| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 171, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.017543859649122806, | |
| "grad_norm": 2.3181249915556488, | |
| "learning_rate": 0.0, | |
| "loss": 0.8483, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03508771929824561, | |
| "grad_norm": 2.3031609820520065, | |
| "learning_rate": 6.666666666666667e-07, | |
| "loss": 0.8765, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05263157894736842, | |
| "grad_norm": 2.5315969251116233, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 1.0209, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.07017543859649122, | |
| "grad_norm": 2.6115330874356784, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.0621, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.08771929824561403, | |
| "grad_norm": 2.6177442321379703, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 0.8563, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 2.2800115956534657, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.0083, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.12280701754385964, | |
| "grad_norm": 1.8440531676316312, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.9982, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.14035087719298245, | |
| "grad_norm": 1.818151618586041, | |
| "learning_rate": 4.666666666666667e-06, | |
| "loss": 0.9474, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.15789473684210525, | |
| "grad_norm": 1.7230747772024742, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 0.9616, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.17543859649122806, | |
| "grad_norm": 1.3680905039669917, | |
| "learning_rate": 6e-06, | |
| "loss": 0.9869, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.19298245614035087, | |
| "grad_norm": 1.4455967664395566, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.0296, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 2.838519352672828, | |
| "learning_rate": 7.333333333333333e-06, | |
| "loss": 0.9008, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.22807017543859648, | |
| "grad_norm": 1.746600094936107, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.9385, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.24561403508771928, | |
| "grad_norm": 1.5993390926077906, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 0.9149, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 1.4190530136292803, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 0.9739, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2807017543859649, | |
| "grad_norm": 1.5476403448451475, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9248, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2982456140350877, | |
| "grad_norm": 1.8296015433545236, | |
| "learning_rate": 9.999661540018812e-06, | |
| "loss": 1.021, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "grad_norm": 1.4666326831664898, | |
| "learning_rate": 9.99864620589731e-06, | |
| "loss": 0.9051, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 1.2470444433529435, | |
| "learning_rate": 9.99695413509548e-06, | |
| "loss": 0.9253, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.3508771929824561, | |
| "grad_norm": 1.1769038399974867, | |
| "learning_rate": 9.994585556692624e-06, | |
| "loss": 0.9595, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3684210526315789, | |
| "grad_norm": 2.6056499740362202, | |
| "learning_rate": 9.991540791356342e-06, | |
| "loss": 0.8963, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.38596491228070173, | |
| "grad_norm": 1.2892838004713705, | |
| "learning_rate": 9.987820251299121e-06, | |
| "loss": 0.7664, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.40350877192982454, | |
| "grad_norm": 1.2011953377876559, | |
| "learning_rate": 9.98342444022253e-06, | |
| "loss": 0.9012, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 0.8570506663016774, | |
| "learning_rate": 9.978353953249023e-06, | |
| "loss": 0.7982, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.43859649122807015, | |
| "grad_norm": 0.8652112443789233, | |
| "learning_rate": 9.972609476841368e-06, | |
| "loss": 0.9253, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.45614035087719296, | |
| "grad_norm": 0.9653001049798393, | |
| "learning_rate": 9.966191788709716e-06, | |
| "loss": 0.8896, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.47368421052631576, | |
| "grad_norm": 0.8846034616258485, | |
| "learning_rate": 9.959101757706308e-06, | |
| "loss": 0.8099, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.49122807017543857, | |
| "grad_norm": 0.7810067582068709, | |
| "learning_rate": 9.951340343707852e-06, | |
| "loss": 0.8782, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5087719298245614, | |
| "grad_norm": 0.6544621596080209, | |
| "learning_rate": 9.942908597485558e-06, | |
