| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.2792022792022792, | |
| "eval_steps": 500, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.011396011396011397, | |
| "grad_norm": 1.0688412189483643, | |
| "learning_rate": 0.0001, | |
| "loss": 2.5442, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.022792022792022793, | |
| "grad_norm": 0.9676798582077026, | |
| "learning_rate": 9.949748743718594e-05, | |
| "loss": 2.4169, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03418803418803419, | |
| "grad_norm": 0.9322265386581421, | |
| "learning_rate": 9.899497487437186e-05, | |
| "loss": 2.4585, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.045584045584045586, | |
| "grad_norm": 1.0203752517700195, | |
| "learning_rate": 9.84924623115578e-05, | |
| "loss": 2.2605, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05698005698005698, | |
| "grad_norm": 1.0615314245224, | |
| "learning_rate": 9.798994974874372e-05, | |
| "loss": 2.2029, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06837606837606838, | |
| "grad_norm": 1.1191452741622925, | |
| "learning_rate": 9.748743718592965e-05, | |
| "loss": 1.977, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07977207977207977, | |
| "grad_norm": 1.0572948455810547, | |
| "learning_rate": 9.698492462311559e-05, | |
| "loss": 1.8041, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.09116809116809117, | |
| "grad_norm": 1.0000556707382202, | |
| "learning_rate": 9.64824120603015e-05, | |
| "loss": 1.8135, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.10256410256410256, | |
| "grad_norm": 0.815621554851532, | |
| "learning_rate": 9.597989949748745e-05, | |
| "loss": 1.6268, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.11396011396011396, | |
| "grad_norm": 0.8220420479774475, | |
| "learning_rate": 9.547738693467337e-05, | |
| "loss": 1.556, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.12535612535612536, | |
| "grad_norm": 0.8398631811141968, | |
| "learning_rate": 9.49748743718593e-05, | |
| "loss": 1.4275, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.13675213675213677, | |
| "grad_norm": 0.7354830503463745, | |
| "learning_rate": 9.447236180904523e-05, | |
| "loss": 1.4687, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.14814814814814814, | |
| "grad_norm": 0.4654861092567444, | |
| "learning_rate": 9.396984924623115e-05, | |
| "loss": 1.3937, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.15954415954415954, | |
| "grad_norm": 0.5842018127441406, | |
| "learning_rate": 9.34673366834171e-05, | |
| "loss": 1.4105, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.17094017094017094, | |
| "grad_norm": 0.6819984912872314, | |
| "learning_rate": 9.296482412060302e-05, | |
| "loss": 1.3469, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.18233618233618235, | |
| "grad_norm": 0.5046871900558472, | |
| "learning_rate": 9.246231155778895e-05, | |
| "loss": 1.3817, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.19373219373219372, | |
| "grad_norm": 0.4931313693523407, | |
| "learning_rate": 9.195979899497488e-05, | |
| "loss": 1.4015, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.20512820512820512, | |
| "grad_norm": 0.5237986445426941, | |
| "learning_rate": 9.14572864321608e-05, | |
| "loss": 1.3251, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.21652421652421652, | |
| "grad_norm": 0.48392462730407715, | |
| "learning_rate": 9.095477386934675e-05, | |
| "loss": 1.2821, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.22792022792022792, | |
| "grad_norm": 0.5208500623703003, | |
| "learning_rate": 9.045226130653267e-05, | |
| "loss": 1.3226, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.23931623931623933, | |
| "grad_norm": 0.5178021192550659, | |
| "learning_rate": 8.99497487437186e-05, | |
| "loss": 1.2268, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.25071225071225073, | |
| "grad_norm": 0.5599659085273743, | |
| "learning_rate": 8.944723618090453e-05, | |
| "loss": 1.3742, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2621082621082621, | |
| "grad_norm": 0.5551819801330566, | |
| "learning_rate": 8.894472361809045e-05, | |
| "loss": 1.3156, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.27350427350427353, | |
| "grad_norm": 0.7024655342102051, | |
| "learning_rate": 8.84422110552764e-05, | |
| "loss": 1.2187, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2849002849002849, | |
