{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.48,
"eval_steps": 500,
"global_step": 60,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.008,
"grad_norm": 29.132591247558594,
"learning_rate": 1e-05,
"loss": 14.6502,
"mean_token_accuracy": 0.44139818847179413,
"step": 1
},
{
"epoch": 0.016,
"grad_norm": 26.889101028442383,
"learning_rate": 2e-05,
"loss": 14.1037,
"mean_token_accuracy": 0.46050838381052017,
"step": 2
},
{
"epoch": 0.024,
"grad_norm": 25.33008575439453,
"learning_rate": 3e-05,
"loss": 13.9492,
"mean_token_accuracy": 0.45162031054496765,
"step": 3
},
{
"epoch": 0.032,
"grad_norm": 21.760456085205078,
"learning_rate": 4e-05,
"loss": 13.5332,
"mean_token_accuracy": 0.47065450996160507,
"step": 4
},
{
"epoch": 0.04,
"grad_norm": 19.281452178955078,
"learning_rate": 5e-05,
"loss": 11.718,
"mean_token_accuracy": 0.5306272506713867,
"step": 5
},
{
"epoch": 0.048,
"grad_norm": 17.71002197265625,
"learning_rate": 4.909090909090909e-05,
"loss": 11.6275,
"mean_token_accuracy": 0.5233139544725418,
"step": 6
},
{
"epoch": 0.056,
"grad_norm": 18.48400115966797,
"learning_rate": 4.8181818181818186e-05,
"loss": 10.769,
"mean_token_accuracy": 0.538578063249588,
"step": 7
},
{
"epoch": 0.064,
"grad_norm": 17.125812530517578,
"learning_rate": 4.7272727272727275e-05,
"loss": 10.1511,
"mean_token_accuracy": 0.57903091609478,
"step": 8
},
{
"epoch": 0.072,
"grad_norm": 16.800792694091797,
"learning_rate": 4.636363636363636e-05,
"loss": 9.5746,
"mean_token_accuracy": 0.5945043712854385,
"step": 9
},
{
"epoch": 0.08,
"grad_norm": 13.959136009216309,
"learning_rate": 4.545454545454546e-05,
"loss": 9.5138,
"mean_token_accuracy": 0.6166220307350159,
"step": 10
},
{
"epoch": 0.088,
"grad_norm": 13.918156623840332,
"learning_rate": 4.454545454545455e-05,
"loss": 9.3013,
"mean_token_accuracy": 0.6187369078397751,
"step": 11
},
{
"epoch": 0.096,
"grad_norm": 13.260101318359375,
"learning_rate": 4.3636363636363636e-05,
"loss": 8.5085,
"mean_token_accuracy": 0.6477507650852203,
"step": 12
},
{
"epoch": 0.104,
"grad_norm": 11.593629837036133,
"learning_rate": 4.2727272727272724e-05,
"loss": 8.5987,
"mean_token_accuracy": 0.6391231864690781,
"step": 13
},
{
"epoch": 0.112,
"grad_norm": 10.642716407775879,
"learning_rate": 4.181818181818182e-05,
"loss": 8.0015,
"mean_token_accuracy": 0.6515648812055588,
"step": 14
},
{
"epoch": 0.12,
"grad_norm": 10.687582015991211,
"learning_rate": 4.0909090909090915e-05,
"loss": 7.8515,
"mean_token_accuracy": 0.659519150853157,
"step": 15
},
{
"epoch": 0.128,
"grad_norm": 11.342368125915527,
"learning_rate": 4e-05,
"loss": 8.1734,
"mean_token_accuracy": 0.6454901546239853,
"step": 16
},
{
"epoch": 0.136,
"grad_norm": 10.889402389526367,
"learning_rate": 3.909090909090909e-05,
"loss": 7.7197,
"mean_token_accuracy": 0.6463980078697205,
"step": 17
},
{
"epoch": 0.144,
"grad_norm": 11.274605751037598,
"learning_rate": 3.818181818181819e-05,
"loss": 7.7246,
"mean_token_accuracy": 0.6666717827320099,
"step": 18
},
{
"epoch": 0.152,
"grad_norm": 9.856607437133789,
"learning_rate": 3.7272727272727276e-05,
"loss": 7.9621,
"mean_token_accuracy": 0.6616508513689041,
"step": 19
},
{
"epoch": 0.16,
"grad_norm": 11.277185440063477,
"learning_rate": 3.6363636363636364e-05,
"loss": 7.6901,
"mean_token_accuracy": 0.6638920903205872,
"step": 20
},
{
"epoch": 0.168,
