307 lines
7.1 KiB
JSON
307 lines
7.1 KiB
JSON
|
{
|
||
|
|
"best_global_step": null,
|
||
|
|
"best_metric": null,
|
||
|
|
"best_model_checkpoint": null,
|
||
|
|
"epoch": 4.0,
|
||
|
|
"eval_steps": 100,
|
||
|
|
"global_step": 644,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 0.12422360248447205,
|
||
|
|
"grad_norm": 15.3125,
|
||
|
|
"learning_rate": 7.755102040816327e-06,
|
||
|
|
"loss": 3.1538,
|
||
|
|
"step": 20
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2484472049689441,
|
||
|
|
"grad_norm": 4.21875,
|
||
|
|
"learning_rate": 1.5918367346938776e-05,
|
||
|
|
"loss": 0.3551,
|
||
|
|
"step": 40
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37267080745341613,
|
||
|
|
"grad_norm": 4.4375,
|
||
|
|
"learning_rate": 1.9994132020337718e-05,
|
||
|
|
"loss": 0.3427,
|
||
|
|
"step": 60
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4968944099378882,
|
||
|
|
"grad_norm": 3.703125,
|
||
|
|
"learning_rate": 1.99472294947797e-05,
|
||
|
|
"loss": 0.3533,
|
||
|
|
"step": 80
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6211180124223602,
|
||
|
|
"grad_norm": 4.28125,
|
||
|
|
"learning_rate": 1.98536445575164e-05,
|
||
|
|
"loss": 0.3258,
|
||
|
|
"step": 100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6211180124223602,
|
||
|
|
"eval_loss": 0.3044624626636505,
|
||
|
|
"eval_runtime": 46.1922,
|
||
|
|
"eval_samples_per_second": 14.007,
|
||
|
|
"eval_steps_per_second": 7.014,
|
||
|
|
"step": 100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7453416149068323,
|
||
|
|
"grad_norm": 4.34375,
|
||
|
|
"learning_rate": 1.971381640325756e-05,
|
||
|
|
"loss": 0.3598,
|
||
|
|
"step": 120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8695652173913043,
|
||
|
|
"grad_norm": 2.875,
|
||
|
|
"learning_rate": 1.952840124642636e-05,
|
||
|
|
"loss": 0.3572,
|
||
|
|
"step": 140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9937888198757764,
|
||
|
|
"grad_norm": 3.359375,
|
||
|
|
"learning_rate": 1.9298269241540908e-05,
|
||
|
|
"loss": 0.3251,
|
||
|
|
"step": 160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1180124223602483,
|
||
|
|
"grad_norm": 2.875,
|
||
|
|
"learning_rate": 1.9024500399573536e-05,
|
||
|
|
"loss": 0.3263,
|
||
|
|
"step": 180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2422360248447206,
|
||
|
|
"grad_norm": 3.515625,
|
||
|
|
"learning_rate": 1.8708379519452372e-05,
|
||
|
|
"loss": 0.326,
|
||
|
|
"step": 200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2422360248447206,
|
||
|
|
"eval_loss": 0.2889421284198761,
|
||
|
|
"eval_runtime": 46.0702,
|
||
|
|
"eval_samples_per_second": 14.044,
|
||
|
|
"eval_steps_per_second": 7.033,
|
||
|
|
"step": 200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3664596273291925,
|
||
|
|
"grad_norm": 2.4375,
|
||
|
|
"learning_rate": 1.835139015849177e-05,
|
||
|
|
"loss": 0.3128,
|
||
|
|
"step": 220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4906832298136645,
|
||
|
|
"grad_norm": 2.65625,
|
||
|
|
"learning_rate": 1.7955207670048306e-05,
|
||
|
|
"loss": 0.3101,
|
||
|
|
"step": 240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6149068322981366,
|
||
|
|
"grad_norm": 4.0625,
|
||
|
|
"learning_rate": 1.7521691341076774e-05,
|
||
|
|
"loss": 0.347,
|
||
|
|
"step": 260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7391304347826086,
|
||
|
|
"grad_norm": 2.984375,
|
||
|
|
"learning_rate": 1.7052875666484584e-05,
|
||
|
|
"loss": 0.3063,
|
||
|
|
"step": 280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8633540372670807,
|
||
|
|
"grad_norm": 3.015625,
|
||
|
|
"learning_rate": 1.6550960801234152e-05,
|
||
|
|
"loss": 0.3115,
|
||
|
|
"step": 300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8633540372670807,
|
||
|
|
"eval_loss": 0.2829993963241577,
|
||
|
|
"eval_runtime": 46.2888,
|
||
|
|
"eval_samples_per_second": 13.977,
|
||
|
|
"eval_steps_per_second": 7.0,
|
||
|
|
"step": 300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9875776397515528,
|
||
|
|
"grad_norm": 2.46875,
|
||
|
|
"learning_rate": 1.601830223500156e-05,
|
||
|
|
"loss": 0.3225,
|
||
|
|
"step": 320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.111801242236025,
|
||
|
|
"grad_norm": 4.40625,
|
||
|
|
"learning_rate": 1.5457399737848445e-05,
|
||
|
|
"loss": 0.2985,
|
||
|
|
