drake / trainer_state.json
AlekseyKorshuk's picture
huggingartists
461ffb6
{
"best_metric": 2.7749314308166504,
"best_model_checkpoint": "output/drake/checkpoint-351",
"epoch": 1.0,
"global_step": 351,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 0.00013713131733182566,
"loss": 2.8146,
"step": 5
},
{
"epoch": 0.03,
"learning_rate": 0.00013692540685817455,
"loss": 2.4938,
"step": 10
},
{
"epoch": 0.04,
"learning_rate": 0.00013658268089626935,
"loss": 2.6848,
"step": 15
},
{
"epoch": 0.06,
"learning_rate": 0.000136103825724055,
"loss": 2.4887,
"step": 20
},
{
"epoch": 0.07,
"learning_rate": 0.0001354898002059893,
"loss": 2.586,
"step": 25
},
{
"epoch": 0.09,
"learning_rate": 0.00013474183387300238,
"loss": 2.4906,
"step": 30
},
{
"epoch": 0.1,
"learning_rate": 0.00013386142446047196,
"loss": 2.4834,
"step": 35
},
{
"epoch": 0.11,
"learning_rate": 0.00013285033490914263,
"loss": 2.6758,
"step": 40
},
{
"epoch": 0.13,
"learning_rate": 0.00013171058983499532,
"loss": 2.4307,
"step": 45
},
{
"epoch": 0.14,
"learning_rate": 0.00013044447147513605,
"loss": 2.8348,
"step": 50
},
{
"epoch": 0.16,
"learning_rate": 0.00012905451511782055,
"loss": 2.7339,
"step": 55
},
{
"epoch": 0.17,
"learning_rate": 0.0001275435040257681,
"loss": 2.4359,
"step": 60
},
{
"epoch": 0.19,
"learning_rate": 0.00012591446386292745,
"loss": 2.7405,
"step": 65
},
{
"epoch": 0.2,
"learning_rate": 0.00012417065663585696,
"loss": 2.5262,
"step": 70
},
{
"epoch": 0.21,
"learning_rate": 0.00012231557416184885,
"loss": 2.5211,
"step": 75
},
{
"epoch": 0.23,
"learning_rate": 0.0001203529310768803,
"loss": 2.4883,
"step": 80
},
{
"epoch": 0.24,
"learning_rate": 0.00011828665739738674,
"loss": 2.6898,
"step": 85
},
{
"epoch": 0.26,
"learning_rate": 0.00011612089065075855,
"loss": 2.5956,
"step": 90
},
{
"epoch": 0.27,
"learning_rate": 0.0001138599675903136,
"loss": 2.4562,
"step": 95
},
{
"epoch": 0.28,
"learning_rate": 0.00011150841551133841,
"loss": 2.457,
"step": 100
},
{
"epoch": 0.3,
"learning_rate": 0.00010907094318558638,
"loss": 2.6934,
"step": 105
},
{
"epoch": 0.31,
"learning_rate": 0.00010655243143238444,
"loss": 2.6493,
"step": 110
},
{
"epoch": 0.33,
"learning_rate": 0.00010395792334523303,
"loss": 2.5976,
"step": 115
},
{
"epoch": 0.34,
"learning_rate": 0.00010129261419346177,
"loss": 2.7521,
"step": 120
},
{
"epoch": 0.36,
"learning_rate": 9.856184101917151e-05,
"loss": 2.5228,
"step": 125
},
{
"epoch": 0.37,
"learning_rate": 9.577107195028614e-05,
"loss": 2.555,
"step": 130
},
{
"epoch": 0.38,
"learning_rate": 9.292589525111789e-05,
"loss": 2.6695,
"step": 135
},
{
"epoch": 0.4,
"learning_rate": 9.003200813237038e-05,
"loss": 2.5893,
"step": 140
},
{
"epoch": 0.41,
"learning_rate": 8.709520534298517e-05,
"loss": 2.4968,
"step": 145
},
{
"epoch": 0.43,
"learning_rate": 8.4121367566679e-05,
"loss": 2.6097,
"step": 150
},
{
"epoch": 0.44,
"learning_rate": 8.111644964640216e-05,
"loss": 2.4365,
"step": 155
},
{
"epoch": 0.46,
"learning_rate": 7.808646866030093e-05,
"loss": 2.6697,
"step": 160
},
{
"epoch": 0.47,
"learning_rate": 7.503749187305804e-05,
"loss": 2.5317,
"step": 165
},
{
"epoch": 0.48,
"learning_rate": 7.197562458674299e-05,
"loss": 2.6247,
"step": 170
},
{
"epoch": 0.5,
"learning_rate": 6.890699791548972e-05,
"loss": 2.4144,
"step": 175
},
{
"epoch": 0.51,
