Qwen-Code-7B / checkpoint-75 /trainer_state.json
JerryFan011018's picture
Upload folder using huggingface_hub
bba0851 verified
{
"best_global_step": 75,
"best_metric": 0.35836032032966614,
"best_model_checkpoint": "/content/output/checkpoint-75",
"epoch": 1.0,
"eval_steps": 500,
"global_step": 75,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.13333333333333333,
"grad_norm": 0.9156391620635986,
"learning_rate": 1.7543859649122806e-05,
"loss": 1.2171,
"mean_token_accuracy": 0.7375583708286285,
"num_tokens": 32413.0,
"step": 10
},
{
"epoch": 0.26666666666666666,
"grad_norm": 0.815642774105072,
"learning_rate": 3.508771929824561e-05,
"loss": 0.9632,
"mean_token_accuracy": 0.7655525088310242,
"num_tokens": 64714.0,
"step": 20
},
{
"epoch": 0.4,
"grad_norm": 0.39404281973838806,
"learning_rate": 5.2631578947368424e-05,
"loss": 0.5951,
"mean_token_accuracy": 0.8334747165441513,
"num_tokens": 96352.0,
"step": 30
},
{
"epoch": 0.5333333333333333,
"grad_norm": 0.41063451766967773,
"learning_rate": 7.017543859649122e-05,
"loss": 0.4305,
"mean_token_accuracy": 0.8707121074199676,
"num_tokens": 125717.0,
"step": 40
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.3393928110599518,
"learning_rate": 8.771929824561403e-05,
"loss": 0.3874,
"mean_token_accuracy": 0.8828857451677322,
"num_tokens": 157446.0,
"step": 50
},
{
"epoch": 0.8,
"grad_norm": 0.2939395606517792,
"learning_rate": 9.905660377358492e-05,
"loss": 0.3613,
"mean_token_accuracy": 0.8902347713708878,
"num_tokens": 187947.0,
"step": 60
},
{
"epoch": 0.9333333333333333,
"grad_norm": 0.30780646204948425,
"learning_rate": 9.59119496855346e-05,
"loss": 0.3545,
"mean_token_accuracy": 0.8908359378576278,
"num_tokens": 219528.0,
"step": 70
},
{
"epoch": 1.0,
"eval_loss": 0.35836032032966614,
"eval_mean_token_accuracy": 0.8900894594192504,
"eval_num_tokens": 234547.0,
"eval_runtime": 47.5564,
"eval_samples_per_second": 2.103,
"eval_steps_per_second": 0.526,
"step": 75
}
],
"logging_steps": 10,
"max_steps": 375,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"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": 1.3215858968223744e+16,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}