| "loss": 0.8991, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 0.650602123344418, | |
| "learning_rate": 9.933807660562898e-06, | |
| "loss": 0.7848, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.543859649122807, | |
| "grad_norm": 0.6643753742626393, | |
| "learning_rate": 9.924038765061042e-06, | |
| "loss": 0.8407, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5614035087719298, | |
| "grad_norm": 0.6382815668704741, | |
| "learning_rate": 9.913603233532067e-06, | |
| "loss": 0.7703, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5789473684210527, | |
| "grad_norm": 0.7892950770322912, | |
| "learning_rate": 9.902502478779897e-06, | |
| "loss": 0.8698, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5964912280701754, | |
| "grad_norm": 0.7708210358290128, | |
| "learning_rate": 9.890738003669029e-06, | |
| "loss": 0.969, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.6140350877192983, | |
| "grad_norm": 0.8189039214769849, | |
| "learning_rate": 9.878311400921072e-06, | |
| "loss": 0.813, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 0.6904369205020999, | |
| "learning_rate": 9.86522435289912e-06, | |
| "loss": 0.8283, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.6491228070175439, | |
| "grad_norm": 0.6148536786862524, | |
| "learning_rate": 9.851478631379982e-06, | |
| "loss": 0.8632, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.5768762634588235, | |
| "learning_rate": 9.83707609731432e-06, | |
| "loss": 0.8763, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6842105263157895, | |
| "grad_norm": 0.6683735778256117, | |
| "learning_rate": 9.822018700574696e-06, | |
| "loss": 0.8983, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.7017543859649122, | |
| "grad_norm": 0.5217247956874884, | |
| "learning_rate": 9.806308479691595e-06, | |
| "loss": 0.8641, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7192982456140351, | |
| "grad_norm": 0.58257740620197, | |
| "learning_rate": 9.789947561577445e-06, | |
| "loss": 0.8692, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.7368421052631579, | |
| "grad_norm": 0.6232683716090549, | |
| "learning_rate": 9.77293816123866e-06, | |
| "loss": 0.8046, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.7543859649122807, | |
| "grad_norm": 0.6283296712711599, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.837, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.7719298245614035, | |
| "grad_norm": 0.5958179302224206, | |
| "learning_rate": 9.736983212571646e-06, | |
| "loss": 0.7531, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 0.5679750039894512, | |
| "learning_rate": 9.718042531967918e-06, | |
| "loss": 0.8698, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.8070175438596491, | |
| "grad_norm": 0.5878126492609907, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.8466, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.8245614035087719, | |
| "grad_norm": 0.5818690232830938, | |
| "learning_rate": 9.678247579197658e-06, | |
| "loss": 0.851, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 0.538107743012865, | |
| "learning_rate": 9.657398694630713e-06, | |
| "loss": 0.835, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.8596491228070176, | |
| "grad_norm": 0.7126873960175786, | |
| "learning_rate": 9.635919272833938e-06, | |
| "loss": 0.9265, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.8771929824561403, | |
| "grad_norm": 0.532600934634179, | |
| "learning_rate": 9.613812221777212e-06, | |
| "loss": 0.6671, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8947368421052632, | |
| "grad_norm": 0.5467907648189101, | |
| "learning_rate": 9.591080534401371e-06, | |
| "loss": 0.8962, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.9122807017543859, | |
| "grad_norm": 0.5277460536404931, | |
| "learning_rate": 9.567727288213005e-06, | |
| "loss": 0.8645, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.9298245614035088, | |
| "grad_norm": 0.49079200391848715, | |
| "learning_rate": 9.543755644867823e-06, | |
| "loss": 0.8165, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "grad_norm": 0.5531673037715197, | |
| "learning_rate": 9.519168849742603e-06, | |
| "loss": 0.7718, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.9649122807017544, | |
| "grad_norm": 0.514036737329238, | |
| "learning_rate": 9.493970231495836e-06, | |
| "loss": 0.8105, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.9824561403508771, | |
| "grad_norm": 0.5133770699616473, | |
| "learning_rate": 9.468163201617063e-06, | |
| "loss": 0.8217, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.5937230846401819, | |
| "learning_rate": 9.441751253965022e-06, | |
| "loss": 0.7339, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.0175438596491229, | |
| "grad_norm": 0.6230297978751398, | |
| "learning_rate": 9.414737964294636e-06, | |
| "loss": 0.8457, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.0350877192982457, | |
