435 lines
12 KiB
JSON
435 lines
12 KiB
JSON
|
{
|
||
|
|
"best_global_step": null,
|
||
|
|
"best_metric": null,
|
||
|
|
"best_model_checkpoint": null,
|
||
|
|
"epoch": 1.5625,
|
||
|
|
"eval_steps": 500,
|
||
|
|
"global_step": 2000,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"entropy": 0.8528947708010673,
|
||
|
|
"epoch": 0.0390625,
|
||
|
|
"grad_norm": 1.140625,
|
||
|
|
"learning_rate": 3.828125000000001e-06,
|
||
|
|
"loss": 1.0035149383544921,
|
||
|
|
"mean_token_accuracy": 0.7955961337685585,
|
||
|
|
"num_tokens": 289985.0,
|
||
|
|
"step": 50
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.19647314386442305,
|
||
|
|
"epoch": 0.078125,
|
||
|
|
"grad_norm": 1.03125,
|
||
|
|
"learning_rate": 7.734375e-06,
|
||
|
|
"loss": 0.1893597412109375,
|
||
|
|
"mean_token_accuracy": 0.9560051780939102,
|
||
|
|
"num_tokens": 578511.0,
|
||
|
|
"step": 100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.09508214401081204,
|
||
|
|
"epoch": 0.1171875,
|
||
|
|
"grad_norm": 1.140625,
|
||
|
|
"learning_rate": 1.1640625000000002e-05,
|
||
|
|
"loss": 0.08917176246643066,
|
||
|
|
"mean_token_accuracy": 0.9763016304373742,
|
||
|
|
"num_tokens": 863089.0,
|
||
|
|
"step": 150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.06972143727354706,
|
||
|
|
"epoch": 0.15625,
|
||
|
|
"grad_norm": 1.03125,
|
||
|
|
"learning_rate": 1.5546875e-05,
|
||
|
|
"loss": 0.0670042610168457,
|
||
|
|
"mean_token_accuracy": 0.9812742775678635,
|
||
|
|
"num_tokens": 1152491.0,
|
||
|
|
"step": 200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.048063146471977236,
|
||
|
|
"epoch": 0.1953125,
|
||
|
|
"grad_norm": 0.78125,
|
||
|
|
"learning_rate": 1.9453125e-05,
|
||
|
|
"loss": 0.04478513240814209,
|
||
|
|
"mean_token_accuracy": 0.9859032621979713,
|
||
|
|
"num_tokens": 1439889.0,
|
||
|
|
"step": 250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.04555465183220804,
|
||
|
|
"epoch": 0.234375,
|
||
|
|
"grad_norm": 0.39453125,
|
||
|
|
"learning_rate": 1.9626736111111114e-05,
|
||
|
|
"loss": 0.04354703426361084,
|
||
|
|
"mean_token_accuracy": 0.986427218914032,
|
||
|
|
"num_tokens": 1724136.0,
|
||
|
|
"step": 300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.03841961392201483,
|
||
|
|
"epoch": 0.2734375,
|
||
|
|
"grad_norm": 0.40234375,
|
||
|
|
"learning_rate": 1.9192708333333335e-05,
|
||
|
|
"loss": 0.03762138843536377,
|
||
|
|
"mean_token_accuracy": 0.9881205716729164,
|
||
|
|
"num_tokens": 2005086.0,
|
||
|
|
"step": 350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.028935400638729333,
|
||
|
|
"epoch": 0.3125,
|
||
|
|
"grad_norm": 0.427734375,
|
||
|
|
"learning_rate": 1.8758680555555557e-05,
|
||
|
|
"loss": 0.027575111389160155,
|
||
|
|
"mean_token_accuracy": 0.9903467008471489,
|
||
|
|
"num_tokens": 2294186.0,
|
||
|
|
"step": 400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.025832002903334796,
|
||
|
|
"epoch": 0.3515625,
|
||
|
|
"grad_norm": 0.1171875,
|
||
|
|
"learning_rate": 1.8324652777777778e-05,
|
||
|
|
"loss": 0.024780447483062743,
|
||
|
|
"mean_token_accuracy": 0.9910794907808304,
|
||
|
|
"num_tokens": 2583118.0,
|
||
|
|
"step": 450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.02405791813507676,
|
||
|
|
"epoch": 0.390625,
|
||
|
|
"grad_norm": 0.34765625,
|
||
|
|
"learning_rate": 1.7890625000000003e-05,
|
||
|
|
"loss": 0.0229555344581604,
|
||
|
|
"mean_token_accuracy": 0.9914542031288147,
|
||
|
|
"num_tokens": 2869922.0,
|
||
|
|
"step": 500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.027367654335685075,
|
||
|
|
"epoch": 0.4296875,
|
||
|
|
"grad_norm": 0.130859375,
|
||
|
|
"learning_rate": 1.7456597222222224e-05,
|
||
|
|
"loss": 0.02657809019088745,
|
||
|
|
"mean_token_accuracy": 0.9907592672109604,
|
||
|
|
"num_tokens": 3151975.0,
|
||
|
|
"step": 550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.02466743137687445,
|
||
|
|
"epoch": 0.46875,
|
||
|
|
"grad_norm": 0.2275390625,
|
||
|
|
"learning_rate": 1.7022569444444446e-05,
|
||
|
|
"loss": 0.023617582321166994,
|
||
|
|
"mean_token_accuracy": 0.9914914935827255,
|
||
|
|
"num_tokens": 3441223.0,
|
||
|
|
"step": 600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.025295174946077168,
|
||
|
|
"epoch": 0.5078125,
|
||
|
|
"grad_norm": 0.23828125,
|
||
|
|
"learning_rate": 1.6588541666666667e-05,
|
||
|
|
"loss": 0.023536417484283447,
|
||
|
|
"mean_token_accuracy": 0.9913991931080818,
|
||
|
|
"num_tokens": 3725476.0,
|
||
|
|
"step": 650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.021516693667508663,
|
||
|
|
"epoch": 0.546875,
|
||
|
|
"grad_norm": 0.1376953125,
|
||
|
|
"learning_rate": 1.615451388888889e-05,
|
||
|
|
"loss": 0.02034649610519409,
|
||
|
|
"mean_token_accuracy": 0.9919455853104592,
|
||
|
|
"num_tokens": 4015608.0,
|
||
|
|
"step": 700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.02162676403298974,
|
||
|
|
"epoch": 0.5859375,
|
||
|
|
"grad_norm": 0.1025390625,
|
||
|
|
"learning_rate": 1.5720486111111114e-05,
|
||
|
|
"loss": 0.02098503351211548,
|
||
|
|
"mean_token_accuracy": 0.9920175784826278,
|
||
|
|
"num_tokens": 4297395.0,
|
||
|
|
"step": 750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.02071377928368747,
|
||
|
|
"epoch": 0.625,
|
||
|
|
"grad_norm": 0.13671875,
|
||
|
|
"learning_rate": 1.5286458333333335e-05,
|
||
|
|
"loss": 0.020009536743164063,
|
||
|
|
"mean_token_accuracy": 0.9920961040258408,
|
||
|
|
"num_tokens": 4584615.0,
|
||
|
|
"step": 800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.022029129918664694,
|
||
|
|
"epoch": 0.6640625,
|
||
|
|
"grad_norm": 0.07861328125,
|
||
|
|
"learning_rate": 1.4852430555555556e-05,
|
||
|
|
"loss": 0.021189706325531008,
|
||
|
|
"mean_token_accuracy": 0.991996766924858,
|
||
|
|
"num_tokens": 4865596.0,
|
||
|
|
"step": 850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.021224360242486,
|
||
|
|
"epoch": 0.703125,
|
||
|
|
"grad_norm": 0.12890625,
|
||
|
|
"learning_rate": 1.4418402777777778e-05,
|
||
|
|
"loss": 0.020408225059509278,
|
||
|
|
"mean_token_accuracy": 0.9920864021778106,
|
||
|
|
"num_tokens": 5151434.0,
|
||
|
|
"step": 900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.02010633623227477,
|
||
|
|
"epoch": 0.7421875,
|
||
|
|
"grad_norm": 0.734375,
|
||
|
|
"learning_rate": 1.3984375000000001e-05,
|
||
|
|
"loss": 0.01937895655632019,
|
||
|
|
"mean_token_accuracy": 0.9925438091158867,
|
||
|
|
"num_tokens": 5444488.0,
|
||
|
|
"step": 950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.01990706148557365,
|
||
|
|
"epoch": 0.78125,
|
||
|
|
"grad_norm": 0.1162109375,
|
||
|
|
"learning_rate": 1.3550347222222224e-05,
|
||
|
|
"loss": 0.019094927310943602,
|
||
|
|
"mean_token_accuracy": 0.9926347577571869,
|
||
|
|
"num_tokens": 5734302.0,
|
||
|
|
"step": 1000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.02041028759907931,
|
||
|
|
"epoch": 0.8203125,
|
||
|
|
"grad_norm": 0.09521484375,
|
||
|
|
"learning_rate": 1.3116319444444446e-05,
|
||
|
|
"loss": 0.019333359003067017,
|
||
|
|
"mean_token_accuracy": 0.9924440628290176,
|
||
|
|
"num_tokens": 6020184.0,
|
||
|
|
"step": 1050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.020887044854462147,
|
||
|
|
"epoch": 0.859375,
|
||
|
|
"grad_norm": 0.1904296875,
|
||
|
|
"learning_rate": 1.2682291666666669e-05,
|
||
|
|
"loss": 0.01991814136505127,
|
||
|
|
"mean_token_accuracy": 0.9922475999593735,
|
||
|
|
"num_tokens": 6307384.0,
|
||
|
|
"step": 1100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.020934463790617884,
|
||
|
|
"epoch": 0.8984375,
|
||
|
|
"grad_norm": 0.10546875,
|
||
|
|
"learning_rate": 1.2248263888888889e-05,
|
||
|
|
"loss": 0.020386738777160643,
|
||
|
|
"mean_token_accuracy": 0.992243467271328,
|
||
|
|
"num_tokens": 6590414.0,
|
||
|
|
"step": 1150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.018531152345240116,
|
||
|
|
"epoch": 0.9375,
|
||
|
|
"grad_norm": 0.072265625,
|
||
|
|
"learning_rate": 1.1814236111111112e-05,
|
||
|
|
"loss": 0.01763615131378174,
|
||
|
|
"mean_token_accuracy": 0.9930036315321922,
|
||
|
|
"num_tokens": 6879445.0,
|
||
|
|
"step": 1200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.018522553648799657,
|
||
|
|
"epoch": 0.9765625,
|
||
|
|
"grad_norm": 0.09521484375,
|
||
|
|
"learning_rate": 1.1380208333333333e-05,
|
||
|
|
"loss": 0.017574281692504884,
|
||
|
|
"mean_token_accuracy": 0.9928433075547218,
|
||
|
|
"num_tokens": 7168537.0,
|
||
|
|
"step": 1250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.01806912823114544,
|
||
|
|
"epoch": 1.015625,
|
||
|
|
"grad_norm": 0.08349609375,
|
||
|
|
"learning_rate": 1.0946180555555556e-05,
|
||
|
|
"loss": 0.017191458940505982,
|
||
|
|
"mean_token_accuracy": 0.9930350634455681,
|
||
|
|
"num_tokens": 7455527.0,
|
||
|
|
"step": 1300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.01792264294810593,
|
||
|
|
"epoch": 1.0546875,
|
||
|
|
"grad_norm": 0.1220703125,
|
||
|
|
"learning_rate": 1.0512152777777778e-05,
|
||
|
|
"loss": 0.017068485021591185,
|
||
|
|
"mean_token_accuracy": 0.9930369177460671,
|
||
|
|
"num_tokens": 7739474.0,
|
||
|
|
"step": 1350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.018599705449305476,
|
||
|
|
"epoch": 1.09375,
|
||
|
|
"grad_norm": 0.08740234375,
|
||
|
|
"learning_rate": 1.0078125000000001e-05,
|
||
|
|
"loss": 0.017695863246917725,
|
||
|
|
"mean_token_accuracy": 0.9929590111970902,
|
||
|
|
"num_tokens": 8026170.0,
|
||
|
|
"step": 1400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.018208509651012717,
|
||
|
|
"epoch": 1.1328125,
|
||
|
|
"grad_norm": 0.11181640625,
|
||
|
|
"learning_rate": 9.644097222222222e-06,
|
||
|
|
"loss": 0.01744949698448181,
|
||
|
|
"mean_token_accuracy": 0.9930833280086517,
|
||
|
|
"num_tokens": 8314626.0,
|
||
|
|
"step": 1450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.01731262981891632,
|
||
|
|
"epoch": 1.171875,
|
||
|
|
"grad_norm": 0.07080078125,
|
||
|
|
"learning_rate": 9.210069444444446e-06,
|
||
|
|
"loss": 0.016677556037902833,
|
||
|
|
"mean_token_accuracy": 0.9931734573841094,
|
||
|
|
"num_tokens": 8601406.0,
|
||
|
|
"step": 1500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.01845017326530069,
|
||
|
|
"epoch": 1.2109375,
|
||
|
|
"grad_norm": 0.087890625,
|
||
|
|
"learning_rate": 8.776041666666667e-06,
|
||
|
|
"loss": 0.017589352130889892,
|
||
|
|
"mean_token_accuracy": 0.9928986424207688,
|
||
|
|
"num_tokens": 8887025.0,
|
||
|
|
"step": 1550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.016624693870544435,
|
||
|
|
"epoch": 1.25,
|
||
|
|
"grad_norm": 0.138671875,
|
||
|
|
"learning_rate": 8.342013888888889e-06,
|
||
|
|
"loss": 0.015853718519210816,
|
||
|
|
"mean_token_accuracy": 0.9933998480439186,
|
||
|
|
"num_tokens": 9173407.0,
|
||
|
|
"step": 1600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.018500901735387744,
|
||
|
|
"epoch": 1.2890625,
|
||
|
|
"grad_norm": 0.1142578125,
|
||
|
|
"learning_rate": 7.907986111111112e-06,
|
||
|
|
"loss": 0.017626932859420776,
|
||
|
|
"mean_token_accuracy": 0.9930863061547279,
|
||
|
|
"num_tokens": 9454970.0,
|
||
|
|
"step": 1650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.01800002105999738,
|
||
|
|
"epoch": 1.328125,
|
||
|
|
"grad_norm": 0.0986328125,
|
||
|
|
"learning_rate": 7.473958333333334e-06,
|
||
|
|
"loss": 0.017116209268569948,
|
||
|
|
"mean_token_accuracy": 0.9930115470290184,
|
||
|
|
"num_tokens": 9739610.0,
|
||
|
|
"step": 1700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.017908857897855342,
|
||
|
|
"epoch": 1.3671875,
|
||
|
|
"grad_norm": 0.1064453125,
|
||
|
|
"learning_rate": 7.039930555555556e-06,
|
||
|
|
"loss": 0.017099602222442625,
|
||
|
|
"mean_token_accuracy": 0.9931422612071037,
|
||
|
|
"num_tokens": 10028723.0,
|
||
|
|
"step": 1750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.017319361912086606,
|
||
|
|
"epoch": 1.40625,
|
||
|
|
"grad_norm": 0.14453125,
|
||
|
|
"learning_rate": 6.605902777777779e-06,
|
||
|
|
"loss": 0.016603636741638183,
|
||
|
|
"mean_token_accuracy": 0.9933136883378029,
|
||
|
|
"num_tokens": 10313957.0,
|
||
|
|
"step": 1800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.017652450683526694,
|
||
|
|
"epoch": 1.4453125,
|
||
|
|
"grad_norm": 0.11572265625,
|
||
|
|
"learning_rate": 6.171875e-06,
|
||
|
|
"loss": 0.01667865037918091,
|
||
|
|
"mean_token_accuracy": 0.9932633358240127,
|
||
|
|
"num_tokens": 10597623.0,
|
||
|
|
"step": 1850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.01626451033167541,
|
||
|
|
"epoch": 1.484375,
|
||
|
|
"grad_norm": 0.119140625,
|
||
|
|
"learning_rate": 5.737847222222222e-06,
|
||
|
|
"loss": 0.015588784217834472,
|
||
|
|
"mean_token_accuracy": 0.993659851551056,
|
||
|
|
"num_tokens": 10887482.0,
|
||
|
|
"step": 1900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.016533851716667415,
|
||
|
|
"epoch": 1.5234375,
|
||
|
|
"grad_norm": 0.1630859375,
|
||
|
|
"learning_rate": 5.303819444444445e-06,
|
||
|
|
"loss": 0.01556604504585266,
|
||
|
|
"mean_token_accuracy": 0.9936296039819718,
|
||
|
|
"num_tokens": 11176455.0,
|
||
|
|
"step": 1950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"entropy": 0.017771934717893602,
|
||
|
|
"epoch": 1.5625,
|
||
|
|
"grad_norm": 0.11572265625,
|
||
|
|
"learning_rate": 4.869791666666667e-06,
|
||
|
|
"loss": 0.017040348052978514,
|
||
|
|
"mean_token_accuracy": 0.9932206523418426,
|
||
|
|
"num_tokens": 11462052.0,
|
||
|
|
"step": 2000
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 50,
|
||
|
|
"max_steps": 2560,
|
||
|
|
"num_input_tokens_seen": 0,
|
||
|
|
"num_train_epochs": 2,
|
||
|
|
"save_steps": 500,
|
||
|
|
"stateful_callbacks": {
|
||
|
|
"TrainerControl": {
|
||
|
|
"args": {
|
||
|
|
"should_epoch_stop": false,
|
||
|
|
"should_evaluate": false,
|
||
|
|
"should_log": false,
|
||
|
|
"should_save": true,
|
||
|
|
"should_training_stop": false
|
||
|
|
},
|
||
|
|
"attributes": {}
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"total_flos": 1.5546599214235546e+17,
|
||
|
|
"train_batch_size": 2,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|