| "grad_norm": 0.6340409517288208, | |
| "learning_rate": 8.793969849246232e-05, | |
| "loss": 1.3423, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "grad_norm": 0.5692094564437866, | |
| "learning_rate": 8.743718592964825e-05, | |
| "loss": 1.2315, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 0.5804877877235413, | |
| "learning_rate": 8.693467336683418e-05, | |
| "loss": 1.2316, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3190883190883191, | |
| "grad_norm": 0.7234011292457581, | |
| "learning_rate": 8.64321608040201e-05, | |
| "loss": 1.2583, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.33048433048433046, | |
| "grad_norm": 0.8010082244873047, | |
| "learning_rate": 8.592964824120603e-05, | |
| "loss": 1.3704, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.3418803418803419, | |
| "grad_norm": 0.6479464173316956, | |
| "learning_rate": 8.542713567839196e-05, | |
| "loss": 1.3294, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.35327635327635326, | |
| "grad_norm": 0.778668224811554, | |
| "learning_rate": 8.49246231155779e-05, | |
| "loss": 1.2566, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.3646723646723647, | |
| "grad_norm": 0.8151825070381165, | |
| "learning_rate": 8.442211055276383e-05, | |
| "loss": 1.2372, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.37606837606837606, | |
| "grad_norm": 0.777619481086731, | |
| "learning_rate": 8.391959798994975e-05, | |
| "loss": 1.2749, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.38746438746438744, | |
| "grad_norm": 0.8822659850120544, | |
| "learning_rate": 8.341708542713568e-05, | |
| "loss": 1.2229, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.39886039886039887, | |
| "grad_norm": 0.870242178440094, | |
| "learning_rate": 8.291457286432161e-05, | |
| "loss": 1.1641, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.41025641025641024, | |
| "grad_norm": 0.8463490605354309, | |
| "learning_rate": 8.241206030150754e-05, | |
| "loss": 1.2443, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.42165242165242167, | |
| "grad_norm": 0.8682456612586975, | |
| "learning_rate": 8.190954773869348e-05, | |
| "loss": 1.2818, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.43304843304843305, | |
| "grad_norm": 0.6559503674507141, | |
| "learning_rate": 8.14070351758794e-05, | |
| "loss": 1.1601, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.603629469871521, | |
| "learning_rate": 8.090452261306533e-05, | |
| "loss": 1.2256, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.45584045584045585, | |
| "grad_norm": 0.5692597031593323, | |
| "learning_rate": 8.040201005025126e-05, | |
| "loss": 1.2869, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4672364672364672, | |
| "grad_norm": 0.5320606231689453, | |
| "learning_rate": 7.989949748743719e-05, | |
| "loss": 1.179, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.47863247863247865, | |
| "grad_norm": 0.5642545819282532, | |
| "learning_rate": 7.939698492462313e-05, | |
| "loss": 1.1968, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.49002849002849, | |
| "grad_norm": 0.4435971975326538, | |
| "learning_rate": 7.889447236180904e-05, | |
| "loss": 1.2094, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5014245014245015, | |
| "grad_norm": 0.4562317430973053, | |
| "learning_rate": 7.839195979899498e-05, | |
| "loss": 1.1734, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.5128205128205128, | |
| "grad_norm": 0.5164006948471069, | |
| "learning_rate": 7.788944723618091e-05, | |
| "loss": 1.2177, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5242165242165242, | |
| "grad_norm": 0.5158300399780273, | |
| "learning_rate": 7.738693467336684e-05, | |
| "loss": 1.2289, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.5356125356125356, | |
| "grad_norm": 0.49650484323501587, | |
| "learning_rate": 7.688442211055277e-05, | |
| "loss": 1.3316, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.5470085470085471, | |
| "grad_norm": 0.4756147563457489, | |
| "learning_rate": 7.638190954773869e-05, | |
| "loss": 1.333, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.5584045584045584, | |
| "grad_norm": 0.4514218866825104, | |
| "learning_rate": 7.587939698492463e-05, | |
| "loss": 1.2733, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5698005698005698, | |
| "grad_norm": 0.445081889629364, | |
| "learning_rate": 7.537688442211056e-05, | |
| "loss": 1.2024, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5811965811965812, | |
| "grad_norm": 0.45231911540031433, | |
| "learning_rate": 7.487437185929649e-05, | |
| "loss": 1.267, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 0.5107349753379822, | |
| "learning_rate": 7.437185929648241e-05, | |
| "loss": 1.2068, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.603988603988604, | |
| "grad_norm": 0.4770635664463043, | |
| "learning_rate": 7.386934673366834e-05, | |
| "loss": 1.2372, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 0.4619287848472595, | |
| "learning_rate": 7.336683417085427e-05, | |
| "loss": 1.2132, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.6267806267806267, | |
| "grad_norm": 0.5125857591629028, | |
| "learning_rate": 7.28643216080402e-05, | |
| "loss": 1.2367, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6381766381766382, | |
| "grad_norm": 0.456436425447464, | |
| "learning_rate": 7.236180904522614e-05, | |
| "loss": 1.1989, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.6495726495726496, | |
| "grad_norm": 0.4466511011123657, | |
| "learning_rate": 7.185929648241206e-05, | |
| "loss": 1.2789, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.6609686609686609, | |
| "grad_norm": 0.45993903279304504, | |
| "learning_rate": 7.135678391959799e-05, | |
| "loss": 1.2228, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.6723646723646723, | |
| "grad_norm": 0.4762590825557709, | |
| "learning_rate": 7.085427135678392e-05, | |
| "loss": 1.1418, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6837606837606838, | |
| "grad_norm": 0.4990002512931824, | |
| "learning_rate": 7.035175879396985e-05, | |
| "loss": 1.1788, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6951566951566952, | |
| "grad_norm": 0.4810471534729004, | |
| "learning_rate": 6.984924623115579e-05, | |
| "loss": 1.1878, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7065527065527065, | |
| "grad_norm": 0.4777512550354004, | |
| "learning_rate": 6.93467336683417e-05, | |
| "loss": 1.2621, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.717948717948718, | |
| "grad_norm": 0.5083452463150024, | |
| "learning_rate": 6.884422110552764e-05, | |
| "loss": 1.1642, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.7293447293447294, | |
| "grad_norm": 0.4965672194957733, | |
| "learning_rate": 6.834170854271357e-05, | |
| "loss": 1.1356, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 0.48566991090774536, | |
| "learning_rate": 6.78391959798995e-05, | |
| "loss": 1.0954, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7521367521367521, | |
| "grad_norm": 0.46452316641807556, | |
| "learning_rate": 6.733668341708544e-05, | |
| "loss": 1.1905, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.7635327635327636, | |
| "grad_norm": 0.51093989610672, | |
| "learning_rate": 6.683417085427135e-05, | |
| "loss": 1.2285, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.7749287749287749, | |
| "grad_norm": 0.5919416546821594, | |
| "learning_rate": 6.633165829145729e-05, | |
| "loss": 1.2368, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.7863247863247863, | |
| "grad_norm": 0.46627846360206604, | |
| "learning_rate": 6.582914572864322e-05, | |
| "loss": 1.1286, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7977207977207977, | |
| "grad_norm": 0.4974450170993805, | |
| "learning_rate": 6.532663316582915e-05, | |
| "loss": 1.2874, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.8091168091168092, | |
| "grad_norm": 0.5373516082763672, | |
| "learning_rate": 6.482412060301508e-05, | |
| "loss": 1.2573, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.8205128205128205, | |
| "grad_norm": 0.4522000849246979, | |
| "learning_rate": 6.4321608040201e-05, | |
| "loss": 1.137, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.8319088319088319, | |
| "grad_norm": 0.4694693088531494, | |
| "learning_rate": 6.381909547738694e-05, | |
| "loss": 1.1971, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.8433048433048433, | |
| "grad_norm": 0.4889134466648102, | |
| "learning_rate": 6.331658291457287e-05, | |
| "loss": 1.1435, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.8547008547008547, | |
| "grad_norm": 0.512048602104187, | |
| "learning_rate": 6.28140703517588e-05, | |
