adapters-opt-bf16-QLORA-super_glue-copa
/
trainer_state-opt-fp16-QLORA-super_glue-copa-sequence_classification.json
{ | |
"best_metric": null, | |
"best_model_checkpoint": null, | |
"epoch": 8.0, | |
"eval_steps": 1, | |
"global_step": 20, | |
"is_hyper_param_search": false, | |
"is_local_process_zero": true, | |
"is_world_process_zero": true, | |
"log_history": [ | |
{ | |
"epoch": 0.4, | |
"grad_norm": 9.25, | |
"learning_rate": 2.5e-05, | |
"loss": 0.7505, | |
"step": 1 | |
}, | |
{ | |
"epoch": 0.4, | |
"eval_accuracy": 0.46, | |
"eval_loss": 0.7447265386581421, | |
"eval_runtime": 0.6628, | |
"eval_samples_per_second": 150.871, | |
"eval_steps_per_second": 4.526, | |
"step": 1 | |
}, | |
{ | |
"epoch": 0.8, | |
"grad_norm": 7.59375, | |
"learning_rate": 5e-05, | |
"loss": 0.7291, | |
"step": 2 | |
}, | |
{ | |
"epoch": 0.8, | |
"eval_accuracy": 0.45, | |
"eval_loss": 0.7388477921485901, | |
"eval_runtime": 0.6576, | |
"eval_samples_per_second": 152.059, | |
"eval_steps_per_second": 4.562, | |
"step": 2 | |
}, | |
{ | |
"epoch": 1.2, | |
"grad_norm": 8.5, | |
"learning_rate": 4.722222222222222e-05, | |
"loss": 0.7301, | |
"step": 3 | |
}, | |
{ | |
"epoch": 1.2, | |
"eval_accuracy": 0.47, | |
"eval_loss": 0.7335740923881531, | |
"eval_runtime": 0.7079, | |
"eval_samples_per_second": 141.262, | |
"eval_steps_per_second": 4.238, | |
"step": 3 | |
}, | |
{ | |
"epoch": 1.6, | |
"grad_norm": 2.34375, | |
"learning_rate": 4.4444444444444447e-05, | |
"loss": 0.6936, | |
"step": 4 | |
}, | |
{ | |
"epoch": 1.6, | |
"eval_accuracy": 0.47, | |
"eval_loss": 0.7240138053894043, | |
"eval_runtime": 0.6067, | |
"eval_samples_per_second": 164.816, | |
"eval_steps_per_second": 4.944, | |
"step": 4 | |
}, | |
{ | |
"epoch": 2.0, | |
"grad_norm": 2.328125, | |
"learning_rate": 4.166666666666667e-05, | |
"loss": 0.6858, | |
"step": 5 | |
}, | |
{ | |
"epoch": 2.0, | |
"eval_accuracy": 0.45, | |
"eval_loss": 0.7132813334465027, | |
"eval_runtime": 0.7063, | |
"eval_samples_per_second": 141.591, | |
"eval_steps_per_second": 4.248, | |
"step": 5 | |
}, | |
{ | |
"epoch": 2.4, | |
"grad_norm": 3.96875, | |
"learning_rate": 3.888888888888889e-05, | |
"loss": 0.71, | |
"step": 6 | |
}, | |
{ | |
"epoch": 2.4, | |
"eval_accuracy": 0.44, | |
"eval_loss": 0.7067969441413879, | |
"eval_runtime": 0.7275, | |
"eval_samples_per_second": 137.448, | |
"eval_steps_per_second": 4.123, | |
"step": 6 | |
}, | |
{ | |
"epoch": 2.8, | |
"grad_norm": 3.359375, | |
"learning_rate": 3.611111111111111e-05, | |
"loss": 0.6762, | |
"step": 7 | |
}, | |
{ | |
"epoch": 2.8, | |
"eval_accuracy": 0.44, | |
"eval_loss": 0.7037889361381531, | |
"eval_runtime": 0.6054, | |
