trainer_state.json

{
 "best_metric": null,
 "best_model_checkpoint": null,
 "epoch": 50.0,
 "global_step": 7400,
 "is_hyper_param_search": false,
 "is_local_process_zero": true,
 "is_world_process_zero": true,
 "log_history": [
 {
 "epoch": 3.38,
 "learning_rate": 4.662162162162162e-05,
 "loss": 3.9918,
 "step": 500
 },
 {
 "epoch": 6.76,
 "learning_rate": 4.324324324324325e-05,
 "loss": 3.6317,
 "step": 1000
 },
 {
 "epoch": 10.14,
 "learning_rate": 3.986486486486487e-05,
 "loss": 3.4414,
 "step": 1500
 },
 {
 "epoch": 13.51,
 "learning_rate": 3.648648648648649e-05,
 "loss": 3.2874,
 "step": 2000
 },
 {
 "epoch": 16.89,
 "learning_rate": 3.310810810810811e-05,
 "loss": 3.1699,
 "step": 2500
 },
 {
 "epoch": 20.27,
 "learning_rate": 2.9729729729729733e-05,
 "loss": 3.0613,
 "step": 3000
 },
 {
 "epoch": 23.65,
 "learning_rate": 2.635135135135135e-05,
 "loss": 2.9674,
 "step": 3500
 },
 {
 "epoch": 27.03,
 "learning_rate": 2.2972972972972976e-05,
 "loss": 2.8905,
 "step": 4000
 },
 {
 "epoch": 30.41,
 "learning_rate": 1.9594594594594595e-05,
 "loss": 2.819,
 "step": 4500
 },
 {
 "epoch": 33.78,
 "learning_rate": 1.6216216216216218e-05,
 "loss": 2.7674,
 "step": 5000
 },
 {
 "epoch": 37.16,
 "learning_rate": 1.2837837837837838e-05,
 "loss": 2.7203,
 "step": 5500
 },
 {
 "epoch": 40.54,
 "learning_rate": 9.45945945945946e-06,
 "loss": 2.6839,
 "step": 6000
 },
 {
 "epoch": 43.92,
 "learning_rate": 6.081081081081082e-06,
 "loss": 2.6523,
 "step": 6500
 },
 {
 "epoch": 47.3,
 "learning_rate": 2.702702702702703e-06,
 "loss": 2.6353,
 "step": 7000
 },
 {
 "epoch": 50.0,
 "step": 7400,
 "total_flos": 1.48213270315008e+16,
 "train_runtime": 2224.787,
 "train_samples_per_second": 3.326
 }
 ],
 "max_steps": 7400,
 "num_train_epochs": 50,
 "total_flos": 1.48213270315008e+16,
 "trial_name": null,
 "trial_params": null
}