| "loss": 1.2086, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.8660968660968661, | |
| "grad_norm": 0.49475014209747314, | |
| "learning_rate": 6.231155778894473e-05, | |
| "loss": 1.2496, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.8774928774928775, | |
| "grad_norm": 0.4658482074737549, | |
| "learning_rate": 6.180904522613065e-05, | |
| "loss": 1.0942, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.5323877334594727, | |
| "learning_rate": 6.130653266331658e-05, | |
| "loss": 1.1853, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.9002849002849003, | |
| "grad_norm": 0.6251657605171204, | |
| "learning_rate": 6.080402010050251e-05, | |
| "loss": 1.2209, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.9116809116809117, | |
| "grad_norm": 0.4768678545951843, | |
| "learning_rate": 6.030150753768844e-05, | |
| "loss": 1.164, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 0.5291458964347839, | |
| "learning_rate": 5.979899497487438e-05, | |
| "loss": 1.2754, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.9344729344729344, | |
| "grad_norm": 0.5092456340789795, | |
| "learning_rate": 5.929648241206031e-05, | |
| "loss": 1.248, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.9458689458689459, | |
| "grad_norm": 0.4834723174571991, | |
| "learning_rate": 5.879396984924623e-05, | |
| "loss": 1.2681, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.9572649572649573, | |
| "grad_norm": 0.5478146076202393, | |
| "learning_rate": 5.829145728643216e-05, | |
| "loss": 1.2045, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.9686609686609686, | |
| "grad_norm": 0.4918864071369171, | |
| "learning_rate": 5.778894472361809e-05, | |
| "loss": 1.1216, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.98005698005698, | |
| "grad_norm": 0.4812568128108978, | |
| "learning_rate": 5.728643216080403e-05, | |
| "loss": 1.1732, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.9914529914529915, | |
| "grad_norm": 0.5308701992034912, | |
| "learning_rate": 5.6783919597989955e-05, | |
| "loss": 1.2643, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.002849002849003, | |
| "grad_norm": 0.7254398465156555, | |
| "learning_rate": 5.628140703517588e-05, | |
| "loss": 1.4278, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.0142450142450143, | |
| "grad_norm": 0.5298280715942383, | |
| "learning_rate": 5.577889447236181e-05, | |
| "loss": 1.2691, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.0256410256410255, | |
| "grad_norm": 0.4411105811595917, | |
| "learning_rate": 5.527638190954774e-05, | |
| "loss": 1.1033, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.037037037037037, | |
| "grad_norm": 0.5160195231437683, | |
| "learning_rate": 5.477386934673368e-05, | |
| "loss": 1.2041, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.0484330484330484, | |
| "grad_norm": 0.4904952943325043, | |
| "learning_rate": 5.4271356783919604e-05, | |
| "loss": 1.1634, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.0598290598290598, | |
| "grad_norm": 0.4598182439804077, | |
| "learning_rate": 5.376884422110553e-05, | |
| "loss": 1.1095, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.0712250712250713, | |
| "grad_norm": 0.4579429626464844, | |
| "learning_rate": 5.3266331658291455e-05, | |
| "loss": 1.0614, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.0826210826210827, | |
| "grad_norm": 0.49531427025794983, | |
| "learning_rate": 5.276381909547739e-05, | |
| "loss": 1.2415, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.0940170940170941, | |
| "grad_norm": 0.4900212287902832, | |
| "learning_rate": 5.226130653266332e-05, | |
| "loss": 1.08, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.1054131054131053, | |
| "grad_norm": 0.5321851372718811, | |
| "learning_rate": 5.175879396984925e-05, | |
| "loss": 1.1464, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.1168091168091168, | |
| "grad_norm": 0.4976498782634735, | |
| "learning_rate": 5.125628140703518e-05, | |
| "loss": 1.1962, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.1282051282051282, | |
| "grad_norm": 0.48046809434890747, | |
| "learning_rate": 5.0753768844221104e-05, | |
| "loss": 1.0534, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.1396011396011396, | |
| "grad_norm": 0.5641231536865234, | |
| "learning_rate": 5.0251256281407036e-05, | |