"eval_samples_per_second": 165.192, | |
"eval_steps_per_second": 4.956, | |
"step": 7 | |
}, | |
{ | |
"epoch": 3.2, | |
"grad_norm": 2.578125, | |
"learning_rate": 3.3333333333333335e-05, | |
"loss": 0.6753, | |
"step": 8 | |
}, | |
{ | |
"epoch": 3.2, | |
"eval_accuracy": 0.46, | |
"eval_loss": 0.704345703125, | |
"eval_runtime": 0.606, | |
"eval_samples_per_second": 165.006, | |
"eval_steps_per_second": 4.95, | |
"step": 8 | |
}, | |
{ | |
"epoch": 3.6, | |
"grad_norm": 7.5, | |
"learning_rate": 3.055555555555556e-05, | |
"loss": 0.7018, | |
"step": 9 | |
}, | |
{ | |
"epoch": 3.6, | |
"eval_accuracy": 0.47, | |
"eval_loss": 0.7046388387680054, | |
"eval_runtime": 0.6548, | |
"eval_samples_per_second": 152.715, | |
"eval_steps_per_second": 4.581, | |
"step": 9 | |
}, | |
{ | |
"epoch": 4.0, | |
"grad_norm": 2.78125, | |
"learning_rate": 2.777777777777778e-05, | |
"loss": 0.7055, | |
"step": 10 | |
}, | |
{ | |
"epoch": 4.0, | |
"eval_accuracy": 0.47, | |
"eval_loss": 0.7057226300239563, | |
"eval_runtime": 0.607, | |
"eval_samples_per_second": 164.734, | |
"eval_steps_per_second": 4.942, | |
"step": 10 | |
}, | |
{ | |
"epoch": 4.4, | |
"grad_norm": 2.1875, | |
"learning_rate": 2.5e-05, | |
"loss": 0.7051, | |
"step": 11 | |
}, | |
{ | |
"epoch": 4.4, | |
"eval_accuracy": 0.47, | |
"eval_loss": 0.7075486779212952, | |
"eval_runtime": 0.6555, | |
"eval_samples_per_second": 152.549, | |
"eval_steps_per_second": 4.576, | |
"step": 11 | |
}, | |
{ | |
"epoch": 4.8, | |
"grad_norm": 1.5859375, | |
"learning_rate": 2.2222222222222223e-05, | |
"loss": 0.702, | |
"step": 12 | |
}, | |
{ | |
"epoch": 4.8, | |
"eval_accuracy": 0.47, | |
"eval_loss": 0.7065821886062622, | |
"eval_runtime": 0.6165, | |
"eval_samples_per_second": 162.207, | |
"eval_steps_per_second": 4.866, | |
"step": 12 | |
}, | |
{ | |
"epoch": 5.2, | |
"grad_norm": 4.03125, | |
"learning_rate": 1.9444444444444445e-05, | |
"loss": 0.681, | |
"step": 13 | |
}, | |
{ | |
"epoch": 5.2, | |
"eval_accuracy": 0.47, | |
"eval_loss": 0.7069336175918579, | |
"eval_runtime": 0.605, | |
"eval_samples_per_second": 165.291, | |
"eval_steps_per_second": 4.959, | |
"step": 13 | |
}, | |
{ | |
"epoch": 5.6, | |
"grad_norm": 5.1875, | |
"learning_rate": 1.6666666666666667e-05, | |
"loss": 0.7026, | |
"step": 14 | |
}, | |
{ | |
"epoch": 5.6, | |
"eval_accuracy": 0.47, | |
"eval_loss": 0.7063087224960327, | |
"eval_runtime": 0.666, | |
"eval_samples_per_second": 150.145, | |
"eval_steps_per_second": 4.504, | |
"step": 14 | |
}, | |
{ | |
"epoch": 6.0, | |
"grad_norm": 4.46875, | |
"learning_rate": 1.388888888888889e-05, | |
"loss": 0.7073, | |
"step": 15 | |
}, | |
{ | |
