430 lines
9.7 KiB
JSON
430 lines
9.7 KiB
JSON
|
{
|
||
|
|
"best_global_step": null,
|
||
|
|
"best_metric": null,
|
||
|
|
"best_model_checkpoint": null,
|
||
|
|
"epoch": 1.0,
|
||
|
|
"eval_steps": 10,
|
||
|
|
"global_step": 50,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 0.020151133501259445,
|
||
|
|
"grad_norm": 200.0,
|
||
|
|
"learning_rate": 0.0,
|
||
|
|
"loss": 11.0118,
|
||
|
|
"step": 1
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.04030226700251889,
|
||
|
|
"grad_norm": 191.0,
|
||
|
|
"learning_rate": 2.0000000000000003e-06,
|
||
|
|
"loss": 11.6099,
|
||
|
|
"step": 2
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.060453400503778336,
|
||
|
|
"grad_norm": 193.0,
|
||
|
|
"learning_rate": 4.000000000000001e-06,
|
||
|
|
"loss": 11.119,
|
||
|
|
"step": 3
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08060453400503778,
|
||
|
|
"grad_norm": 187.0,
|
||
|
|
"learning_rate": 6e-06,
|
||
|
|
"loss": 10.1865,
|
||
|
|
"step": 4
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10075566750629723,
|
||
|
|
"grad_norm": 159.0,
|
||
|
|
"learning_rate": 8.000000000000001e-06,
|
||
|
|
"loss": 8.6306,
|
||
|
|
"step": 5
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12090680100755667,
|
||
|
|
"grad_norm": 130.0,
|
||
|
|
"learning_rate": 1e-05,
|
||
|
|
"loss": 6.7238,
|
||
|
|
"step": 6
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14105793450881612,
|
||
|
|
"grad_norm": 99.5,
|
||
|
|
"learning_rate": 9.777777777777779e-06,
|
||
|
|
"loss": 4.568,
|
||
|
|
"step": 7
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16120906801007556,
|
||
|
|
"grad_norm": 62.25,
|
||
|
|
"learning_rate": 9.555555555555556e-06,
|
||
|
|
"loss": 3.7907,
|
||
|
|
"step": 8
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.181360201511335,
|
||
|
|
"grad_norm": 38.5,
|
||
|
|
"learning_rate": 9.333333333333334e-06,
|
||
|
|
"loss": 2.9669,
|
||
|
|
"step": 9
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20151133501259447,
|
||
|
|
"grad_norm": 36.0,
|
||
|
|
"learning_rate": 9.111111111111112e-06,
|
||
|
|
"loss": 3.1071,
|
||
|
|
"step": 10
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20151133501259447,
|
||
|
|
"eval_loss": 2.249746799468994,
|
||
|
|
"eval_model_preparation_time": 0.0215,
|
||
|
|
"eval_runtime": 2.9615,
|
||
|
|
"eval_samples_per_second": 32.078,
|
||
|
|
"eval_steps_per_second": 16.208,
|
||
|
|
"step": 10
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2216624685138539,
|
||
|
|
"grad_norm": 29.125,
|
||
|
|
"learning_rate": 8.888888888888888e-06,
|
||
|
|
"loss": 2.2238,
|
||
|
|
"step": 11
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24181360201511334,
|
||
|
|
"grad_norm": 27.125,
|
||
|
|
"learning_rate": 8.666666666666668e-06,
|
||
|
|
"loss": 1.978,
|
||
|
|
"step": 12
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2619647355163728,
|
||
|
|
"grad_norm": 21.625,
|
||
|
|
"learning_rate": 8.444444444444446e-06,
|
||
|
|
"loss": 1.7079,
|
||
|
|
"step": 13
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28211586901763225,
|
||
|
|
"grad_norm": 22.5,
|
||
|
|
"learning_rate": 8.222222222222222e-06,
|
||
|
|
"loss": 1.7878,
|
||
|
|
"step": 14
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3022670025188917,
|
||
|
|
"grad_norm": 15.8125,
|
||
|
|
"learning_rate": 8.000000000000001e-06,
|
||
|
|
"loss": 1.1862,
|
||
|
|
"step": 15
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3224181360201511,
|
||
|
|
"grad_norm": 13.9375,
|
||
|
|
"learning_rate": 7.77777777777778e-06,
|
||
|
|
"loss": 1.2518,
|
||
|
|
"step": 16
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3425692695214106,
|
||
|
|
"grad_norm": 17.25,
|
||
|
|
"learning_rate": 7.555555555555556e-06,
|
||
|
|
"loss": 1.107,
|
||
|
|
"step": 17
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.36272040302267,
|
||
|
|
"grad_norm": 13.5625,
|
||
|
|
"learning_rate": 7.333333333333333e-06,
|
||
|
|
"loss": 0.9791,
|
||
|
|
"step": 18