| "loss": 1.2342, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.150997150997151, | |
| "grad_norm": 0.5218859314918518, | |
| "learning_rate": 4.974874371859297e-05, | |
| "loss": 1.1125, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.1623931623931625, | |
| "grad_norm": 0.672106146812439, | |
| "learning_rate": 4.92462311557789e-05, | |
| "loss": 1.2577, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.173789173789174, | |
| "grad_norm": 0.5047236680984497, | |
| "learning_rate": 4.874371859296483e-05, | |
| "loss": 1.0518, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 0.5189758539199829, | |
| "learning_rate": 4.824120603015075e-05, | |
| "loss": 1.1672, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.1965811965811965, | |
| "grad_norm": 0.5212465524673462, | |
| "learning_rate": 4.7738693467336685e-05, | |
| "loss": 1.0869, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.207977207977208, | |
| "grad_norm": 0.5158497095108032, | |
| "learning_rate": 4.723618090452262e-05, | |
| "loss": 1.0674, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.2193732193732194, | |
| "grad_norm": 0.6164978742599487, | |
| "learning_rate": 4.673366834170855e-05, | |
| "loss": 1.2204, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.5276105403900146, | |
| "learning_rate": 4.6231155778894475e-05, | |
| "loss": 1.0782, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.242165242165242, | |
| "grad_norm": 0.5990796685218811, | |
| "learning_rate": 4.57286432160804e-05, | |
| "loss": 1.1748, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.2535612535612537, | |
| "grad_norm": 0.5942894816398621, | |
| "learning_rate": 4.522613065326633e-05, | |
| "loss": 1.1417, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.264957264957265, | |
| "grad_norm": 0.5517327189445496, | |
| "learning_rate": 4.4723618090452266e-05, | |
| "loss": 1.0214, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.2763532763532763, | |
| "grad_norm": 0.599429190158844, | |
| "learning_rate": 4.42211055276382e-05, | |
| "loss": 1.2503, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.2877492877492878, | |
| "grad_norm": 0.5922709107398987, | |
| "learning_rate": 4.3718592964824124e-05, | |
| "loss": 1.1294, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.2991452991452992, | |
| "grad_norm": 0.569146990776062, | |
| "learning_rate": 4.321608040201005e-05, | |
| "loss": 1.1523, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.3105413105413106, | |
| "grad_norm": 0.5592817664146423, | |
| "learning_rate": 4.271356783919598e-05, | |
| "loss": 1.1484, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.3219373219373218, | |
| "grad_norm": 0.5917912125587463, | |
| "learning_rate": 4.2211055276381914e-05, | |
| "loss": 1.0667, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.6044990420341492, | |
| "learning_rate": 4.170854271356784e-05, | |
| "loss": 1.1646, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.3447293447293447, | |
| "grad_norm": 0.5149083137512207, | |
| "learning_rate": 4.120603015075377e-05, | |
| "loss": 1.0567, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.3561253561253561, | |
| "grad_norm": 0.5668403506278992, | |
| "learning_rate": 4.07035175879397e-05, | |
| "loss": 1.2072, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.3675213675213675, | |
| "grad_norm": 0.5478379726409912, | |
| "learning_rate": 4.020100502512563e-05, | |
| "loss": 1.1353, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.378917378917379, | |
| "grad_norm": 0.5570518374443054, | |
| "learning_rate": 3.969849246231156e-05, | |
| "loss": 1.0336, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.3903133903133904, | |
| "grad_norm": 0.5794707536697388, | |
| "learning_rate": 3.919597989949749e-05, | |
| "loss": 1.1696, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.4017094017094016, | |
| "grad_norm": 0.567593514919281, | |
| "learning_rate": 3.869346733668342e-05, | |
| "loss": 1.0635, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.413105413105413, | |
| "grad_norm": 0.5602433085441589, | |
| "learning_rate": 3.8190954773869346e-05, | |
| "loss": 1.1443, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.4245014245014245, | |
| "grad_norm": 0.578921377658844, | |
| "learning_rate": 3.768844221105528e-05, | |
| "loss": 1.1843, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.435897435897436, | |
