trainer_state.json

{
 "best_metric": null,
 "best_model_checkpoint": null,
 "epoch": 5.0,
 "eval_steps": 500,
 "global_step": 7430,
 "is_hyper_param_search": false,
 "is_local_process_zero": true,
 "is_world_process_zero": true,
 "log_history": [
 {
 "epoch": 0.3364737550471063,
 "grad_norm": 0.923697829246521,
 "learning_rate": 0.00046635262449528937,
 "loss": 1.9724,
 "step": 500
 },
 {
 "epoch": 0.6729475100942126,
 "grad_norm": 0.9243040084838867,
 "learning_rate": 0.0004327052489905787,
 "loss": 1.4423,
 "step": 1000
 },
 {
 "epoch": 1.0,
 "eval_loss": 1.0461400747299194,
 "eval_runtime": 36.2574,
 "eval_samples_per_second": 81.942,
 "eval_steps_per_second": 10.26,
 "step": 1486
 },
 {
 "epoch": 1.009421265141319,
 "grad_norm": 0.9555139541625977,
 "learning_rate": 0.0003990578734858681,
 "loss": 1.2854,
 "step": 1500
 },
 {
 "epoch": 1.3458950201884252,
 "grad_norm": 0.8507774472236633,
 "learning_rate": 0.0003654104979811575,
 "loss": 0.9929,
 "step": 2000
 },
 {
 "epoch": 1.6823687752355316,
 "grad_norm": 1.1206731796264648,
 "learning_rate": 0.00033176312247644685,
 "loss": 0.9408,
 "step": 2500
 },
 {
 "epoch": 2.0,
 "eval_loss": 0.9026183485984802,
 "eval_runtime": 36.3158,
 "eval_samples_per_second": 81.81,
 "eval_steps_per_second": 10.243,
 "step": 2972
 },
 {
 "epoch": 2.018842530282638,
 "grad_norm": 0.7318525910377502,
 "learning_rate": 0.0002981157469717362,
 "loss": 0.8886,
 "step": 3000
 },
 {
 "epoch": 2.3553162853297445,
 "grad_norm": 1.1639642715454102,
 "learning_rate": 0.00026446837146702556,
 "loss": 0.6969,
 "step": 3500
 },
 {
 "epoch": 2.6917900403768504,
 "grad_norm": 0.7347049117088318,
 "learning_rate": 0.00023082099596231497,
 "loss": 0.692,
 "step": 4000
 },
 {
 "epoch": 3.0,
 "eval_loss": 0.8661695122718811,
 "eval_runtime": 36.5469,
 "eval_samples_per_second": 81.293,
 "eval_steps_per_second": 10.179,
 "step": 4458
 },
 {
 "epoch": 3.028263795423957,
 "grad_norm": 0.7746924757957458,
 "learning_rate": 0.00019717362045760433,
 "loss": 0.6564,
 "step": 4500
 },
 {
 "epoch": 3.3647375504710633,
 "grad_norm": 0.7316901087760925,
 "learning_rate": 0.00016352624495289368,
 "loss": 0.4934,
 "step": 5000
 },
 {
 "epoch": 3.7012113055181697,
 "grad_norm": 0.9040531516075134,
 "learning_rate": 0.00012987886944818307,
 "loss": 0.5261,
 "step": 5500
 },
 {
 "epoch": 4.0,
 "eval_loss": 0.8571327924728394,
 "eval_runtime": 36.3399,
 "eval_samples_per_second": 81.756,
 "eval_steps_per_second": 10.237,
 "step": 5944
 },
 {
 "epoch": 4.037685060565276,
 "grad_norm": 0.9058707356452942,
 "learning_rate": 9.623149394347241e-05,
 "loss": 0.4785,
 "step": 6000
 },
 {
 "epoch": 4.3741588156123825,
 "grad_norm": 0.7362410426139832,
 "learning_rate": 6.258411843876178e-05,
 "loss": 0.3714,
 "step": 6500
 },
 {
 "epoch": 4.710632570659489,
 "grad_norm": 0.6890231370925903,
 "learning_rate": 2.8936742934051144e-05,
 "loss": 0.3846,
 "step": 7000
 }
 ],
 "logging_steps": 500,
 "max_steps": 7430,
 "num_input_tokens_seen": 0,
 "num_train_epochs": 5,
 "save_steps": 10000,
 "stateful_callbacks": {
 "TrainerControl": {
 "args": {
 "should_epoch_stop": false,
 "should_evaluate": false,
 "should_log": false,
 "should_save": true,
 "should_training_stop": true
 },
 "attributes": {}
 }
 },
 "total_flos": 7016439606285312.0,
 "train_batch_size": 8,
 "trial_name": null,
 "trial_params": null
}