trainer_state.json

{
 "best_metric": null,
 "best_model_checkpoint": null,
 "epoch": 0.9977431999049768,
 "global_step": 800,
 "is_hyper_param_search": false,
 "is_local_process_zero": true,
 "is_world_process_zero": true,
 "log_history": [
 {
 "epoch": 0.25,
 "eval_loss": 1.9490387439727783,
 "eval_runtime": 1630.042,
 "eval_samples_per_second": 40.118,
 "eval_steps_per_second": 0.836,
 "step": 200
 },
 {
 "epoch": 0.5,
 "eval_loss": 1.9350184202194214,
 "eval_runtime": 1608.0969,
 "eval_samples_per_second": 40.665,
 "eval_steps_per_second": 0.848,
 "step": 400
 },
 {
 "epoch": 0.62,
 "learning_rate": 0.0001,
 "loss": 1.0689,
 "step": 500
 },
 {
 "epoch": 0.75,
 "eval_loss": 1.8705594539642334,
 "eval_runtime": 1614.7147,
 "eval_samples_per_second": 40.499,
 "eval_steps_per_second": 0.844,
 "step": 600
 },
 {
 "epoch": 1.0,
 "eval_loss": 1.70987069606781,
 "eval_runtime": 1610.6806,
 "eval_samples_per_second": 40.6,
 "eval_steps_per_second": 0.846,
 "step": 800
 }
 ],
 "max_steps": 801,
 "num_train_epochs": 1,
 "total_flos": 2081757562142720.0,
 "trial_name": null,
 "trial_params": null
}