| "grad_norm": 0.5648573040962219, | |
| "learning_rate": 3.7185929648241204e-05, | |
| "loss": 1.1076, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.4472934472934473, | |
| "grad_norm": 0.5635711550712585, | |
| "learning_rate": 3.668341708542714e-05, | |
| "loss": 1.0589, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.4586894586894588, | |
| "grad_norm": 0.6067689061164856, | |
| "learning_rate": 3.618090452261307e-05, | |
| "loss": 1.1941, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.4700854700854702, | |
| "grad_norm": 0.5677550435066223, | |
| "learning_rate": 3.5678391959798995e-05, | |
| "loss": 1.0384, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.4814814814814814, | |
| "grad_norm": 0.599470853805542, | |
| "learning_rate": 3.517587939698493e-05, | |
| "loss": 1.2259, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.4928774928774928, | |
| "grad_norm": 0.6475313305854797, | |
| "learning_rate": 3.467336683417085e-05, | |
| "loss": 1.1618, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.5042735042735043, | |
| "grad_norm": 0.6463358402252197, | |
| "learning_rate": 3.4170854271356785e-05, | |
| "loss": 1.1712, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.5156695156695157, | |
| "grad_norm": 0.5662721395492554, | |
| "learning_rate": 3.366834170854272e-05, | |
| "loss": 1.0731, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.5270655270655271, | |
| "grad_norm": 0.5981451869010925, | |
| "learning_rate": 3.3165829145728643e-05, | |
| "loss": 1.0298, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.5991083979606628, | |
| "learning_rate": 3.2663316582914576e-05, | |
| "loss": 1.2169, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.54985754985755, | |
| "grad_norm": 0.6036386489868164, | |
| "learning_rate": 3.21608040201005e-05, | |
| "loss": 1.0498, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.5612535612535612, | |
| "grad_norm": 0.5927392840385437, | |
| "learning_rate": 3.1658291457286434e-05, | |
| "loss": 1.0389, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.5726495726495726, | |
| "grad_norm": 0.5733420848846436, | |
| "learning_rate": 3.1155778894472366e-05, | |
| "loss": 1.0608, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.584045584045584, | |
| "grad_norm": 0.6083365678787231, | |
| "learning_rate": 3.065326633165829e-05, | |
| "loss": 1.1203, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.5954415954415955, | |
| "grad_norm": 0.6153535842895508, | |
| "learning_rate": 3.015075376884422e-05, | |
| "loss": 1.1729, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.606837606837607, | |
| "grad_norm": 0.6425400376319885, | |
| "learning_rate": 2.9648241206030153e-05, | |
| "loss": 1.0155, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.618233618233618, | |
| "grad_norm": 0.6353899240493774, | |
| "learning_rate": 2.914572864321608e-05, | |
| "loss": 1.245, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.6296296296296298, | |
| "grad_norm": 0.6624664664268494, | |
| "learning_rate": 2.8643216080402015e-05, | |
| "loss": 0.9795, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.641025641025641, | |
| "grad_norm": 0.6005571484565735, | |
| "learning_rate": 2.814070351758794e-05, | |
| "loss": 1.1581, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.6524216524216524, | |
| "grad_norm": 0.6583634614944458, | |
| "learning_rate": 2.763819095477387e-05, | |
| "loss": 1.0538, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.6638176638176638, | |
| "grad_norm": 0.5750309228897095, | |
| "learning_rate": 2.7135678391959802e-05, | |
| "loss": 0.9153, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.6752136752136753, | |
| "grad_norm": 0.6019430160522461, | |
| "learning_rate": 2.6633165829145728e-05, | |
| "loss": 1.159, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.6866096866096867, | |
| "grad_norm": 0.6395899057388306, | |
| "learning_rate": 2.613065326633166e-05, | |
| "loss": 1.1699, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.698005698005698, | |
| "grad_norm": 0.635377824306488, | |
| "learning_rate": 2.562814070351759e-05, | |
| "loss": 1.0286, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.7094017094017095, | |
| "grad_norm": 0.6403070688247681, | |
| "learning_rate": 2.5125628140703518e-05, | |
| "loss": 1.1748, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.7207977207977208, | |
