213 lines
5.5 KiB
JSON
213 lines
5.5 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": 2.8274196604015085,
|
|
"learning_rate": 2e-05,
|
|
"loss": 0.422,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.13153071701526642,
|
|
"step": 5,
|
|
"valid_targets_mean": 13794.3,
|
|
"valid_targets_min": 7848
|
|
},
|
|
{
|
|
"epoch": 0.9375,
|
|
"grad_norm": 0.6788499642545621,
|
|
"learning_rate": 3.9979273472934556e-05,
|
|
"loss": 0.3467,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.10864634811878204,
|
|
"step": 10,
|
|
"valid_targets_mean": 14449.4,
|
|
"valid_targets_min": 8164
|
|
},
|
|
{
|
|
"epoch": 1.375,
|
|
"grad_norm": 0.3305783167470126,
|
|
"learning_rate": 3.925834574695599e-05,
|
|
"loss": 0.3089,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.09869768470525742,
|
|
"step": 15,
|
|
"valid_targets_mean": 13925.9,
|
|
"valid_targets_min": 9503
|
|
},
|
|
{
|
|
"epoch": 1.84375,
|
|
"grad_norm": 0.24414082828506373,
|
|
"learning_rate": 3.754365204805189e-05,
|
|
"loss": 0.2889,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.08986084163188934,
|
|
"step": 20,
|
|
"valid_targets_mean": 13211.0,
|
|
"valid_targets_min": 5976
|
|
},
|
|
{
|
|
"epoch": 2.28125,
|
|
"grad_norm": 0.1871479801692981,
|
|
"learning_rate": 3.4923673716953717e-05,
|
|
"loss": 0.2639,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.08749322593212128,
|
|
"step": 25,
|
|
"valid_targets_mean": 14030.9,
|
|
"valid_targets_min": 6036
|
|
},
|
|
{
|
|
"epoch": 2.75,
|
|
"grad_norm": 0.1741112049371486,
|
|
"learning_rate": 3.153360644229735e-05,
|
|
"loss": 0.2441,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.07998865842819214,
|
|
"step": 30,
|
|
"valid_targets_mean": 14805.3,
|
|
"valid_targets_min": 9784
|
|
},
|
|
{
|
|
"epoch": 3.1875,
|
|
"grad_norm": 0.15024766068678258,
|
|
"learning_rate": 2.7548383915052287e-05,
|
|
"loss": 0.2406,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.08076930791139603,
|
|
"step": 35,
|
|
"valid_targets_mean": 14291.8,
|
|
"valid_targets_min": 7913
|
|
},
|
|
{
|
|
"epoch": 3.65625,
|
|
"grad_norm": 0.13765463973479342,
|
|
"learning_rate": 2.317365092912503e-05,
|
|
"loss": 0.2271,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.07251711189746857,
|
|
"step": 40,
|
|
"valid_targets_mean": 14492.5,
|
|
"valid_targets_min": 7853
|
|
},
|
|
{
|
|
"epoch": 4.09375,
|
|
"grad_norm": 0.14044311527289763,
|
|
"learning_rate": 1.8635151732706586e-05,
|
|
"loss": 0.225,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.07657542824745178,
|
|
"step": 45,
|
|
"valid_targets_mean": 14069.8,
|
|
"valid_targets_min": 5516
|
|
},
|
|
{
|
|
"epoch": 4.5625,
|
|
"grad_norm": 0.11766963528888084,
|
|
"learning_rate": 1.4167081210903501e-05,
|
|
"loss": 0.216,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.06715231388807297,
|
|
"step": 50,
|
|
"valid_targets_mean": 14619.7,
|
|
"valid_targets_min": 9401
|
|
},
|
|
{
|
|
"epoch": 5.0,
|
|
"grad_norm": 0.1564461173013084,
|
|
"learning_rate": 1.0000000000000006e-05,
|
|
"loss": 0.2168,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.10675714910030365,
|
|
"step": 55,
|
|
"valid_targets_mean": 13425.2,
|
|
"valid_targets_min": 6387
|
|
},
|
|
{
|
|
"epoch": 5.46875,
|
|
"grad_norm": 0.13399235229390016,
|
|
"learning_rate": 6.348937135626922e-06,
|
|
"loss": 0.2129,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.06979517638683319,
|
|
"step": 60,
|
|
"valid_targets_mean": 14429.8,
|
|
"valid_targets_min": 3326
|
|
},
|
|
{
|
|
"epoch": 5.9375,
|
|
"grad_norm": 0.14157691568897496,
|
|
"learning_rate": 3.402294160019499e-06,
|
|
"loss": 0.2098,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.07008929550647736,
|
|
"step": 65,
|
|
"valid_targets_mean": 14258.1,
|
|
"valid_targets_min": 3977
|
|
},
|
|
{
|
|
"epoch": 6.375,
|
|
"grad_norm": 0.1307087732739413,
|
|
"learning_rate": 1.3121232567865793e-06,
|
|
"loss": 0.2121,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.06869778037071228,
|
|
"step": 70,
|
|
"valid_targets_mean": 13733.4,
|
|
"valid_targets_min": 6025
|
|
},
|
|
{
|
|
"epoch": 6.84375,
|
|
"grad_norm": 0.11156698536027285,
|
|
"learning_rate": 1.862810792733849e-07,
|
|
"loss": 0.2096,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.07188820093870163,
|
|
"step": 75,
|
|
"valid_targets_mean": 14534.2,
|
|
"valid_targets_min": 9307
|
|
},
|
|
{
|
|
"epoch": 7.0,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.10573843121528625,
|
|
"step": 77,
|
|
"total_flos": 5.7154397037972685e+17,
|
|
"train_loss": 0.2553465281214033,
|
|
"train_runtime": 2318.0705,
|
|
"train_samples_per_second": 3.02,
|
|
"train_steps_per_second": 0.033,
|
|
"valid_targets_mean": 13469.1,
|
|
"valid_targets_min": 4791
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 77,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 7,
|
|
"save_steps": 300,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 5.7154397037972685e+17,
|
|
"train_batch_size": 1,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|