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38287153652392947,
|
||
|
|
"grad_norm": 14.875,
|
||
|
|
"learning_rate": 7.111111111111112e-06,
|
||
|
|
"loss": 1.1575,
|
||
|
|
"step": 19
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.40302267002518893,
|
||
|
|
"grad_norm": 14.9375,
|
||
|
|
"learning_rate": 6.88888888888889e-06,
|
||
|
|
"loss": 0.818,
|
||
|
|
"step": 20
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.40302267002518893,
|
||
|
|
"eval_loss": 0.7283493876457214,
|
||
|
|
"eval_model_preparation_time": 0.0215,
|
||
|
|
"eval_runtime": 3.0654,
|
||
|
|
"eval_samples_per_second": 30.991,
|
||
|
|
"eval_steps_per_second": 15.659,
|
||
|
|
"step": 20
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.42317380352644834,
|
||
|
|
"grad_norm": 13.75,
|
||
|
|
"learning_rate": 6.666666666666667e-06,
|
||
|
|
"loss": 0.785,
|
||
|
|
"step": 21
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4433249370277078,
|
||
|
|
"grad_norm": 14.9375,
|
||
|
|
"learning_rate": 6.444444444444445e-06,
|
||
|
|
"loss": 0.6213,
|
||
|
|
"step": 22
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4634760705289673,
|
||
|
|
"grad_norm": 18.375,
|
||
|
|
"learning_rate": 6.222222222222223e-06,
|
||
|
|
"loss": 0.6943,
|
||
|
|
"step": 23
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4836272040302267,
|
||
|
|
"grad_norm": 14.3125,
|
||
|
|
"learning_rate": 6e-06,
|
||
|
|
"loss": 0.5944,
|
||
|
|
"step": 24
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5037783375314862,
|
||
|
|
"grad_norm": 10.6875,
|
||
|
|
"learning_rate": 5.777777777777778e-06,
|
||
|
|
"loss": 0.5768,
|
||
|
|
"step": 25
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5239294710327456,
|
||
|
|
"grad_norm": 14.8125,
|
||
|
|
"learning_rate": 5.555555555555557e-06,
|
||
|
|
"loss": 0.6205,
|
||
|
|
"step": 26
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5440806045340051,
|
||
|
|
"grad_norm": 10.25,
|
||
|
|
"learning_rate": 5.333333333333334e-06,
|
||
|
|
"loss": 0.5576,
|
||
|
|
"step": 27
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5642317380352645,
|
||
|
|
"grad_norm": 8.6875,
|
||
|
|
"learning_rate": 5.1111111111111115e-06,
|
||
|
|
"loss": 0.585,
|
||
|
|
"step": 28
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5843828715365239,
|
||
|
|
"grad_norm": 9.3125,
|
||
|
|
"learning_rate": 4.888888888888889e-06,
|
||
|
|
"loss": 0.4842,
|
||
|
|
"step": 29
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6045340050377834,
|
||
|
|
"grad_norm": 9.625,
|
||
|
|
"learning_rate": 4.666666666666667e-06,
|
||
|
|
"loss": 0.5162,
|
||
|
|
"step": 30
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6045340050377834,
|
||
|
|
"eval_loss": 0.5592977404594421,
|
||
|
|
"eval_model_preparation_time": 0.0215,
|
||
|
|
"eval_runtime": 3.3756,
|
||
|
|
"eval_samples_per_second": 28.143,
|
||
|
|
"eval_steps_per_second": 14.22,
|
||
|
|
"step": 30
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6246851385390428,
|
||
|
|
"grad_norm": 6.9375,
|
||
|
|
"learning_rate": 4.444444444444444e-06,
|
||
|
|
"loss": 0.4891,
|
||
|
|
"step": 31
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6448362720403022,
|
||
|
|
"grad_norm": 13.5,
|
||
|
|
"learning_rate": 4.222222222222223e-06,
|
||
|
|
"loss": 0.4537,
|
||
|
|
"step": 32
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6649874055415617,
|
||
|
|
"grad_norm": 12.5625,
|
||
|
|
"learning_rate": 4.000000000000001e-06,
|
||
|
|
"loss": 0.5064,
|
||
|
|
"step": 33
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6851385390428212,
|
||
|
|
"grad_norm": 14.375,
|
||
|
|
"learning_rate": 3.777777777777778e-06,
|
||
|
|
"loss": 0.5586,
|
||
|
|
"step": 34
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7052896725440806,
|
||
|
|
"grad_norm": 7.9375,
|
||
|
|
"learning_rate": 3.555555555555556e-06,
|
||
|
|
"loss": 0.4467,
|
||
|
|
"step": 35
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.72544080604534,
|
||
|
|
"grad_norm": 11.3125,
|
||
|
|
