|
{ |
|
"best_metric": 0.9010989010989011, |
|
"best_model_checkpoint": "ktp-kk-crop/checkpoint-10", |
|
"epoch": 25.0, |
|
"eval_steps": 500, |
|
"global_step": 50, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 1.0, |
|
"eval_accuracy": 0.6153846153846154, |
|
"eval_loss": 0.6706601977348328, |
|
"eval_runtime": 4.349, |
|
"eval_samples_per_second": 20.924, |
|
"eval_steps_per_second": 1.38, |
|
"step": 2 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.6153846153846154, |
|
"eval_loss": 0.6923023462295532, |
|
"eval_runtime": 3.6322, |
|
"eval_samples_per_second": 25.053, |
|
"eval_steps_per_second": 1.652, |
|
"step": 4 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_accuracy": 0.7032967032967034, |
|
"eval_loss": 0.7208701372146606, |
|
"eval_runtime": 3.5659, |
|
"eval_samples_per_second": 25.52, |
|
"eval_steps_per_second": 1.683, |
|
"step": 6 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.8021978021978022, |
|
"eval_loss": 0.4788942337036133, |
|
"eval_runtime": 3.617, |
|
"eval_samples_per_second": 25.159, |
|
"eval_steps_per_second": 1.659, |
|
"step": 8 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"eval_accuracy": 0.9010989010989011, |
|
"eval_loss": 0.3554236590862274, |
|
"eval_runtime": 3.6003, |
|
"eval_samples_per_second": 25.275, |
|
"eval_steps_per_second": 1.667, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_accuracy": 0.8241758241758241, |
|
"eval_loss": 0.8337829113006592, |
|
"eval_runtime": 3.6007, |
|
"eval_samples_per_second": 25.273, |
|
"eval_steps_per_second": 1.666, |
|
"step": 12 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"eval_accuracy": 0.7912087912087912, |
|
"eval_loss": 1.073412537574768, |
|
"eval_runtime": 3.6715, |
|
"eval_samples_per_second": 24.786, |
|
"eval_steps_per_second": 1.634, |
|
"step": 14 |
|
}, |
|
{ |
|
"epoch": 7.5, |
|
"grad_norm": 38.24232864379883, |
|
"learning_rate": 3.888888888888889e-05, |
|
"loss": 0.2476, |
|
"step": 15 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_accuracy": 0.8241758241758241, |
|
"eval_loss": 1.0161423683166504, |
|
"eval_runtime": 3.6667, |
|
"eval_samples_per_second": 24.818, |
|
"eval_steps_per_second": 1.636, |
|
"step": 16 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"eval_accuracy": 0.8571428571428571, |
|
"eval_loss": 0.8966610431671143, |
|
"eval_runtime": 3.6608, |
|
"eval_samples_per_second": 24.858, |
|
"eval_steps_per_second": 1.639, |
|
"step": 18 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"eval_accuracy": 0.8681318681318682, |
|
"eval_loss": 0.8477412462234497, |
|
"eval_runtime": 3.4998, |
|
"eval_samples_per_second": 26.001, |
|
"eval_steps_per_second": 1.714, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 11.0, |
|
"eval_accuracy": 0.8351648351648352, |
|
"eval_loss": 1.3332176208496094, |
|
"eval_runtime": 3.6008, |
|
"eval_samples_per_second": 25.272, |
|
"eval_steps_per_second": 1.666, |
|
"step": 22 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"eval_accuracy": 0.7252747252747253, |
|
"eval_loss": 3.477609395980835, |
|
"eval_runtime": 3.6298, |
|
"eval_samples_per_second": 25.071, |
|
"eval_steps_per_second": 1.653, |
|
"step": 24 |
|
}, |
|
{ |
|
"epoch": 13.0, |
|
"eval_accuracy": 0.6373626373626373, |
|
"eval_loss": 6.0097527503967285, |
|
"eval_runtime": 3.6304, |
|
"eval_samples_per_second": 25.066, |
|
"eval_steps_per_second": 1.653, |
|
"step": 26 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"eval_accuracy": 0.7252747252747253, |
|
"eval_loss": 3.8649513721466064, |