"learning_rate": 6.583775650849418e-05,
"loss": 2.6802,
"step": 180
},
{
"epoch": 0.53,
"learning_rate": 6.277404624590518e-05,
"loss": 2.7703,
"step": 185
},
{
"epoch": 0.54,
"learning_rate": 5.9722001932251785e-05,
"loss": 2.8038,
"step": 190
},
{
"epoch": 0.56,
"learning_rate": 5.6687735012048764e-05,
"loss": 2.5009,
"step": 195
},
{
"epoch": 0.57,
"learning_rate": 5.3677321332176824e-05,
"loss": 2.7397,
"step": 200
},
{
"epoch": 0.58,
"learning_rate": 5.069678897554777e-05,
"loss": 2.2651,
"step": 205
},
{
"epoch": 0.6,
"learning_rate": 4.775210619040673e-05,
"loss": 2.6494,
"step": 210
},
{
"epoch": 0.61,
"learning_rate": 4.4849169439453994e-05,
"loss": 2.4319,
"step": 215
},
{
"epoch": 0.63,
"learning_rate": 4.199379159270695e-05,
"loss": 2.5083,
"step": 220
},
{
"epoch": 0.64,
"learning_rate": 3.919169028775044e-05,
"loss": 2.4238,
"step": 225
},
{
"epoch": 0.66,
"learning_rate": 3.6448476480681456e-05,
"loss": 2.7361,
"step": 230
},
{
"epoch": 0.67,
"learning_rate": 3.3769643210672815e-05,
"loss": 2.3582,
"step": 235
},
{
"epoch": 0.68,
"learning_rate": 3.1160554600657045e-05,
"loss": 2.5586,
"step": 240
},
{
"epoch": 0.7,
"learning_rate": 2.862643511615132e-05,
"loss": 2.6506,
"step": 245
},
{
"epoch": 0.71,
"learning_rate": 2.6172359103734907e-05,
"loss": 2.4642,
"step": 250
},
{
"epoch": 0.73,
"learning_rate": 2.3803240630126026e-05,
"loss": 2.4471,
"step": 255
},
{
"epoch": 0.74,
"learning_rate": 2.152382364220499e-05,
"loss": 2.3953,
"step": 260
},
{
"epoch": 0.75,
"learning_rate": 1.9338672467686093e-05,
"loss": 2.5485,
"step": 265
},
{
"epoch": 0.77,
"learning_rate": 1.725216267546251e-05,
"loss": 2.483,
"step": 270
},
{
"epoch": 0.78,
"learning_rate": 1.5268472313922077e-05,
"loss": 2.8104,
"step": 275
},
{
"epoch": 0.8,
"learning_rate": 1.3391573544780878e-05,
"loss": 2.4582,
"step": 280
},
{
"epoch": 0.81,
"learning_rate": 1.1625224689186508e-05,
"loss": 2.698,
"step": 285
},
{
"epoch": 0.83,
"learning_rate": 9.97296270201696e-06,
"loss": 2.3153,
"step": 290
},
{
"epoch": 0.84,
"learning_rate": 8.438096089447732e-06,
"loss": 2.4751,
"step": 295
},
{
"epoch": 0.85,
"learning_rate": 7.023698283963675e-06,
"loss": 2.7106,
"step": 300
},
{
"epoch": 0.87,
"learning_rate": 5.73260149008795e-06,
"loss": 2.6915,
"step": 305
},
{
"epoch": 0.88,
"learning_rate": 4.567391013146352e-06,
"loss": 2.8011,
"step": 310
},
{
"epoch": 0.9,
"learning_rate": 3.530400082425805e-06,
"loss": 2.65,
"step": 315
},
{
"epoch": 0.91,
"learning_rate": 2.623705179092382e-06,
"loss": 2.4158,
"step": 320
},
{
"epoch": 0.93,
"learning_rate": 1.8491218782241098e-06,
"loss": 2.3891,
"step": 325
},
{
"epoch": 0.94,
"learning_rate": 1.2082012132854124e-06,
"loss": 2.5672,
"step": 330
},
{
"epoch": 0.95,
"learning_rate": 7.022265703216476e-07,
"loss": 2.5956,
"step": 335
},
{
"epoch": 0.97,
"learning_rate": 3.322111180939607e-07,
"loss": 2.4543,
"step": 340
},
{
"epoch": 0.98,
"learning_rate": 9.889577929986752e-08,
"loss": 2.6378,
"step": 345
},
{
"epoch": 1.0,
"learning_rate": 2.7477469421299584e-09,
"loss": 2.6626,
"step": 350
},
{
"epoch": 1.0,
"eval_loss": 2.7749314308166504,
"eval_runtime": 12.2562,
"eval_samples_per_second": 45.365,
"eval_steps_per_second": 5.711,
"step": 351
}
],
"max_steps": 3510,
"num_train_epochs": 10,
"total_flos": 366592720896000.0,
"trial_name": null,
"trial_params": null
}