trainer_state.json

{
 "best_metric": null,
 "best_model_checkpoint": null,
 "epoch": 3.8095238095238093,
 "eval_steps": 500,
 "global_step": 10,
 "is_hyper_param_search": false,
 "is_local_process_zero": true,
 "is_world_process_zero": true,
 "log_history": [
 {
 "epoch": 0.38,
 "learning_rate": 0.0,
 "loss": 3.0607,
 "step": 1
 },
 {
 "epoch": 0.76,
 "learning_rate": 0.0002,
 "loss": 2.9844,
 "step": 2
 },
 {
 "epoch": 1.14,
 "learning_rate": 0.0002,
 "loss": 3.0551,
 "step": 3
 },
 {
 "epoch": 1.52,
 "learning_rate": 0.0002,
 "loss": 2.8438,
 "step": 4
 },
 {
 "epoch": 1.9,
 "learning_rate": 0.0002,
 "loss": 2.7193,
 "step": 5
 },
 {
 "epoch": 2.29,
 "learning_rate": 0.0002,
 "loss": 2.6074,
 "step": 6
 },
 {
 "epoch": 2.67,
 "learning_rate": 0.0002,
 "loss": 2.5219,
 "step": 7
 },
 {
 "epoch": 3.05,
 "learning_rate": 0.0002,
 "loss": 2.4533,
 "step": 8
 },
 {
 "epoch": 3.43,
 "learning_rate": 0.0002,
 "loss": 2.3453,
 "step": 9
 },
 {
 "epoch": 3.81,
 "learning_rate": 0.0002,
 "loss": 2.2306,
 "step": 10
 },
 {
 "epoch": 3.81,
 "step": 10,
 "total_flos": 8222269440000.0,
 "train_loss": 2.682184100151062,
 "train_runtime": 273.7695,
 "train_samples_per_second": 1.516,
 "train_steps_per_second": 0.037
 }
 ],
 "logging_steps": 1.0,
 "max_steps": 10,
 "num_input_tokens_seen": 0,
 "num_train_epochs": 5,
 "save_steps": 10,
 "total_flos": 8222269440000.0,
 "train_batch_size": 1,
 "trial_name": null,
 "trial_params": null
}