| "grad_norm": 0.6614859104156494, | |
| "learning_rate": 2.462311557788945e-05, | |
| "loss": 1.0518, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.7321937321937322, | |
| "grad_norm": 0.7036588788032532, | |
| "learning_rate": 2.4120603015075376e-05, | |
| "loss": 1.2191, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.7435897435897436, | |
| "grad_norm": 0.6717102527618408, | |
| "learning_rate": 2.361809045226131e-05, | |
| "loss": 1.1376, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.7549857549857548, | |
| "grad_norm": 0.6435654759407043, | |
| "learning_rate": 2.3115577889447238e-05, | |
| "loss": 1.1601, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.7663817663817665, | |
| "grad_norm": 0.6151710748672485, | |
| "learning_rate": 2.2613065326633167e-05, | |
| "loss": 1.0929, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.6438819169998169, | |
| "learning_rate": 2.21105527638191e-05, | |
| "loss": 1.111, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.7891737891737893, | |
| "grad_norm": 0.6459853649139404, | |
| "learning_rate": 2.1608040201005025e-05, | |
| "loss": 1.2303, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.8005698005698005, | |
| "grad_norm": 0.6052287220954895, | |
| "learning_rate": 2.1105527638190957e-05, | |
| "loss": 1.1341, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.811965811965812, | |
| "grad_norm": 0.6797654032707214, | |
| "learning_rate": 2.0603015075376886e-05, | |
| "loss": 1.1724, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.8233618233618234, | |
| "grad_norm": 0.6900933980941772, | |
| "learning_rate": 2.0100502512562815e-05, | |
| "loss": 1.0881, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.8347578347578346, | |
| "grad_norm": 0.6317200064659119, | |
| "learning_rate": 1.9597989949748744e-05, | |
| "loss": 1.1256, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.6044368743896484, | |
| "learning_rate": 1.9095477386934673e-05, | |
| "loss": 1.0595, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.8575498575498575, | |
| "grad_norm": 0.6719862818717957, | |
| "learning_rate": 1.8592964824120602e-05, | |
| "loss": 0.9983, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.868945868945869, | |
| "grad_norm": 0.6419474482536316, | |
| "learning_rate": 1.8090452261306535e-05, | |
| "loss": 1.0876, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.8803418803418803, | |
| "grad_norm": 0.6861122250556946, | |
| "learning_rate": 1.7587939698492464e-05, | |
| "loss": 1.0904, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.8917378917378918, | |
| "grad_norm": 0.6277052760124207, | |
| "learning_rate": 1.7085427135678393e-05, | |
| "loss": 1.1036, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.9031339031339032, | |
| "grad_norm": 0.7358347177505493, | |
| "learning_rate": 1.6582914572864322e-05, | |
| "loss": 1.0499, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.9145299145299144, | |
| "grad_norm": 0.6961327195167542, | |
| "learning_rate": 1.608040201005025e-05, | |
| "loss": 1.097, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.925925925925926, | |
| "grad_norm": 0.6499162316322327, | |
| "learning_rate": 1.5577889447236183e-05, | |
| "loss": 1.0566, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.9373219373219372, | |
| "grad_norm": 0.6426655650138855, | |
| "learning_rate": 1.507537688442211e-05, | |
| "loss": 1.1127, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.9487179487179487, | |
| "grad_norm": 0.6038071513175964, | |
| "learning_rate": 1.457286432160804e-05, | |
| "loss": 1.0345, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.96011396011396, | |
| "grad_norm": 0.6887624263763428, | |
| "learning_rate": 1.407035175879397e-05, | |
| "loss": 1.2775, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.9715099715099715, | |
| "grad_norm": 0.6664908528327942, | |
| "learning_rate": 1.3567839195979901e-05, | |
| "loss": 1.099, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.982905982905983, | |
| "grad_norm": 0.6395720839500427, | |
| "learning_rate": 1.306532663316583e-05, | |
| "loss": 1.0795, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.9943019943019942, | |
| "grad_norm": 0.8258576989173889, | |
| "learning_rate": 1.2562814070351759e-05, | |
| "loss": 1.4081, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.005698005698006, | |