"epoch": 6.0, | |
"eval_accuracy": 0.47, | |
"eval_loss": 0.705986499786377, | |
"eval_runtime": 0.7202, | |
"eval_samples_per_second": 138.849, | |
"eval_steps_per_second": 4.165, | |
"step": 15 | |
}, | |
{ | |
"epoch": 6.4, | |
"grad_norm": 2.4375, | |
"learning_rate": 1.1111111111111112e-05, | |
"loss": 0.7072, | |
"step": 16 | |
}, | |
{ | |
"epoch": 6.4, | |
"eval_accuracy": 0.47, | |
"eval_loss": 0.7048340439796448, | |
"eval_runtime": 0.7767, | |
"eval_samples_per_second": 128.745, | |
"eval_steps_per_second": 3.862, | |
"step": 16 | |
}, | |
{ | |
"epoch": 6.8, | |
"grad_norm": 2.09375, | |
"learning_rate": 8.333333333333334e-06, | |
"loss": 0.6854, | |
"step": 17 | |
}, | |
{ | |
"epoch": 6.8, | |
"eval_accuracy": 0.47, | |
"eval_loss": 0.705859363079071, | |
"eval_runtime": 0.7695, | |
"eval_samples_per_second": 129.949, | |
"eval_steps_per_second": 3.898, | |
"step": 17 | |
}, | |
{ | |
"epoch": 7.2, | |
"grad_norm": 4.375, | |
"learning_rate": 5.555555555555556e-06, | |
"loss": 0.6735, | |
"step": 18 | |
}, | |
{ | |
"epoch": 7.2, | |
"eval_accuracy": 0.47, | |
"eval_loss": 0.7044140696525574, | |
"eval_runtime": 0.6581, | |
"eval_samples_per_second": 151.945, | |
"eval_steps_per_second": 4.558, | |
"step": 18 | |
}, | |
{ | |
"epoch": 7.6, | |
"grad_norm": 7.0625, | |
"learning_rate": 2.777777777777778e-06, | |
"loss": 0.6894, | |
"step": 19 | |
}, | |
{ | |
"epoch": 7.6, | |
"eval_accuracy": 0.47, | |
"eval_loss": 0.7049902081489563, | |
"eval_runtime": 0.6094, | |
"eval_samples_per_second": 164.108, | |
"eval_steps_per_second": 4.923, | |
"step": 19 | |
}, | |
{ | |
"epoch": 8.0, | |
"grad_norm": 4.84375, | |
"learning_rate": 0.0, | |
"loss": 0.6885, | |
"step": 20 | |
}, | |
{ | |
"epoch": 8.0, | |
"eval_accuracy": 0.47, | |
"eval_loss": 0.704902172088623, | |
"eval_runtime": 0.6067, | |
"eval_samples_per_second": 164.835, | |
"eval_steps_per_second": 4.945, | |
"step": 20 | |
}, | |
{ | |
"epoch": 8.0, | |
"step": 20, | |
"total_flos": 6418587971485696.0, | |
"train_loss": 0.6999931305646896, | |
"train_runtime": 64.2055, | |
"train_samples_per_second": 62.3, | |
"train_steps_per_second": 0.311 | |
} | |
], | |
"logging_steps": 1, | |
"max_steps": 20, | |
"num_input_tokens_seen": 0, | |
"num_train_epochs": 10, | |
"save_steps": 500, | |
"stateful_callbacks": { | |
"TrainerControl": { | |
"args": { | |
"should_epoch_stop": false, | |
"should_evaluate": false, | |
"should_log": false, | |
"should_save": false, | |
"should_training_stop": false | |
}, | |
"attributes": {} | |
} | |
}, | |
"total_flos": 6418587971485696.0, | |
"train_batch_size": 10, | |
"trial_name": null, | |
"trial_params": null | |
} | |