| "grad_norm": 0.6369933256359207, | |
| "learning_rate": 9.38712698977291e-06, | |
| "loss": 0.7362, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 0.589397997874693, | |
| "learning_rate": 9.358922068483813e-06, | |
| "loss": 0.8198, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0701754385964912, | |
| "grad_norm": 1.1922827708837067, | |
| "learning_rate": 9.330127018922195e-06, | |
| "loss": 0.6925, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.087719298245614, | |
| "grad_norm": 0.5755259625359568, | |
| "learning_rate": 9.30074573947683e-06, | |
| "loss": 0.7759, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.1052631578947367, | |
| "grad_norm": 0.6904382226436003, | |
| "learning_rate": 9.27078220790263e-06, | |
| "loss": 0.8048, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.1228070175438596, | |
| "grad_norm": 0.593152822599225, | |
| "learning_rate": 9.24024048078213e-06, | |
| "loss": 0.7966, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.1403508771929824, | |
| "grad_norm": 0.5119476599255871, | |
| "learning_rate": 9.209124692976287e-06, | |
| "loss": 0.7187, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.1578947368421053, | |
| "grad_norm": 0.563089185925041, | |
| "learning_rate": 9.177439057064684e-06, | |
| "loss": 0.722, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.1754385964912282, | |
| "grad_norm": 0.6121427386079497, | |
| "learning_rate": 9.145187862775208e-06, | |
| "loss": 0.7187, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.1929824561403508, | |
| "grad_norm": 0.5626488603303788, | |
| "learning_rate": 9.112375476403313e-06, | |
| "loss": 0.7695, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.2105263157894737, | |
| "grad_norm": 0.655350389412972, | |
| "learning_rate": 9.079006340220862e-06, | |
| "loss": 0.8125, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.2280701754385965, | |
| "grad_norm": 0.5513747526227838, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.7634, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2456140350877192, | |
| "grad_norm": 0.5335653254863885, | |
| "learning_rate": 9.01061596377522e-06, | |
| "loss": 0.8119, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.263157894736842, | |
| "grad_norm": 0.8092883052562865, | |
| "learning_rate": 8.97560398247424e-06, | |
| "loss": 0.7782, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.280701754385965, | |
| "grad_norm": 0.5792420164253886, | |
| "learning_rate": 8.94005376803361e-06, | |
| "loss": 0.7796, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.2982456140350878, | |
| "grad_norm": 0.5621117795245402, | |
| "learning_rate": 8.903970133383297e-06, | |
| "loss": 0.6995, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 0.47945869477443037, | |
| "learning_rate": 8.867357963669821e-06, | |
| "loss": 0.8251, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.8108211089894015, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.7235, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.3508771929824561, | |
| "grad_norm": 0.566581952297288, | |
| "learning_rate": 8.792567916744346e-06, | |
| "loss": 0.6927, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.368421052631579, | |
| "grad_norm": 0.7115261845014149, | |
| "learning_rate": 8.754400164907496e-06, | |
| "loss": 0.8149, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.3859649122807016, | |
| "grad_norm": 0.6136017997874252, | |
| "learning_rate": 8.715724127386971e-06, | |
| "loss": 0.7969, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.4035087719298245, | |
| "grad_norm": 0.5470125280490145, | |
| "learning_rate": 8.676545040299145e-06, | |
| "loss": 0.7421, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.4210526315789473, | |
| "grad_norm": 0.5632703929336123, | |
| "learning_rate": 8.636868207865244e-06, | |
| "loss": 0.7666, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.4385964912280702, | |
| "grad_norm": 0.5527321729817085, | |
| "learning_rate": 8.596699001693257e-06, | |
| "loss": 0.753, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.456140350877193, | |
| "grad_norm": 0.5442038276676865, | |
| "learning_rate": 8.556042860050686e-06, | |
| "loss": 0.6878, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.4736842105263157, | |
| "grad_norm": 0.6575613398984559, | |
| "learning_rate": 8.51490528712831e-06, | |
| "loss": 0.6168, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.4912280701754386, | |
| "grad_norm": 0.49844691265568325, | |
| "learning_rate": 8.473291852294986e-06, | |
| "loss": 0.7702, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.5087719298245614, | |
| "grad_norm": 0.9498809185392928, | |
| "learning_rate": 8.43120818934367e-06, | |
| "loss": 0.7551, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.526315789473684, | |
| "grad_norm": 0.49422861901011766, | |
| "learning_rate": 8.388659995728662e-06, | |
| "loss": 0.8855, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.543859649122807, | |
| "grad_norm": 0.46298742993415604, | |
| "learning_rate": 8.345653031794292e-06, | |
| "loss": 0.7727, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.5614035087719298, | |
| "grad_norm": 0.4936268735476307, | |
| "learning_rate": 8.302193119995038e-06, | |
| "loss": 0.8101, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 0.5085429430911613, | |
| "learning_rate": 8.258286144107277e-06, | |
| "loss": 0.7133, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5964912280701755, | |
| "grad_norm": 0.566577344849338, | |
| "learning_rate": 8.213938048432697e-06, | |
| "loss": 0.7486, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.6140350877192984, | |
| "grad_norm": 0.6180793961535552, | |
| "learning_rate": 8.16915483699355e-06, | |
| "loss": 0.7587, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.631578947368421, | |
| "grad_norm": 1.5446348618156136, | |
| "learning_rate": 8.123942572719801e-06, | |
| "loss": 0.7006, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.6491228070175439, | |
| "grad_norm": 0.4275568847926208, | |
| "learning_rate": 8.078307376628292e-06, | |
| "loss": 0.7049, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.5526953863989619, | |
| "learning_rate": 8.032255426994069e-06, | |
| "loss": 0.8032, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.6842105263157894, | |
| "grad_norm": 0.478174074472972, | |
| "learning_rate": 7.985792958513932e-06, | |
| "loss": 0.7065, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.7017543859649122, | |
| "grad_norm": 0.5015195769908033, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.8757, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.719298245614035, | |
| "grad_norm": 0.4956314811473689, | |
| "learning_rate": 7.891661680839932e-06, | |
| "loss": 0.7509, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.736842105263158, | |
| "grad_norm": 0.5672360300883434, | |
| "learning_rate": 7.84400561551426e-06, | |
| "loss": 0.7892, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.7543859649122808, | |
| "grad_norm": 0.5215781458691335, | |
| "learning_rate": 7.795964517353734e-06, | |
| "loss": 0.7305, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.7719298245614035, | |
| "grad_norm": 0.5643537700134045, | |
| "learning_rate": 7.747544890354031e-06, | |
| "loss": 0.8008, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.7894736842105263, | |
| "grad_norm": 0.5320255912476924, | |
| "learning_rate": 7.698753289757565e-06, | |
| "loss": 0.7972, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.807017543859649, | |
| "grad_norm": 0.5525517793925226, | |
| "learning_rate": 7.649596321166024e-06, | |
| "loss": 0.8186, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.8245614035087718, | |
| "grad_norm": 0.8168598036576498, | |
| "learning_rate": 7.600080639646077e-06, | |
| "loss": 0.6691, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.8421052631578947, | |
| "grad_norm": 0.6865587544638265, | |
| "learning_rate": 7.550212948828377e-06, | |
| "loss": 0.648, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.8596491228070176, | |
| "grad_norm": 1.1185527286362207, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.8352, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.8771929824561404, | |
| "grad_norm": 0.6112393692749535, | |
| "learning_rate": 7.449448591190436e-06, | |
| "loss": 0.7669, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.8947368421052633, | |
| "grad_norm": 0.9620940166649647, | |
| "learning_rate": 7.398565566251232e-06, | |
| "loss": 0.7878, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.912280701754386, | |
| "grad_norm": 0.61784270717678, | |
| "learning_rate": 7.347357813929455e-06, | |
| "loss": 0.7035, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.9298245614035088, | |
| "grad_norm": 0.631437859687955, | |
| "learning_rate": 7.295832266935059e-06, | |
| "loss": 0.8081, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.9473684210526314, | |
| "grad_norm": 0.5199165972432765, | |
| "learning_rate": 7.243995901002312e-06, | |
| "loss": 0.8047, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.9649122807017543, | |
| "grad_norm": 0.5695554248113175, | |
| "learning_rate": 7.191855733945388e-06, | |
| "loss": 0.7765, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.9824561403508771, | |
| "grad_norm": 0.9237970152596228, | |
| "learning_rate": 7.1394188247082715e-06, | |
| "loss": 0.8523, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.6858125378850107, | |
| "learning_rate": 7.08669227240909e-06, | |
| "loss": 0.7581, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.017543859649123, | |
| "grad_norm": 0.785994369925905, | |
| "learning_rate": 7.033683215379002e-06, | |
| "loss": 0.6335, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.0350877192982457, | |
