|
{ |
|
"best_metric": 0.4311259388923645, |
|
"best_model_checkpoint": "mikhail-panzo/fil_b64_le5_s8000/checkpoint-1500", |
|
"epoch": 65.21739130434783, |
|
"eval_steps": 500, |
|
"global_step": 1500, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 2.1739130434782608, |
|
"grad_norm": 4.140745639801025, |
|
"learning_rate": 2.4500000000000004e-07, |
|
"loss": 0.8272, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 4.3478260869565215, |
|
"grad_norm": 1.78262197971344, |
|
"learning_rate": 4.95e-07, |
|
"loss": 0.7463, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 6.521739130434782, |
|
"grad_norm": 4.001489639282227, |
|
"learning_rate": 7.450000000000001e-07, |
|
"loss": 0.7504, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 8.695652173913043, |
|
"grad_norm": 1.0013818740844727, |
|
"learning_rate": 9.950000000000002e-07, |
|
"loss": 0.7295, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 10.869565217391305, |
|
"grad_norm": 1.9014511108398438, |
|
"learning_rate": 1.2450000000000002e-06, |
|
"loss": 0.7182, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 13.043478260869565, |
|
"grad_norm": 13.358899116516113, |
|
"learning_rate": 1.495e-06, |
|
"loss": 0.7037, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 15.217391304347826, |
|
"grad_norm": 1.4039109945297241, |
|
"learning_rate": 1.745e-06, |
|
"loss": 0.6966, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 17.391304347826086, |
|
"grad_norm": 3.0226972103118896, |
|
"learning_rate": 1.9950000000000004e-06, |
|
"loss": 0.6702, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 19.565217391304348, |
|
"grad_norm": 1.716300129890442, |
|
"learning_rate": 2.24e-06, |
|
"loss": 0.6487, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 21.73913043478261, |
|
"grad_norm": 2.9351744651794434, |
|
"learning_rate": 2.4900000000000003e-06, |
|
"loss": 0.6071, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 21.73913043478261, |
|
"eval_loss": 0.521294116973877, |
|
"eval_runtime": 6.6451, |
|
"eval_samples_per_second": 23.927, |
|
"eval_steps_per_second": 3.01, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 23.91304347826087, |
|
"grad_norm": 1.9992552995681763, |
|
"learning_rate": 2.7400000000000004e-06, |
|
"loss": 0.5759, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 26.08695652173913, |
|
"grad_norm": 1.2594470977783203, |
|
"learning_rate": 2.99e-06, |
|
"loss": 0.56, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 28.26086956521739, |
|
"grad_norm": 1.0602363348007202, |
|
"learning_rate": 3.2400000000000003e-06, |
|
"loss": 0.5709, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 30.434782608695652, |
|
"grad_norm": 1.116231083869934, |
|
"learning_rate": 3.49e-06, |
|
"loss": 0.5322, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 32.608695652173914, |
|
"grad_norm": 1.5203620195388794, |
|
"learning_rate": 3.74e-06, |
|
"loss": 0.5264, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 34.78260869565217, |
|
"grad_norm": 1.6139357089996338, |
|
"learning_rate": 3.990000000000001e-06, |
|
"loss": 0.5191, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 36.95652173913044, |
|
"grad_norm": 1.436273217201233, |
|
"learning_rate": 4.24e-06, |
|
"loss": 0.5272, |
|
"step": 850 |
|
}, |
|
{ |
|
"epoch": 39.130434782608695, |
|
"grad_norm": 1.757323145866394, |
|
"learning_rate": 4.49e-06, |
|
"loss": 0.5016, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 41.30434782608695, |
|
"grad_norm": 0.9879368543624878, |
|
"learning_rate": 4.74e-06, |
|
"loss": 0.4975, |
|
"step": 950 |
|
}, |
|
{ |
|
"epoch": 43.47826086956522, |
|
"grad_norm": 1.3402222394943237, |
|
"learning_rate": 4.9900000000000005e-06, |
|
"loss": 0.5126, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 43.47826086956522, |
|
"eval_loss": 0.4506620466709137, |
|
"eval_runtime": 6.4644, |
|
"eval_samples_per_second": 24.596, |
|
"eval_steps_per_second": 3.094, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 45.65217391304348, |
|
"grad_norm": 1.9366440773010254, |
|
"learning_rate": 5.240000000000001e-06, |
|
"loss": 0.5043, |
|
"step": 1050 |
|
}, |
|
{ |
|
"epoch": 47.82608695652174, |
|
"grad_norm": 1.5358505249023438, |
|
"learning_rate": 5.490000000000001e-06, |
|
"loss": 0.5034, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 50.0, |
|
"grad_norm": 1.6836323738098145, |
|
"learning_rate": 5.74e-06, |
|
"loss": 0.499, |
|
"step": 1150 |
|
}, |
|
{ |
|
"epoch": 52.17391304347826, |
|
"grad_norm": 1.5092076063156128, |
|
"learning_rate": 5.99e-06, |
|
"loss": 0.487, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 54.34782608695652, |
|
"grad_norm": 1.0751235485076904, |
|
"learning_rate": 6.24e-06, |
|
"loss": 0.4871, |
|
"step": 1250 |
|
}, |
|
{ |
|
"epoch": 56.52173913043478, |
|
"grad_norm": 1.327528476715088, |
|
"learning_rate": 6.4900000000000005e-06, |
|
"loss": 0.4911, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 58.69565217391305, |
|
"grad_norm": 1.7827174663543701, |
|
"learning_rate": 6.740000000000001e-06, |
|
"loss": 0.4772, |
|
"step": 1350 |
|
}, |
|
{ |
|
"epoch": 60.869565217391305, |
|
"grad_norm": 1.2707558870315552, |
|
"learning_rate": 6.99e-06, |
|
"loss": 0.4786, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 63.04347826086956, |
|
"grad_norm": 1.0794321298599243, |
|
"learning_rate": 7.24e-06, |
|
"loss": 0.4741, |
|
"step": 1450 |
|
}, |
|
{ |
|
"epoch": 65.21739130434783, |
|
"grad_norm": 2.0180580615997314, |
|
"learning_rate": 7.49e-06, |
|
"loss": 0.4749, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 65.21739130434783, |
|
"eval_loss": 0.4311259388923645, |
|
"eval_runtime": 6.4494, |
|
"eval_samples_per_second": 24.653, |
|
"eval_steps_per_second": 3.101, |
|
"step": 1500 |
|
} |
|
], |
|
"logging_steps": 50, |
|
"max_steps": 8000, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 348, |
|
"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": 2.226780270809741e+16, |
|
"train_batch_size": 64, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|