"learning_rate": 3.3333333333333333e-06,
|
||
|
|
"loss": 0.4768,
|
||
|
|
"step": 36
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7455919395465995,
|
||
|
|
"grad_norm": 9.0,
|
||
|
|
"learning_rate": 3.1111111111111116e-06,
|
||
|
|
"loss": 0.5006,
|
||
|
|
"step": 37
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7657430730478589,
|
||
|
|
"grad_norm": 9.6875,
|
||
|
|
"learning_rate": 2.888888888888889e-06,
|
||
|
|
"loss": 0.4835,
|
||
|
|
"step": 38
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7858942065491183,
|
||
|
|
"grad_norm": 6.5625,
|
||
|
|
"learning_rate": 2.666666666666667e-06,
|
||
|
|
"loss": 0.4814,
|
||
|
|
"step": 39
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8060453400503779,
|
||
|
|
"grad_norm": 6.375,
|
||
|
|
"learning_rate": 2.4444444444444447e-06,
|
||
|
|
"loss": 0.3869,
|
||
|
|
"step": 40
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8060453400503779,
|
||
|
|
"eval_loss": 0.5029374957084656,
|
||
|
|
"eval_model_preparation_time": 0.0215,
|
||
|
|
"eval_runtime": 2.8351,
|
||
|
|
"eval_samples_per_second": 33.508,
|
||
|
|
"eval_steps_per_second": 16.931,
|
||
|
|
"step": 40
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8261964735516373,
|
||
|
|
"grad_norm": 8.3125,
|
||
|
|
"learning_rate": 2.222222222222222e-06,
|
||
|
|
"loss": 0.5224,
|
||
|
|
"step": 41
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8463476070528967,
|
||
|
|
"grad_norm": 7.625,
|
||
|
|
"learning_rate": 2.0000000000000003e-06,
|
||
|
|
"loss": 0.4716,
|
||
|
|
"step": 42
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8664987405541562,
|
||
|
|
"grad_norm": 5.6875,
|
||
|
|
"learning_rate": 1.777777777777778e-06,
|
||
|
|
"loss": 0.3948,
|
||
|
|
"step": 43
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8866498740554156,
|
||
|
|
"grad_norm": 7.5625,
|
||
|
|
"learning_rate": 1.5555555555555558e-06,
|
||
|
|
"loss": 0.4883,
|
||
|
|
"step": 44
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.906801007556675,
|
||
|
|
"grad_norm": 7.375,
|
||
|
|
"learning_rate": 1.3333333333333334e-06,
|
||
|
|
"loss": 0.4354,
|
||
|
|
"step": 45
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9269521410579346,
|
||
|
|
"grad_norm": 8.375,
|
||
|
|
"learning_rate": 1.111111111111111e-06,
|
||
|
|
"loss": 0.4541,
|
||
|
|
"step": 46
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.947103274559194,
|
||
|
|
"grad_norm": 7.5,
|
||
|
|
"learning_rate": 8.88888888888889e-07,
|
||
|
|
"loss": 0.4197,
|
||
|
|
"step": 47
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9672544080604534,
|
||
|
|
"grad_norm": 6.0,
|
||
|
|
"learning_rate": 6.666666666666667e-07,
|
||
|
|
"loss": 0.4558,
|
||
|
|
"step": 48
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9874055415617129,
|
||
|
|
"grad_norm": 6.09375,
|
||
|
|
"learning_rate": 4.444444444444445e-07,
|
||
|
|
"loss": 0.4534,
|
||
|
|
"step": 49
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0,
|
||
|
|
"grad_norm": 7.65625,
|
||
|
|
"learning_rate": 2.2222222222222224e-07,
|
||
|
|
"loss": 0.4926,
|
||
|
|
"step": 50
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0,
|
||
|
|
"eval_loss": 0.49504080414772034,
|
||
|
|
"eval_model_preparation_time": 0.0215,
|
||
|
|
"eval_runtime": 3.2667,
|
||
|
|
"eval_samples_per_second": 29.082,
|
||
|
|
"eval_steps_per_second": 14.694,
|
||
|
|
"step": 50
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 1,
|
||
|
|
"max_steps": 50,
|
||
|
|
"num_input_tokens_seen": 0,
|
||
|
|
"num_train_epochs": 1,
|
||
|
|
"save_steps": 5000,
|
||
|
|
"stateful_callbacks": {
|
||
|
|
"TrainerControl": {
|
||
|
|
"args": {
|
||
|
|
"should_epoch_stop": false,
|
||
|
|
"should_evaluate": false,
|
||
|
|
"should_log": false,
|
||
|
|
"should_save": true,
|
||
|
|
"should_training_stop": true
|
||
|
|
},
|
||
|
|
"attributes": {}
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"total_flos": 1033611294953472.0,
|
||
|
|
"train_batch_size": 2,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|