "loss": 0.3024, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.15104, | |
| "grad_norm": 0.39516572600823124, | |
| "learning_rate": 4.864724057185834e-05, | |
| "loss": 0.2743, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.15168, | |
| "grad_norm": 0.4548277706908328, | |
| "learning_rate": 4.86300025810253e-05, | |
| "loss": 0.2877, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.15232, | |
| "grad_norm": 0.3720089945799359, | |
| "learning_rate": 4.861265854742559e-05, | |
| "loss": 0.2819, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.15296, | |
| "grad_norm": 0.47047832876789597, | |
| "learning_rate": 4.859520854889287e-05, | |
| "loss": 0.3021, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 0.4554923160199729, | |
| "learning_rate": 4.857765266373631e-05, | |
| "loss": 0.3033, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.15424, | |
| "grad_norm": 0.5479982370735748, | |
| "learning_rate": 4.855999097074027e-05, | |
| "loss": 0.3016, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.15488, | |
| "grad_norm": 0.38490547613491893, | |
| "learning_rate": 4.8542223549163946e-05, | |
| "loss": 0.2791, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.15552, | |
| "grad_norm": 0.5010326131588188, | |
| "learning_rate": 4.8524350478740985e-05, | |
| "loss": 0.2758, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.15616, | |
| "grad_norm": 0.408628306302986, | |
| "learning_rate": 4.850637183967917e-05, | |
| "loss": 0.2914, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.1568, | |
| "grad_norm": 0.4342781570295902, | |
| "learning_rate": 4.848828771266002e-05, | |
| "loss": 0.2977, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.15744, | |
| "grad_norm": 0.43152902680446176, | |
| "learning_rate": 4.847009817883844e-05, | |
| "loss": 0.2909, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.15808, | |
| "grad_norm": 0.3551063718621289, | |
| "learning_rate": 4.845180331984239e-05, | |
| "loss": 0.2788, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.15872, | |
| "grad_norm": 0.4023732387829879, | |
| "learning_rate": 4.843340321777246e-05, | |
| "loss": 0.2727, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.15936, | |
| "grad_norm": 0.3631153670647941, | |
| "learning_rate": 4.8414897955201556e-05, | |
| "loss": 0.2674, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.39442642636399, | |
| "learning_rate": 4.8396287615174484e-05, | |
| "loss": 0.2653, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.16064, | |
| "grad_norm": 0.3999569444426015, | |
| "learning_rate": 4.837757228120762e-05, | |
| "loss": 0.297, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.16128, | |
| "grad_norm": 0.3348408050068023, | |
| "learning_rate": 4.835875203728849e-05, | |
| "loss": 0.2608, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.16192, | |
| "grad_norm": 0.4012425485757949, | |
| "learning_rate": 4.8339826967875443e-05, | |
| "loss": 0.2966, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.16256, | |
| "grad_norm": 0.39410881558128635, | |
| "learning_rate": 4.832079715789724e-05, | |
| "loss": 0.3064, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.1632, | |
| "grad_norm": 0.43770904619065176, | |
| "learning_rate": 4.830166269275267e-05, | |
| "loss": 0.2988, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.16384, | |
| "grad_norm": 0.34230150447086477, | |
| "learning_rate": 4.828242365831016e-05, | |
| "loss": 0.283, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.16448, | |
| "grad_norm": 0.4037086139344497, | |
| "learning_rate": 4.8263080140907455e-05, | |
| "loss": 0.2982, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.16512, | |
| "grad_norm": 0.44544678937787124, | |
| "learning_rate": 4.824363222735114e-05, | |
| "loss": 0.2821, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.16576, | |
| "grad_norm": 0.4065131032313394, | |
| "learning_rate": 4.82240800049163e-05, | |
| "loss": 0.2855, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 0.5004502107420943, | |
| "learning_rate": 4.820442356134612e-05, | |
| "loss": 0.2787, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.16704, | |
| "grad_norm": 0.4602519114887701, | |
| "learning_rate": 4.8184662984851505e-05, | |
| "loss": 0.3095, | |
| "step": 261 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1562, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 261, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 119007652085760.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |