209 lines
5.4 KiB
JSON
209 lines
5.4 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 7.0,
|
|
"eval_steps": 500,
|
|
"global_step": 77,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.46875,
|
|
"grad_norm": 1.7063705735800943,
|
|
"learning_rate": 2e-05,
|
|
"loss": 0.3477,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.0856635719537735,
|
|
"step": 5,
|
|
"valid_targets_mean": 4223.8,
|
|
"valid_targets_min": 1764
|
|
},
|
|
{
|
|
"epoch": 0.9375,
|
|
"grad_norm": 0.3174963435320657,
|
|
"learning_rate": 3.9979273472934556e-05,
|
|
"loss": 0.2066,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.056782372295856476,
|
|
"step": 10,
|
|
"valid_targets_mean": 3980.9,
|
|
"valid_targets_min": 1525
|
|
},
|
|
{
|
|
"epoch": 1.375,
|
|
"grad_norm": 0.19940629712969432,
|
|
"learning_rate": 3.925834574695599e-05,
|
|
"loss": 0.1441,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.04278196394443512,
|
|
"step": 15,
|
|
"valid_targets_mean": 3834.6,
|
|
"valid_targets_min": 2180
|
|
},
|
|
{
|
|
"epoch": 1.84375,
|
|
"grad_norm": 0.16723919345138785,
|
|
"learning_rate": 3.754365204805189e-05,
|
|
"loss": 0.1333,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.037383563816547394,
|
|
"step": 20,
|
|
"valid_targets_mean": 3959.7,
|
|
"valid_targets_min": 1891
|
|
},
|
|
{
|
|
"epoch": 2.28125,
|
|
"grad_norm": 0.21085194266590884,
|
|
"learning_rate": 3.4923673716953717e-05,
|
|
"loss": 0.108,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.03128531947731972,
|
|
"step": 25,
|
|
"valid_targets_mean": 3111.5,
|
|
"valid_targets_min": 1912
|
|
},
|
|
{
|
|
"epoch": 2.75,
|
|
"grad_norm": 0.19698767350301571,
|
|
"learning_rate": 3.153360644229735e-05,
|
|
"loss": 0.0902,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.029110876843333244,
|
|
"step": 30,
|
|
"valid_targets_mean": 3842.9,
|
|
"valid_targets_min": 1877
|
|
},
|
|
{
|
|
"epoch": 3.1875,
|
|
"grad_norm": 0.21338454966150336,
|
|
"learning_rate": 2.7548383915052287e-05,
|
|
"loss": 0.0792,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.025294136255979538,
|
|
"step": 35,
|
|
"valid_targets_mean": 3693.0,
|
|
"valid_targets_min": 1891
|
|
},
|
|
{
|
|
"epoch": 3.65625,
|
|
"grad_norm": 0.20227319700357463,
|
|
"learning_rate": 2.317365092912503e-05,
|
|
"loss": 0.0594,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.015729054808616638,
|
|
"step": 40,
|
|
"valid_targets_mean": 4400.2,
|
|
"valid_targets_min": 1921
|
|
},
|
|
{
|
|
"epoch": 4.09375,
|
|
"grad_norm": 0.188311500063878,
|
|
"learning_rate": 1.8635151732706586e-05,
|
|
"loss": 0.0577,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.013538211584091187,
|
|
"step": 45,
|
|
"valid_targets_mean": 3079.8,
|
|
"valid_targets_min": 1942
|
|
},
|
|
{
|
|
"epoch": 4.5625,
|
|
"grad_norm": 0.1953658413119939,
|
|
"learning_rate": 1.4167081210903501e-05,
|
|
"loss": 0.0415,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.015336105599999428,
|
|
"step": 50,
|
|
"valid_targets_mean": 4404.6,
|
|
"valid_targets_min": 1929
|
|
},
|
|
{
|
|
"epoch": 5.0,
|
|
"grad_norm": 0.23684064615294104,
|
|
"learning_rate": 1.0000000000000006e-05,
|
|
"loss": 0.041,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.019942108541727066,
|
|
"step": 55,
|
|
"valid_targets_mean": 4031.6,
|
|
"valid_targets_min": 1907
|
|
},
|
|
{
|
|
"epoch": 5.46875,
|
|
"grad_norm": 0.1969832814002483,
|
|
"learning_rate": 6.348937135626922e-06,
|
|
"loss": 0.0294,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.008506393060088158,
|
|
"step": 60,
|
|
"valid_targets_mean": 4410.8,
|
|
"valid_targets_min": 1900
|
|
},
|
|
{
|
|
"epoch": 5.9375,
|
|
"grad_norm": 0.18836987621083745,
|
|
"learning_rate": 3.402294160019499e-06,
|
|
"loss": 0.0304,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.011393106542527676,
|
|
"step": 65,
|
|
"valid_targets_mean": 3430.5,
|
|
"valid_targets_min": 1907
|
|
},
|
|
{
|
|
"epoch": 6.375,
|
|
"grad_norm": 0.13487516265968516,
|
|
"learning_rate": 1.3121232567865793e-06,
|
|
"loss": 0.0259,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.007582570891827345,
|
|
"step": 70,
|
|
"valid_targets_mean": 3291.6,
|
|
"valid_targets_min": 1904
|
|
},
|
|
{
|
|
"epoch": 6.84375,
|
|
"grad_norm": 0.1432999963460923,
|
|
"learning_rate": 1.862810792733849e-07,
|
|
"loss": 0.0248,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.00842520035803318,
|
|
"step": 75,
|
|
"valid_targets_mean": 3544.7,
|
|
"valid_targets_min": 1889
|
|
},
|
|
{
|
|
"epoch": 7.0,
|
|
"step": 77,
|
|
"total_flos": 9.529203217976525e+16,
|
|
"train_loss": 0.0,
|
|
"train_runtime": 0.8525,
|
|
"train_samples_per_second": 8211.578,
|
|
"train_steps_per_second": 90.327
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 77,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 7,
|
|
"save_steps": 50,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 9.529203217976525e+16,
|
|
"train_batch_size": 1,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|