| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 7467, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2008838891120932, | |
| "grad_norm": 3.7622594833374023, | |
| "learning_rate": 1.1860881974423135e-05, | |
| "loss": 0.3148, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4017677782241864, | |
| "grad_norm": 2.0420267581939697, | |
| "learning_rate": 1.1364403030328403e-05, | |
| "loss": 0.1783, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6026516673362796, | |
| "grad_norm": 5.3519697189331055, | |
| "learning_rate": 1.0867924086233672e-05, | |
| "loss": 0.1565, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8035355564483728, | |
| "grad_norm": 3.6524667739868164, | |
| "learning_rate": 1.037144514213894e-05, | |
| "loss": 0.1528, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9531602968973307, | |
| "eval_f1": 0.7930374238468234, | |
| "eval_loss": 0.12813745439052582, | |
| "eval_precision": 0.732947232947233, | |
| "eval_recall": 0.8638604474781949, | |
| "eval_runtime": 24.1379, | |
| "eval_samples_per_second": 176.735, | |
| "eval_steps_per_second": 22.123, | |
| "step": 2489 | |
| }, | |
| { | |
| "epoch": 1.004419445560466, | |
| "grad_norm": 2.0776000022888184, | |
| "learning_rate": 9.874966198044209e-06, | |
| "loss": 0.1338, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2053033346725592, | |
| "grad_norm": 1.6920737028121948, | |
| "learning_rate": 9.378487253949478e-06, | |
| "loss": 0.1009, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4061872237846524, | |
| "grad_norm": 0.6098091006278992, | |
| "learning_rate": 8.882008309854746e-06, | |
| "loss": 0.1044, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6070711128967456, | |
| "grad_norm": 5.741237640380859, | |
| "learning_rate": 8.385529365760015e-06, | |
| "loss": 0.1001, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8079550020088389, | |
| "grad_norm": 1.5740350484848022, | |
| "learning_rate": 7.889050421665283e-06, | |
| "loss": 0.0944, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9635200906234975, | |
| "eval_f1": 0.8353142997181718, | |
| "eval_loss": 0.10682477056980133, | |
| "eval_precision": 0.810486267982404, | |
| "eval_recall": 0.8617115408924283, | |
| "eval_runtime": 24.1496, | |
| "eval_samples_per_second": 176.649, | |
| "eval_steps_per_second": 22.112, | |
| "step": 4978 | |
| }, | |
| { | |
| "epoch": 2.008838891120932, | |
| "grad_norm": 1.4677562713623047, | |
| "learning_rate": 7.392571477570552e-06, | |
| "loss": 0.0933, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.2097227802330255, | |
| "grad_norm": 6.143584251403809, | |
| "learning_rate": 6.89609253347582e-06, | |
| "loss": 0.0666, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.4106066693451185, | |
| "grad_norm": 1.6898033618927002, | |
| "learning_rate": 6.399613589381089e-06, | |
| "loss": 0.0659, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.611490558457212, | |
| "grad_norm": 6.529832363128662, | |
| "learning_rate": 5.903134645286357e-06, | |
| "loss": 0.0685, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.812374447569305, | |
| "grad_norm": 6.338042736053467, | |
| "learning_rate": 5.4066557011916255e-06, | |
| "loss": 0.0659, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9666155153223522, | |
| "eval_f1": 0.8455699277342564, | |
| "eval_loss": 0.10882943123579025, | |
| "eval_precision": 0.81369798971482, | |
| "eval_recall": 0.8800404500063204, | |
| "eval_runtime": 22.3687, | |
| "eval_samples_per_second": 190.713, | |
| "eval_steps_per_second": 23.873, | |
| "step": 7467 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 12445, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1975400912140920.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 1.2357360918517866e-05, | |
| "num_train_epochs": 5, | |
| "per_device_train_batch_size": 8, | |
| "seed": 35 | |
| } | |
| } | |