Qwen2.5-1.5B-Open-R1-Distill / trainer_state.json
chentianqi's picture
Model save
b79840a verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 268,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.018656716417910446,
"grad_norm": 1.8347556405877465,
"learning_rate": 1.785714285714286e-05,
"loss": 0.8455,
"step": 5
},
{
"epoch": 0.03731343283582089,
"grad_norm": 0.8138895515621757,
"learning_rate": 3.571428571428572e-05,
"loss": 0.7683,
"step": 10
},
{
"epoch": 0.055970149253731345,
"grad_norm": 0.5304912873979699,
"learning_rate": 4.999827900623038e-05,
"loss": 0.7025,
"step": 15
},
{
"epoch": 0.07462686567164178,
"grad_norm": 0.43348641951391986,
"learning_rate": 4.993807186343243e-05,
"loss": 0.6745,
"step": 20
},
{
"epoch": 0.09328358208955224,
"grad_norm": 0.44737860611801894,
"learning_rate": 4.979207812402531e-05,
"loss": 0.6437,
"step": 25
},
{
"epoch": 0.11194029850746269,
"grad_norm": 0.34788482160414036,
"learning_rate": 4.956085596012407e-05,
"loss": 0.6365,
"step": 30
},
{
"epoch": 0.13059701492537312,
"grad_norm": 0.37612788213843273,
"learning_rate": 4.924528939432311e-05,
"loss": 0.6203,
"step": 35
},
{
"epoch": 0.14925373134328357,
"grad_norm": 0.30080903467219317,
"learning_rate": 4.884658491984735e-05,
"loss": 0.611,
"step": 40
},
{
"epoch": 0.16791044776119404,
"grad_norm": 0.2587626020744064,
"learning_rate": 4.8366266887814235e-05,
"loss": 0.6117,
"step": 45
},
{
"epoch": 0.1865671641791045,
"grad_norm": 0.4346155151035265,
"learning_rate": 4.780617167924209e-05,
"loss": 0.5946,
"step": 50
},
{
"epoch": 0.20522388059701493,
"grad_norm": 0.35789197328214567,
"learning_rate": 4.716844068408693e-05,
"loss": 0.5971,
"step": 55
},
{
"epoch": 0.22388059701492538,
"grad_norm": 0.26978493002328224,
"learning_rate": 4.6455512114150546e-05,
"loss": 0.5924,
"step": 60
},
{
"epoch": 0.24253731343283583,
"grad_norm": 0.3559670356396701,
"learning_rate": 4.5670111681161296e-05,
"loss": 0.5832,
"step": 65
},
{
"epoch": 0.26119402985074625,
"grad_norm": 0.3741132211101352,
"learning_rate": 4.481524217566783e-05,
"loss": 0.5805,
"step": 70
},
{
"epoch": 0.2798507462686567,
"grad_norm": 0.6131512648194827,
"learning_rate": 4.3894171986588217e-05,
"loss": 0.5803,
"step": 75
},
{
"epoch": 0.29850746268656714,
"grad_norm": 0.48316310875225843,
"learning_rate": 4.29104226053073e-05,
"loss": 0.5788,
"step": 80
},
{
"epoch": 0.31716417910447764,
"grad_norm": 0.35265961474990465,
"learning_rate": 4.186775516209732e-05,
"loss": 0.5738,
"step": 85
},
{
"epoch": 0.3358208955223881,
"grad_norm": 0.35170242329056867,
"learning_rate": 4.077015604633669e-05,
"loss": 0.5762,
"step": 90
},
{
"epoch": 0.35447761194029853,
"grad_norm": 0.3319752030097362,
"learning_rate": 3.962182166550441e-05,
"loss": 0.5746,
"step": 95
},
{
"epoch": 0.373134328358209,
"grad_norm": 0.28806706591519804,
"learning_rate": 3.8427142401220634e-05,
"loss": 0.5705,
"step": 100
},
{
"epoch": 0.3917910447761194,
"grad_norm": 0.2824496074425936,
"learning_rate": 3.71906858236735e-05,
"loss": 0.5668,
"step": 105
},
{
"epoch": 0.41044776119402987,
"grad_norm": 0.28224550664809256,
"learning_rate": 3.591717922860785e-05,
"loss": 0.574,
"step": 110
},
{
"epoch": 0.4291044776119403,
"grad_norm": 0.2614488086229778,
"learning_rate": 3.46114915636416e-05,
"loss": 0.5649,
"step": 115
},
{
"epoch": 0.44776119402985076,
"grad_norm": 0.3105122853298353,
"learning_rate": 3.3278614813010034e-05,
"loss": 0.5659,
"step": 120
},
{
"epoch": 0.4664179104477612,
"grad_norm": 0.30904610340823935,
"learning_rate": 3.1923644911909e-05,
"loss": 0.5627,
"step": 125
},
{
"epoch": 0.48507462686567165,
"grad_norm": 0.2404641182221627,
"learning_rate": 3.0551762263406576e-05,
"loss": 0.5615,
"step": 130
},
{
"epoch": 0.503731343283582,
"grad_norm": 0.2514924850527194,
"learning_rate": 2.9168211932412042e-05,
"loss": 0.5586,
"step": 135
},
{
"epoch": 0.5223880597014925,
"grad_norm": 0.21507830724346733,
"learning_rate": 2.777828359242567e-05,
"loss": 0.564,
"step": 140
},
{
"epoch": 0.5410447761194029,
"grad_norm": 0.234838056524818,
"learning_rate": 2.6387291301738377e-05,
"loss": 0.5567,
"step": 145
},
{
"epoch": 0.5597014925373134,
"grad_norm": 0.23127993294001392,
"learning_rate": 2.50005531864019e-05,
"loss": 0.5545,
"step": 150
},
{
"epoch": 0.5783582089552238,
"grad_norm": 0.21437316758602867,
"learning_rate": 2.362337110764688e-05,
"loss": 0.5549,
"step": 155
},
{
"epoch": 0.5970149253731343,
"grad_norm": 0.19908887535771141,
"learning_rate": 2.226101039148557e-05,
"loss": 0.5529,
"step": 160
},
{
"epoch": 0.6156716417910447,
"grad_norm": 0.2556247686724649,
"learning_rate": 2.0918679697998252e-05,
"loss": 0.5519,
"step": 165
},
{
"epoch": 0.6343283582089553,
"grad_norm": 0.23026257493147687,
"learning_rate": 1.9601511107268255e-05,
"loss": 0.5523,
"step": 170
},
{
"epoch": 0.6529850746268657,
"grad_norm": 0.20786079342807257,
"learning_rate": 1.8314540498102216e-05,
"loss": 0.5521,
"step": 175
},
{
"epoch": 0.6716417910447762,
"grad_norm": 0.18862181222368476,
"learning_rate": 1.7062688294552992e-05,
"loss": 0.5441,
"step": 180
},
{
"epoch": 0.6902985074626866,
"grad_norm": 0.1968489941519543,
"learning_rate": 1.5850740653856096e-05,
"loss": 0.5475,
"step": 185
},
{
"epoch": 0.7089552238805971,
"grad_norm": 0.19127327607250483,
"learning_rate": 1.4683331167703218e-05,
"loss": 0.5511,
"step": 190
},
{
"epoch": 0.7276119402985075,
"grad_norm": 0.1782083084377943,
"learning_rate": 1.356492314681356e-05,
"loss": 0.5539,
"step": 195
},
{
"epoch": 0.746268656716418,
"grad_norm": 0.17219697852845983,
"learning_rate": 1.2499792556533716e-05,
"loss": 0.5482,
"step": 200
},
{
"epoch": 0.7649253731343284,
"grad_norm": 0.18622284000499204,
"learning_rate": 1.1492011668707753e-05,
"loss": 0.5457,
"step": 205
},
{
"epoch": 0.7835820895522388,
"grad_norm": 0.17736767965673514,
"learning_rate": 1.0545433492320603e-05,
"loss": 0.5508,
"step": 210
},
{
"epoch": 0.8022388059701493,
"grad_norm": 0.15925309459619932,
"learning_rate": 9.663677042440537e-06,
"loss": 0.5451,
"step": 215
},
{
"epoch": 0.8208955223880597,
"grad_norm": 0.16712076838059856,
"learning_rate": 8.850113503781367e-06,
"loss": 0.5451,
"step": 220
},
{
"epoch": 0.8395522388059702,
"grad_norm": 0.14902482359583405,
"learning_rate": 8.107853341784671e-06,
"loss": 0.5515,
"step": 225
},
{
"epoch": 0.8582089552238806,
"grad_norm": 0.17432109815408722,
"learning_rate": 7.439734410499752e-06,
"loss": 0.5478,
"step": 230
},
{
"epoch": 0.8768656716417911,
"grad_norm": 0.18977796427900687,
"learning_rate": 6.848311102728011e-06,
"loss": 0.548,
"step": 235
},
{
"epoch": 0.8955223880597015,
"grad_norm": 0.1779226032502227,
"learning_rate": 6.335844583913515e-06,
"loss": 0.5441,
"step": 240
},
{
"epoch": 0.914179104477612,
"grad_norm": 0.15085938404473737,
"learning_rate": 5.904294147118193e-06,
"loss": 0.5478,
"step": 245
},
{
"epoch": 0.9328358208955224,
"grad_norm": 0.14863412008011967,
"learning_rate": 5.555309722133842e-06,
"loss": 0.5444,
"step": 250
},
{
"epoch": 0.9514925373134329,
"grad_norm": 0.1485398957249496,
"learning_rate": 5.290225567370509e-06,
"loss": 0.5405,
"step": 255
},
{
"epoch": 0.9701492537313433,
"grad_norm": 0.15494251083627367,
"learning_rate": 5.110055168638854e-06,
"loss": 0.5441,
"step": 260
},
{
"epoch": 0.9888059701492538,
"grad_norm": 0.14954414034114885,
"learning_rate": 5.0154873643297575e-06,
"loss": 0.5479,
"step": 265
},
{
"epoch": 1.0,
"step": 268,
"total_flos": 487820711952384.0,
"train_loss": 0.5798122989597605,
"train_runtime": 9517.8097,
"train_samples_per_second": 3.604,
"train_steps_per_second": 0.028
}
],
"logging_steps": 5,
"max_steps": 268,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 487820711952384.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}