"step": 340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.2360248447204967,
|
||
|
|
"grad_norm": 2.828125,
|
||
|
|
"learning_rate": 1.4870885628785075e-05,
|
||
|
|
"loss": 0.2862,
|
||
|
|
"step": 360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.360248447204969,
|
||
|
|
"grad_norm": 2.40625,
|
||
|
|
"learning_rate": 1.426151242228039e-05,
|
||
|
|
"loss": 0.2951,
|
||
|
|
"step": 380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4844720496894412,
|
||
|
|
"grad_norm": 3.015625,
|
||
|
|
"learning_rate": 1.3632139910693971e-05,
|
||
|
|
"loss": 0.2743,
|
||
|
|
"step": 400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4844720496894412,
|
||
|
|
"eval_loss": 0.27418315410614014,
|
||
|
|
"eval_runtime": 46.3638,
|
||
|
|
"eval_samples_per_second": 13.955,
|
||
|
|
"eval_steps_per_second": 6.988,
|
||
|
|
"step": 400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.608695652173913,
|
||
|
|
"grad_norm": 2.28125,
|
||
|
|
"learning_rate": 1.2985721743252206e-05,
|
||
|
|
"loss": 0.3003,
|
||
|
|
"step": 420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.732919254658385,
|
||
|
|
"grad_norm": 2.53125,
|
||
|
|
"learning_rate": 1.232529156455353e-05,
|
||
|
|
"loss": 0.272,
|
||
|
|
"step": 440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.857142857142857,
|
||
|
|
"grad_norm": 2.796875,
|
||
|
|
"learning_rate": 1.1653948777654925e-05,
|
||
|
|
"loss": 0.2959,
|
||
|
|
"step": 460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.981366459627329,
|
||
|
|
"grad_norm": 2.390625,
|
||
|
|
"learning_rate": 1.097484399855352e-05,
|
||
|
|
"loss": 0.2921,
|
||
|
|
"step": 480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1055900621118013,
|
||
|
|
"grad_norm": 5.0625,
|
||
|
|
"learning_rate": 1.029116427032555e-05,
|
||
|
|
"loss": 0.2574,
|
||
|
|
"step": 500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1055900621118013,
|
||
|
|
"eval_loss": 0.27571436762809753,
|
||
|
|
"eval_runtime": 46.2058,
|
||
|
|
"eval_samples_per_second": 14.003,
|
||
|
|
"eval_steps_per_second": 7.012,
|
||
|
|
"step": 500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.229813664596273,
|
||
|
|
"grad_norm": 3.625,
|
||
|
|
"learning_rate": 9.606118106312788e-06,
|
||
|
|
"loss": 0.2499,
|
||
|
|
"step": 520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.3540372670807455,
|
||
|
|
"grad_norm": 4.875,
|
||
|
|
"learning_rate": 8.922920432548891e-06,
|
||
|
|
"loss": 0.2531,
|
||
|
|
"step": 540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4782608695652173,
|
||
|
|
"grad_norm": 4.59375,
|
||
|
|
"learning_rate": 8.24477750009095e-06,
|
||
|
|
"loss": 0.2449,
|
||
|
|
"step": 560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.6024844720496896,
|
||
|
|
"grad_norm": 3.390625,
|
||
|
|
"learning_rate": 7.574871838062751e-06,
|
||
|
|
"loss": 0.2618,
|
||
|
|
"step": 580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7267080745341614,
|
||
|
|
"grad_norm": 5.375,
|
||
|
|
"learning_rate": 6.916347318025194e-06,
|
||
|
|
"loss": 0.2613,
|
||
|
|
"step": 600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7267080745341614,
|
||
|
|
"eval_loss": 0.26501935720443726,
|
||
|
|
"eval_runtime": 46.1464,
|
||
|
|
"eval_samples_per_second": 14.021,
|
||
|
|
"eval_steps_per_second": 7.021,
|
||
|
|
"step": 600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.8509316770186337,
|
||
|
|
"grad_norm": 2.84375,
|
||
|
|
"learning_rate": 6.2722943997669695e-06,
|
||
|
|
"loss": 0.2476,
|
||
|
|
"step": 620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.9751552795031055,
|
||
|
|
"grad_norm": 5.34375,
|
||
|
|
"learning_rate": 5.645735627756951e-06,
|
||
|
|
"loss": 0.247,
|
||
|
|
"step": 640
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 20,
|
||
|
|
"max_steps": 966,
|
||
|
|
"num_input_tokens_seen": 0,
|
||
|
|
"num_train_epochs": 6,
|
||
|
|
"save_steps": 161,
|
||
|
|
"stateful_callbacks": {
|
||
|
|
"TrainerControl": {
|
||
|
|
"args": {
|
||
|
|
"should_epoch_stop": false,
|
||
|
|
"should_evaluate": false,
|
||
|
|
"should_log": false,
|
||
|
|
"should_save": true,
|
||
|
|
"should_training_stop": false
|
||
|
|
},
|
||
|
|
"attributes": {}
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"total_flos": 6.782602179417293e+16,
|
||
|
|
"train_batch_size": 2,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|