zlm_b64_le4_s12000 / checkpoint-500 /trainer_state.json
mikhail-panzo's picture
Training in progress, step 500, checkpoint
286ec5a verified
raw
history blame
2.65 kB
{
"best_metric": 0.47458764910697937,
"best_model_checkpoint": "mikhail_panzo/zlm_b64_le4_s12000/checkpoint-500",
"epoch": 0.4187604690117253,
"eval_steps": 500,
"global_step": 500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.04187604690117253,
"grad_norm": 3.0184223651885986,
"learning_rate": 2.35e-06,
"loss": 1.1206,
"step": 50
},
{
"epoch": 0.08375209380234507,
"grad_norm": 3.4442005157470703,
"learning_rate": 4.85e-06,
"loss": 0.8185,
"step": 100
},
{
"epoch": 0.12562814070351758,
"grad_norm": 2.6812493801116943,
"learning_rate": 7.35e-06,
"loss": 0.7576,
"step": 150
},
{
"epoch": 0.16750418760469013,
"grad_norm": 4.518482685089111,
"learning_rate": 9.85e-06,
"loss": 0.6938,
"step": 200
},
{
"epoch": 0.20938023450586266,
"grad_norm": 4.683146953582764,
"learning_rate": 1.235e-05,
"loss": 0.6386,
"step": 250
},
{
"epoch": 0.25125628140703515,
"grad_norm": 2.7392578125,
"learning_rate": 1.485e-05,
"loss": 0.6226,
"step": 300
},
{
"epoch": 0.2931323283082077,
"grad_norm": 2.017056465148926,
"learning_rate": 1.7349999999999998e-05,
"loss": 0.6128,
"step": 350
},
{
"epoch": 0.33500837520938026,
"grad_norm": 3.6669230461120605,
"learning_rate": 1.985e-05,
"loss": 0.5675,
"step": 400
},
{
"epoch": 0.3768844221105528,
"grad_norm": 4.16896915435791,
"learning_rate": 2.235e-05,
"loss": 0.5655,
"step": 450
},
{
"epoch": 0.4187604690117253,
"grad_norm": 3.691459894180298,
"learning_rate": 2.485e-05,
"loss": 0.5487,
"step": 500
},
{
"epoch": 0.4187604690117253,
"eval_loss": 0.47458764910697937,
"eval_runtime": 214.7196,
"eval_samples_per_second": 39.535,
"eval_steps_per_second": 4.946,
"step": 500
}
],
"logging_steps": 50,
"max_steps": 12000,
"num_input_tokens_seen": 0,
"num_train_epochs": 11,
"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": 4873240581631488.0,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}