"grad_norm": 11.014278411865234,
"learning_rate": 3.545454545454546e-05,
"loss": 7.4989,
"mean_token_accuracy": 0.6643838137388229,
"step": 21
},
{
"epoch": 0.176,
"grad_norm": 10.972683906555176,
"learning_rate": 3.454545454545455e-05,
"loss": 6.8229,
"mean_token_accuracy": 0.7015579491853714,
"step": 22
},
{
"epoch": 0.184,
"grad_norm": 9.975464820861816,
"learning_rate": 3.3636363636363636e-05,
"loss": 7.3764,
"mean_token_accuracy": 0.6741979718208313,
"step": 23
},
{
"epoch": 0.192,
"grad_norm": 9.958491325378418,
"learning_rate": 3.272727272727273e-05,
"loss": 7.0166,
"mean_token_accuracy": 0.698300376534462,
"step": 24
},
{
"epoch": 0.2,
"grad_norm": 10.115056037902832,
"learning_rate": 3.181818181818182e-05,
"loss": 6.4697,
"mean_token_accuracy": 0.7078270465135574,
"step": 25
},
{
"epoch": 0.208,
"grad_norm": 10.057324409484863,
"learning_rate": 3.090909090909091e-05,
"loss": 6.6724,
"mean_token_accuracy": 0.7051044702529907,
"step": 26
},
{
"epoch": 0.216,
"grad_norm": 9.80183219909668,
"learning_rate": 3e-05,
"loss": 7.0968,
"mean_token_accuracy": 0.6842809170484543,
"step": 27
},
{
"epoch": 0.224,
"grad_norm": 10.466309547424316,
"learning_rate": 2.909090909090909e-05,
"loss": 6.8699,
"mean_token_accuracy": 0.7123119533061981,
"step": 28
},
{
"epoch": 0.232,
"grad_norm": 10.540712356567383,
"learning_rate": 2.818181818181818e-05,
"loss": 6.8986,
"mean_token_accuracy": 0.6990974545478821,
"step": 29
},
{
"epoch": 0.24,
"grad_norm": 10.496529579162598,
"learning_rate": 2.7272727272727273e-05,
"loss": 7.1639,
"mean_token_accuracy": 0.6841171979904175,
"step": 30
},
{
"epoch": 0.248,
"grad_norm": 10.347450256347656,
"learning_rate": 2.636363636363636e-05,
"loss": 7.1243,
"mean_token_accuracy": 0.7038989663124084,
"step": 31
},
{
"epoch": 0.256,
"grad_norm": 10.918871879577637,
"learning_rate": 2.5454545454545454e-05,
"loss": 6.5305,
"mean_token_accuracy": 0.7208158075809479,
"step": 32
},
{
"epoch": 0.264,
"grad_norm": 9.658543586730957,
"learning_rate": 2.4545454545454545e-05,
"loss": 6.4891,
"mean_token_accuracy": 0.7414443045854568,
"step": 33
},
{
"epoch": 0.272,
"grad_norm": 9.635473251342773,
"learning_rate": 2.3636363636363637e-05,
"loss": 6.6357,
"mean_token_accuracy": 0.73957559466362,
"step": 34
},
{
"epoch": 0.28,
"grad_norm": 9.881387710571289,
"learning_rate": 2.272727272727273e-05,
"loss": 6.8124,
"mean_token_accuracy": 0.7286674231290817,
"step": 35
},
{
"epoch": 0.288,
"grad_norm": 9.951563835144043,
"learning_rate": 2.1818181818181818e-05,
"loss": 6.8861,
"mean_token_accuracy": 0.7150181531906128,
"step": 36
},
{
"epoch": 0.296,
"grad_norm": 8.651752471923828,
"learning_rate": 2.090909090909091e-05,
"loss": 6.1332,
"mean_token_accuracy": 0.7416634857654572,
"step": 37
},
{
"epoch": 0.304,
"grad_norm": 9.848931312561035,
"learning_rate": 2e-05,
"loss": 6.2172,
"mean_token_accuracy": 0.737366572022438,
"step": 38
},
{
"epoch": 0.312,
"grad_norm": 9.531893730163574,
"learning_rate": 1.9090909090909094e-05,
"loss": 6.8838,
"mean_token_accuracy": 0.6978575885295868,
"step": 39
},
{
"epoch": 0.32,
"grad_norm": 9.349637031555176,
"learning_rate": 1.8181818181818182e-05,
"loss": 6.739,
"mean_token_accuracy": 0.7097857445478439,
"step": 40
},
{
"epoch": 0.328,
"grad_norm": 9.532185554504395,
"learning_rate": 1.7272727272727274e-05,
"loss": 6.0505,