| "grad_norm": 0.5064549061271131, | |
| "learning_rate": 6.980398830195785e-06, | |
| "loss": 0.676, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.0526315789473686, | |
| "grad_norm": 0.7179603442135754, | |
| "learning_rate": 6.9268463307122425e-06, | |
| "loss": 0.8513, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.0701754385964914, | |
| "grad_norm": 1.4797085153108631, | |
| "learning_rate": 6.873032967079562e-06, | |
| "loss": 0.7187, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.087719298245614, | |
| "grad_norm": 0.8049866394990322, | |
| "learning_rate": 6.818966024765758e-06, | |
| "loss": 0.697, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 0.6394181440634134, | |
| "learning_rate": 6.7646528235693445e-06, | |
| "loss": 0.6782, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.1228070175438596, | |
| "grad_norm": 0.7126202450367011, | |
| "learning_rate": 6.710100716628345e-06, | |
| "loss": 0.671, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.1403508771929824, | |
| "grad_norm": 0.9011490981866277, | |
| "learning_rate": 6.655317089424791e-06, | |
| "loss": 0.6242, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.1578947368421053, | |
| "grad_norm": 0.5865060751427547, | |
| "learning_rate": 6.600309358784858e-06, | |
| "loss": 0.7732, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.175438596491228, | |
| "grad_norm": 0.5460452862369597, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.6756, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.192982456140351, | |
| "grad_norm": 1.100575175010165, | |
| "learning_rate": 6.48965140519241e-06, | |
| "loss": 0.773, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.2105263157894735, | |
| "grad_norm": 0.7724724448009763, | |
| "learning_rate": 6.434016163555452e-06, | |
| "loss": 0.6874, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.2280701754385963, | |
| "grad_norm": 0.586955110593518, | |
| "learning_rate": 6.378186779084996e-06, | |
| "loss": 0.6274, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.245614035087719, | |
| "grad_norm": 0.548706035385071, | |
| "learning_rate": 6.322170810186013e-06, | |
| "loss": 0.6548, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.263157894736842, | |
| "grad_norm": 0.8491842498083872, | |
| "learning_rate": 6.26597584052401e-06, | |
| "loss": 0.6965, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.280701754385965, | |
| "grad_norm": 0.7101171530615974, | |
| "learning_rate": 6.209609477998339e-06, | |
| "loss": 0.7097, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.2982456140350878, | |
| "grad_norm": 0.5785602672886258, | |
| "learning_rate": 6.153079353712201e-06, | |
| "loss": 0.6871, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.3157894736842106, | |
| "grad_norm": 0.5301807594357945, | |
| "learning_rate": 6.0963931209395165e-06, | |
| "loss": 0.6499, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 0.7961140456990252, | |
| "learning_rate": 6.039558454088796e-06, | |
| "loss": 0.7494, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.3508771929824563, | |
| "grad_norm": 0.6843893164143875, | |
| "learning_rate": 5.982583047664151e-06, | |
| "loss": 0.6616, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.3684210526315788, | |
| "grad_norm": 0.6978093548119424, | |
| "learning_rate": 5.925474615223573e-06, | |
| "loss": 0.6797, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.3859649122807016, | |
| "grad_norm": 1.0874752950243343, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.6717, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.4035087719298245, | |
| "grad_norm": 1.4461585738563913, | |
| "learning_rate": 5.810889615527839e-06, | |
| "loss": 0.7796, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.4210526315789473, | |
| "grad_norm": 2.7738789376173196, | |
| "learning_rate": 5.753428561247416e-06, | |
| "loss": 0.7642, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.43859649122807, | |
| "grad_norm": 0.7995584014288799, | |
| "learning_rate": 5.695865504800328e-06, | |
| "loss": 0.7226, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.456140350877193, | |
| "grad_norm": 0.5776142257886474, | |
| "learning_rate": 5.638208239302975e-06, | |
| "loss": 0.6788, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.473684210526316, | |
| "grad_norm": 0.5511179277284277, | |
| "learning_rate": 5.5804645706261515e-06, | |
| "loss": 0.7219, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.4912280701754383, | |
| "grad_norm": 0.6521291861111554, | |
| "learning_rate": 5.522642316338268e-06, | |
| "loss": 0.7287, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.5087719298245617, | |
| "grad_norm": 0.5320162974492163, | |
| "learning_rate": 5.464749304646963e-06, | |
| "loss": 0.5747, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.526315789473684, | |
| "grad_norm": 0.5343013681765985, | |
