DEMO / stage1 /trainer_state.json
Xusy2333's picture
Fix: add tfevents to LFS
c3d7fcc
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 4632,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.3238341968911917,
"grad_norm": 3.4535982608795166,
"learning_rate": 4.460276338514681e-05,
"loss": 2.5692,
"step": 500
},
{
"epoch": 0.6476683937823834,
"grad_norm": 3.00873064994812,
"learning_rate": 3.920552677029361e-05,
"loss": 2.3948,
"step": 1000
},
{
"epoch": 0.9715025906735751,
"grad_norm": 2.5873517990112305,
"learning_rate": 3.380829015544041e-05,
"loss": 2.3669,
"step": 1500
},
{
"epoch": 1.2953367875647668,
"grad_norm": 2.6994502544403076,
"learning_rate": 2.841105354058722e-05,
"loss": 2.3413,
"step": 2000
},
{
"epoch": 1.6191709844559585,
"grad_norm": 2.5752243995666504,
"learning_rate": 2.3013816925734024e-05,
"loss": 2.3107,
"step": 2500
},
{
"epoch": 1.9430051813471503,
"grad_norm": 2.985656261444092,
"learning_rate": 1.761658031088083e-05,
"loss": 2.286,
"step": 3000
},
{
"epoch": 2.266839378238342,
"grad_norm": 2.945868492126465,
"learning_rate": 1.2219343696027635e-05,
"loss": 2.2789,
"step": 3500
},
{
"epoch": 2.5906735751295336,
"grad_norm": 3.70044207572937,
"learning_rate": 6.822107081174439e-06,
"loss": 2.2627,
"step": 4000
},
{
"epoch": 2.9145077720207255,
"grad_norm": 4.432509899139404,
"learning_rate": 1.4248704663212437e-06,
"loss": 2.2396,
"step": 4500
},
{
"epoch": 3.0,
"step": 4632,
"total_flos": 9.934475607251354e+16,
"train_loss": 2.3373115948033045,
"train_runtime": 795.3925,
"train_samples_per_second": 93.124,
"train_steps_per_second": 5.824
}
],
"logging_steps": 500,
"max_steps": 4632,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 9.934475607251354e+16,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}