"mean_token_accuracy": 0.7424566149711609,
"step": 41
},
{
"epoch": 0.336,
"grad_norm": 8.855209350585938,
"learning_rate": 1.6363636363636366e-05,
"loss": 6.4006,
"mean_token_accuracy": 0.7251224219799042,
"step": 42
},
{
"epoch": 0.344,
"grad_norm": 8.668252944946289,
"learning_rate": 1.5454545454545454e-05,
"loss": 6.0441,
"mean_token_accuracy": 0.7409052848815918,
"step": 43
},
{
"epoch": 0.352,
"grad_norm": 10.111790657043457,
"learning_rate": 1.4545454545454545e-05,
"loss": 6.173,
"mean_token_accuracy": 0.7369689494371414,
"step": 44
},
{
"epoch": 0.36,
"grad_norm": 8.958393096923828,
"learning_rate": 1.3636363636363637e-05,
"loss": 6.2257,
"mean_token_accuracy": 0.7140295207500458,
"step": 45
},
{
"epoch": 0.368,
"grad_norm": 10.362472534179688,
"learning_rate": 1.2727272727272727e-05,
"loss": 5.2547,
"mean_token_accuracy": 0.7552689164876938,
"step": 46
},
{
"epoch": 0.376,
"grad_norm": 9.94664478302002,
"learning_rate": 1.1818181818181819e-05,
"loss": 5.992,
"mean_token_accuracy": 0.7395860105752945,
"step": 47
},
{
"epoch": 0.384,
"grad_norm": 8.77272891998291,
"learning_rate": 1.0909090909090909e-05,
"loss": 6.4631,
"mean_token_accuracy": 0.7185734361410141,
"step": 48
},
{
"epoch": 0.392,
"grad_norm": 8.110837936401367,
"learning_rate": 1e-05,
"loss": 6.3143,
"mean_token_accuracy": 0.7158133089542389,
"step": 49
},
{
"epoch": 0.4,
"grad_norm": 8.038437843322754,
"learning_rate": 9.090909090909091e-06,
"loss": 6.8103,
"mean_token_accuracy": 0.7134602963924408,
"step": 50
},
{
"epoch": 0.408,
"grad_norm": 8.795354843139648,
"learning_rate": 8.181818181818183e-06,
"loss": 6.0269,
"mean_token_accuracy": 0.7261971086263657,
"step": 51
},
{
"epoch": 0.416,
"grad_norm": 8.536953926086426,
"learning_rate": 7.272727272727272e-06,
"loss": 5.355,
"mean_token_accuracy": 0.7585936486721039,
"step": 52
},
{
"epoch": 0.424,
"grad_norm": 8.2533540725708,
"learning_rate": 6.363636363636363e-06,
"loss": 6.5612,
"mean_token_accuracy": 0.7138941884040833,
"step": 53
},
{
"epoch": 0.432,
"grad_norm": 9.274043083190918,
"learning_rate": 5.4545454545454545e-06,
"loss": 6.4098,
"mean_token_accuracy": 0.7215629369020462,
"step": 54
},
{
"epoch": 0.44,
"grad_norm": 9.127202987670898,
"learning_rate": 4.5454545454545455e-06,
"loss": 6.5323,
"mean_token_accuracy": 0.7295349985361099,
"step": 55
},
{
"epoch": 0.448,
"grad_norm": 8.477712631225586,
"learning_rate": 3.636363636363636e-06,
"loss": 5.4184,
"mean_token_accuracy": 0.7586539834737778,
"step": 56
},
{
"epoch": 0.456,
"grad_norm": 7.758520126342773,
"learning_rate": 2.7272727272727272e-06,
"loss": 6.2795,
"mean_token_accuracy": 0.7345724701881409,
"step": 57
},
{
"epoch": 0.464,
"grad_norm": 8.064261436462402,
"learning_rate": 1.818181818181818e-06,
"loss": 5.3877,
"mean_token_accuracy": 0.7539169639348984,
"step": 58
},
{
"epoch": 0.472,
"grad_norm": 8.512259483337402,
"learning_rate": 9.09090909090909e-07,
"loss": 5.8114,
"mean_token_accuracy": 0.7418429106473923,
"step": 59
},
{
"epoch": 0.48,
"grad_norm": 8.13984203338623,
"learning_rate": 0.0,
"loss": 5.8393,
"mean_token_accuracy": 0.7534664273262024,
"step": 60
}
],
"logging_steps": 1,
"max_steps": 60,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"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": 313206736158720.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}