| "learning_rate": 5.406793373339292e-06, | |
| "loss": 0.6644, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.543859649122807, | |
| "grad_norm": 0.5505777650291347, | |
| "learning_rate": 5.348782368720627e-06, | |
| "loss": 0.6802, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.56140350877193, | |
| "grad_norm": 0.7316466874755883, | |
| "learning_rate": 5.290724144552379e-06, | |
| "loss": 0.7391, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.5789473684210527, | |
| "grad_norm": 0.5222045626855123, | |
| "learning_rate": 5.232626560988735e-06, | |
| "loss": 0.6765, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.5964912280701755, | |
| "grad_norm": 0.5265557612972067, | |
| "learning_rate": 5.174497483512506e-06, | |
| "loss": 0.6991, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.6140350877192984, | |
| "grad_norm": 0.6547315249725427, | |
| "learning_rate": 5.116344781870282e-06, | |
| "loss": 0.6726, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 1.3071345213590821, | |
| "learning_rate": 5.0581763290069865e-06, | |
| "loss": 0.7574, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.6491228070175437, | |
| "grad_norm": 0.5473493446643941, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7411, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.638114136219989, | |
| "learning_rate": 4.941823670993016e-06, | |
| "loss": 0.7339, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.6842105263157894, | |
| "grad_norm": 0.5676468027606216, | |
| "learning_rate": 4.883655218129719e-06, | |
| "loss": 0.7287, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.7017543859649122, | |
| "grad_norm": 0.6545780377534246, | |
| "learning_rate": 4.825502516487497e-06, | |
| "loss": 0.7263, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.719298245614035, | |
| "grad_norm": 0.6457797041030401, | |
| "learning_rate": 4.767373439011267e-06, | |
| "loss": 0.7444, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.736842105263158, | |
| "grad_norm": 0.4862748398288923, | |
| "learning_rate": 4.7092758554476215e-06, | |
| "loss": 0.6798, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.754385964912281, | |
| "grad_norm": 0.4750150888754816, | |
| "learning_rate": 4.651217631279374e-06, | |
| "loss": 0.6983, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.7719298245614032, | |
| "grad_norm": 0.48750976133668983, | |
| "learning_rate": 4.59320662666071e-06, | |
| "loss": 0.6084, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.7894736842105265, | |
| "grad_norm": 0.5033516808053461, | |
| "learning_rate": 4.53525069535304e-06, | |
| "loss": 0.6704, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.807017543859649, | |
| "grad_norm": 0.646269259484673, | |
| "learning_rate": 4.477357683661734e-06, | |
| "loss": 0.5749, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.824561403508772, | |
| "grad_norm": 0.5246351704666878, | |
| "learning_rate": 4.4195354293738484e-06, | |
| "loss": 0.6792, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.8421052631578947, | |
| "grad_norm": 0.5760992862285343, | |
| "learning_rate": 4.361791760697027e-06, | |
| "loss": 0.7105, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.8596491228070176, | |
| "grad_norm": 0.5316646124664971, | |
| "learning_rate": 4.304134495199675e-06, | |
| "loss": 0.7188, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.8771929824561404, | |
| "grad_norm": 0.4419039659691918, | |
| "learning_rate": 4.246571438752585e-06, | |
| "loss": 0.6904, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.8947368421052633, | |
| "grad_norm": 0.5976560284785514, | |
| "learning_rate": 4.189110384472164e-06, | |
| "loss": 0.7557, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.912280701754386, | |
| "grad_norm": 0.6056940699390365, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.6911, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.9298245614035086, | |
| "grad_norm": 0.5705520786902705, | |
| "learning_rate": 4.074525384776428e-06, | |
| "loss": 0.6305, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.9473684210526314, | |
| "grad_norm": 0.5134337635276869, | |
| "learning_rate": 4.017416952335849e-06, | |
| "loss": 0.6758, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.9649122807017543, | |
| "grad_norm": 0.9115143820579311, | |
| "learning_rate": 3.960441545911205e-06, | |
| "loss": 0.6977, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.982456140350877, | |
| "grad_norm": 0.5332516662466225, | |
| "learning_rate": 3.903606879060483e-06, | |
| "loss": 0.7112, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.687899070827854, | |
| "learning_rate": 3.8469206462878e-06, | |
| "loss": 0.5381, | |
| "step": 171 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 285, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 39317700280320.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |