| { | |
| "best_metric": 0.19211259484291077, | |
| "best_model_checkpoint": "./beans_outputs/checkpoint-165", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 165, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.30303030303030304, | |
| "grad_norm": 1.740174412727356, | |
| "learning_rate": 1.8787878787878792e-05, | |
| "loss": 1.0024, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 1.4151499271392822, | |
| "learning_rate": 1.7575757575757576e-05, | |
| "loss": 0.8405, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 1.3221044540405273, | |
| "learning_rate": 1.6363636363636366e-05, | |
| "loss": 0.7212, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9624060150375939, | |
| "eval_loss": 0.5817888379096985, | |
| "eval_runtime": 1.1163, | |
| "eval_samples_per_second": 119.146, | |
| "eval_steps_per_second": 4.479, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.2121212121212122, | |
| "grad_norm": 1.3216980695724487, | |
| "learning_rate": 1.5151515151515153e-05, | |
| "loss": 0.5934, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5151515151515151, | |
| "grad_norm": 1.5411862134933472, | |
| "learning_rate": 1.3939393939393942e-05, | |
| "loss": 0.5163, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 1.4787555932998657, | |
| "learning_rate": 1.2727272727272728e-05, | |
| "loss": 0.4355, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9774436090225563, | |
| "eval_loss": 0.3262147307395935, | |
| "eval_runtime": 0.9678, | |
| "eval_samples_per_second": 137.421, | |
| "eval_steps_per_second": 5.166, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.121212121212121, | |
| "grad_norm": 1.2977631092071533, | |
| "learning_rate": 1.1515151515151517e-05, | |
| "loss": 0.3642, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.4242424242424243, | |
| "grad_norm": 2.0464847087860107, | |
| "learning_rate": 1.0303030303030304e-05, | |
| "loss": 0.3188, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 1.2298272848129272, | |
| "learning_rate": 9.090909090909091e-06, | |
| "loss": 0.3134, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9699248120300752, | |
| "eval_loss": 0.23712603747844696, | |
| "eval_runtime": 1.0214, | |
| "eval_samples_per_second": 130.21, | |
| "eval_steps_per_second": 4.895, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.0303030303030303, | |
| "grad_norm": 1.2645708322525024, | |
| "learning_rate": 7.87878787878788e-06, | |
| "loss": 0.2831, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 1.1787359714508057, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.2644, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 2.6098392009735107, | |
| "learning_rate": 5.4545454545454545e-06, | |
| "loss": 0.2457, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.9393939393939394, | |
| "grad_norm": 1.1771011352539062, | |
| "learning_rate": 4.242424242424243e-06, | |
| "loss": 0.2207, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9774436090225563, | |
| "eval_loss": 0.19545504450798035, | |
| "eval_runtime": 0.9619, | |
| "eval_samples_per_second": 138.27, | |
| "eval_steps_per_second": 5.198, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 4.242424242424242, | |
| "grad_norm": 0.6839556097984314, | |
| "learning_rate": 3.0303030303030305e-06, | |
| "loss": 0.2271, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 4.545454545454545, | |
| "grad_norm": 1.096548318862915, | |
| "learning_rate": 1.8181818181818183e-06, | |
| "loss": 0.2074, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 4.848484848484849, | |
| "grad_norm": 0.5964490175247192, | |
| "learning_rate": 6.060606060606061e-07, | |
| "loss": 0.2149, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9699248120300752, | |
| "eval_loss": 0.19211259484291077, | |
| "eval_runtime": 1.0109, | |
| "eval_samples_per_second": 131.57, | |
| "eval_steps_per_second": 4.946, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 165, | |
| "total_flos": 4.006371770595533e+17, | |
| "train_loss": 0.0, | |
| "train_runtime": 0.1246, | |
| "train_samples_per_second": 41491.687, | |
| "train_steps_per_second": 1324.203 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 165, | |
| "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": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.006371770595533e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |