Files
Qwen_SLM_Reasoning-Model/trainer_state.json
ModelHub XC efcc43e905 初始化项目,由ModelHub XC社区提供模型
Model: Prithwiraj731/Qwen_SLM_Reasoning-Model
Source: Original Platform
2026-04-21 18:32:33 +08:00

51 lines
1.1 KiB
JSON

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 9,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.8,
"grad_norm": 0.49833059310913086,
"learning_rate": 8e-05,
"loss": 2.501791763305664,
"step": 5
},
{
"epoch": 3.0,
"step": 9,
"total_flos": 135178931404800.0,
"train_loss": 2.427226278516981,
"train_runtime": 33.6069,
"train_samples_per_second": 1.785,
"train_steps_per_second": 0.268
}
],
"logging_steps": 5,
"max_steps": 9,
"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": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 135178931404800.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}