|
"eval_runtime": 3.5214, |
|
"eval_samples_per_second": 25.842, |
|
"eval_steps_per_second": 1.704, |
|
"step": 28 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"grad_norm": 0.0022278681863099337, |
|
"learning_rate": 2.2222222222222223e-05, |
|
"loss": 0.0277, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"eval_accuracy": 0.8131868131868132, |
|
"eval_loss": 2.2409703731536865, |
|
"eval_runtime": 3.6708, |
|
"eval_samples_per_second": 24.79, |
|
"eval_steps_per_second": 1.635, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"eval_accuracy": 0.8131868131868132, |
|
"eval_loss": 1.45852792263031, |
|
"eval_runtime": 3.7348, |
|
"eval_samples_per_second": 24.365, |
|
"eval_steps_per_second": 1.607, |
|
"step": 32 |
|
}, |
|
{ |
|
"epoch": 17.0, |
|
"eval_accuracy": 0.8021978021978022, |
|
"eval_loss": 1.4789769649505615, |
|
"eval_runtime": 3.7258, |
|
"eval_samples_per_second": 24.424, |
|
"eval_steps_per_second": 1.61, |
|
"step": 34 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"eval_accuracy": 0.8241758241758241, |
|
"eval_loss": 1.8427152633666992, |
|
"eval_runtime": 3.6118, |
|
"eval_samples_per_second": 25.195, |
|
"eval_steps_per_second": 1.661, |
|
"step": 36 |
|
}, |
|
{ |
|
"epoch": 19.0, |
|
"eval_accuracy": 0.8241758241758241, |
|
"eval_loss": 2.0873501300811768, |
|
"eval_runtime": 3.6114, |
|
"eval_samples_per_second": 25.198, |
|
"eval_steps_per_second": 1.661, |
|
"step": 38 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"eval_accuracy": 0.8131868131868132, |
|
"eval_loss": 2.245572090148926, |
|
"eval_runtime": 3.651, |
|
"eval_samples_per_second": 24.924, |
|
"eval_steps_per_second": 1.643, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 21.0, |
|
"eval_accuracy": 0.8021978021978022, |
|
"eval_loss": 2.366342782974243, |
|
"eval_runtime": 3.6311, |
|
"eval_samples_per_second": 25.061, |
|
"eval_steps_per_second": 1.652, |
|
"step": 42 |
|
}, |
|
{ |
|
"epoch": 22.0, |
|
"eval_accuracy": 0.8021978021978022, |
|
"eval_loss": 2.4484336376190186, |
|
"eval_runtime": 3.6882, |
|
"eval_samples_per_second": 24.673, |
|
"eval_steps_per_second": 1.627, |
|
"step": 44 |
|
}, |
|
{ |
|
"epoch": 22.5, |
|
"grad_norm": 0.0005057471571490169, |
|
"learning_rate": 5.555555555555556e-06, |
|
"loss": 0.0001, |
|
"step": 45 |
|
}, |
|
{ |
|
"epoch": 23.0, |
|
"eval_accuracy": 0.8021978021978022, |
|
"eval_loss": 2.4973790645599365, |
|
"eval_runtime": 3.6432, |
|
"eval_samples_per_second": 24.978, |
|
"eval_steps_per_second": 1.647, |
|
"step": 46 |
|
}, |
|
{ |
|
"epoch": 24.0, |
|
"eval_accuracy": 0.8021978021978022, |
|
"eval_loss": 2.5220789909362793, |
|
"eval_runtime": 3.6924, |
|
"eval_samples_per_second": 24.645, |
|
"eval_steps_per_second": 1.625, |
|
"step": 48 |
|
}, |
|
{ |
|
"epoch": 25.0, |
|
"eval_accuracy": 0.8021978021978022, |
|
"eval_loss": 2.530649185180664, |
|
"eval_runtime": 3.7584, |
|
"eval_samples_per_second": 24.212, |
|
"eval_steps_per_second": 1.596, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 25.0, |
|
"step": 50, |
|
"total_flos": 2.427840666897408e+17, |
|
"train_loss": 0.08263867741210561, |
|
"train_runtime": 466.1075, |
|
"train_samples_per_second": 6.597, |
|
"train_steps_per_second": 0.107 |
|
} |
|
], |
|
"logging_steps": 15, |
|
"max_steps": 50, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 25, |
|
"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": 2.427840666897408e+17, |
|
"train_batch_size": 16, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|