| "grad_norm": 0.7664657831192017, | |
| "learning_rate": 1.2060301507537688e-05, | |
| "loss": 1.2999, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.017094017094017, | |
| "grad_norm": 0.6419854164123535, | |
| "learning_rate": 1.1557788944723619e-05, | |
| "loss": 1.0661, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.0284900284900287, | |
| "grad_norm": 0.6126803755760193, | |
| "learning_rate": 1.105527638190955e-05, | |
| "loss": 1.0989, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.03988603988604, | |
| "grad_norm": 0.5971490740776062, | |
| "learning_rate": 1.0552763819095479e-05, | |
| "loss": 0.9485, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.051282051282051, | |
| "grad_norm": 0.6343738436698914, | |
| "learning_rate": 1.0050251256281408e-05, | |
| "loss": 1.1226, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.0626780626780628, | |
| "grad_norm": 0.6217495203018188, | |
| "learning_rate": 9.547738693467337e-06, | |
| "loss": 1.0628, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.074074074074074, | |
| "grad_norm": 0.5966653823852539, | |
| "learning_rate": 9.045226130653267e-06, | |
| "loss": 1.0092, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.0854700854700856, | |
| "grad_norm": 0.6323566436767578, | |
| "learning_rate": 8.542713567839196e-06, | |
| "loss": 1.0992, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.096866096866097, | |
| "grad_norm": 0.6440880298614502, | |
| "learning_rate": 8.040201005025125e-06, | |
| "loss": 1.0804, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.1082621082621085, | |
| "grad_norm": 0.5995816588401794, | |
| "learning_rate": 7.537688442211055e-06, | |
| "loss": 1.1067, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.1196581196581197, | |
| "grad_norm": 0.6618144512176514, | |
| "learning_rate": 7.035175879396985e-06, | |
| "loss": 1.0689, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.131054131054131, | |
| "grad_norm": 0.6532097458839417, | |
| "learning_rate": 6.532663316582915e-06, | |
| "loss": 1.1028, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.1424501424501425, | |
| "grad_norm": 0.5830849409103394, | |
| "learning_rate": 6.030150753768844e-06, | |
| "loss": 0.9996, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.6783652901649475, | |
| "learning_rate": 5.527638190954775e-06, | |
| "loss": 1.2311, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.1652421652421654, | |
| "grad_norm": 0.6712796688079834, | |
| "learning_rate": 5.025125628140704e-06, | |
| "loss": 0.9609, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.1766381766381766, | |
| "grad_norm": 0.6146546006202698, | |
| "learning_rate": 4.522613065326634e-06, | |
| "loss": 1.1135, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.1880341880341883, | |
| "grad_norm": 0.6589621901512146, | |
| "learning_rate": 4.020100502512563e-06, | |
| "loss": 1.0524, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.1994301994301995, | |
| "grad_norm": 0.648345947265625, | |
| "learning_rate": 3.5175879396984926e-06, | |
| "loss": 1.1136, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.2108262108262107, | |
| "grad_norm": 0.6554787158966064, | |
| "learning_rate": 3.015075376884422e-06, | |
| "loss": 1.0393, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 0.60575270652771, | |
| "learning_rate": 2.512562814070352e-06, | |
| "loss": 1.0688, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.2336182336182335, | |
| "grad_norm": 0.609583854675293, | |
| "learning_rate": 2.0100502512562813e-06, | |
| "loss": 1.047, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.245014245014245, | |
| "grad_norm": 0.6172504425048828, | |
| "learning_rate": 1.507537688442211e-06, | |
| "loss": 1.0835, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.2564102564102564, | |
| "grad_norm": 0.6250450015068054, | |
| "learning_rate": 1.0050251256281407e-06, | |
| "loss": 1.0684, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.267806267806268, | |
| "grad_norm": 0.5692541599273682, | |
| "learning_rate": 5.025125628140703e-07, | |
| "loss": 0.9156, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.2792022792022792, | |
| "grad_norm": 0.6259585022926331, | |
| "learning_rate": 0.0, | |
| "loss": 1.2106, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.469981127242547e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |