SALAMA_C3 / checkpoint-500 /trainer_state.json
EYEDOL's picture
Upload folder using huggingface_hub
32e5646 verified
{
"best_global_step": 500,
"best_metric": 19.59291781998664,
"best_model_checkpoint": "./SALAMA_C3/checkpoint-500",
"epoch": 0.8375209380234506,
"eval_steps": 500,
"global_step": 500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.03350083752093802,
"grad_norm": 11.757207870483398,
"learning_rate": 6.333333333333334e-07,
"loss": 0.5112,
"step": 20
},
{
"epoch": 0.06700167504187604,
"grad_norm": 10.197953224182129,
"learning_rate": 1.3e-06,
"loss": 0.5283,
"step": 40
},
{
"epoch": 0.10050251256281408,
"grad_norm": 12.811074256896973,
"learning_rate": 1.9666666666666668e-06,
"loss": 0.474,
"step": 60
},
{
"epoch": 0.13400335008375208,
"grad_norm": 8.559342384338379,
"learning_rate": 2.6e-06,
"loss": 0.424,
"step": 80
},
{
"epoch": 0.16750418760469013,
"grad_norm": 6.9629225730896,
"learning_rate": 3.266666666666667e-06,
"loss": 0.4465,
"step": 100
},
{
"epoch": 0.20100502512562815,
"grad_norm": 7.833058834075928,
"learning_rate": 3.9333333333333335e-06,
"loss": 0.3764,
"step": 120
},
{
"epoch": 0.23450586264656617,
"grad_norm": 6.882424831390381,
"learning_rate": 4.600000000000001e-06,
"loss": 0.3705,
"step": 140
},
{
"epoch": 0.26800670016750416,
"grad_norm": 5.4183244705200195,
"learning_rate": 5.2666666666666665e-06,
"loss": 0.4116,
"step": 160
},
{
"epoch": 0.3015075376884422,
"grad_norm": 6.103787899017334,
"learning_rate": 5.933333333333335e-06,
"loss": 0.3903,
"step": 180
},
{
"epoch": 0.33500837520938026,
"grad_norm": 6.084160804748535,
"learning_rate": 6.600000000000001e-06,
"loss": 0.3794,
"step": 200
},
{
"epoch": 0.3685092127303183,
"grad_norm": 4.531406879425049,
"learning_rate": 7.266666666666668e-06,
"loss": 0.3904,
"step": 220
},
{
"epoch": 0.4020100502512563,
"grad_norm": 6.702854156494141,
"learning_rate": 7.933333333333334e-06,
"loss": 0.3672,
"step": 240
},
{
"epoch": 0.4355108877721943,
"grad_norm": 6.386377811431885,
"learning_rate": 8.6e-06,
"loss": 0.3379,
"step": 260
},
{
"epoch": 0.46901172529313234,
"grad_norm": 5.33281135559082,
"learning_rate": 9.266666666666667e-06,
"loss": 0.4102,
"step": 280
},
{
"epoch": 0.5025125628140703,
"grad_norm": 6.031435012817383,
"learning_rate": 9.933333333333334e-06,
"loss": 0.3815,
"step": 300
},
{
"epoch": 0.5360134003350083,
"grad_norm": 4.8638176918029785,
"learning_rate": 9.968253968253969e-06,
"loss": 0.358,
"step": 320
},
{
"epoch": 0.5695142378559463,
"grad_norm": 5.352113723754883,
"learning_rate": 9.932980599647268e-06,
"loss": 0.3578,
"step": 340
},
{
"epoch": 0.6030150753768844,
"grad_norm": 5.560739040374756,
"learning_rate": 9.897707231040565e-06,
"loss": 0.3795,
"step": 360
},
{
"epoch": 0.6365159128978225,
"grad_norm": 6.186940670013428,
"learning_rate": 9.862433862433864e-06,
"loss": 0.3852,
"step": 380
},
{
"epoch": 0.6700167504187605,
"grad_norm": 5.37507438659668,
"learning_rate": 9.827160493827161e-06,
"loss": 0.3761,
"step": 400
},
{
"epoch": 0.7035175879396985,
"grad_norm": 6.373449802398682,
"learning_rate": 9.79188712522046e-06,
"loss": 0.3674,
"step": 420
},
{
"epoch": 0.7370184254606366,
"grad_norm": 5.702625274658203,
"learning_rate": 9.756613756613757e-06,
"loss": 0.4013,
"step": 440
},
{
"epoch": 0.7705192629815746,
"grad_norm": 5.5675153732299805,
"learning_rate": 9.721340388007056e-06,
"loss": 0.3495,
"step": 460
},
{
"epoch": 0.8040201005025126,
"grad_norm": 6.296374320983887,
"learning_rate": 9.686067019400353e-06,
"loss": 0.3704,
"step": 480
},
{
"epoch": 0.8375209380234506,
"grad_norm": 4.830463886260986,
"learning_rate": 9.650793650793652e-06,
"loss": 0.36,
"step": 500
},
{
"epoch": 0.8375209380234506,
"eval_loss": 0.27764827013015747,
"eval_runtime": 1752.3826,
"eval_samples_per_second": 2.725,
"eval_steps_per_second": 0.341,
"eval_wer": 19.59291781998664,
"step": 500
}
],
"logging_steps": 20,
"max_steps": 5970,
"num_input_tokens_seen": 0,
"num_train_epochs": 10,
"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": 4.61736640512e+18,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}