kurt-cobain / trainer_state.json
AlekseyKorshuk's picture
huggingartists
2b2cb11
raw
history blame
3.77 kB
{
"best_metric": 1.8867709636688232,
"best_model_checkpoint": "output/kurt-cobain/checkpoint-126",
"epoch": 2.0,
"global_step": 126,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.08,
"learning_rate": 0.00013507870183531476,
"loss": 3.0032,
"step": 5
},
{
"epoch": 0.16,
"learning_rate": 0.00012884599993319768,
"loss": 2.5844,
"step": 10
},
{
"epoch": 0.24,
"learning_rate": 0.00011888735840752609,
"loss": 2.5435,
"step": 15
},
{
"epoch": 0.32,
"learning_rate": 0.0001058186737011911,
"loss": 2.4834,
"step": 20
},
{
"epoch": 0.4,
"learning_rate": 9.044818420726556e-05,
"loss": 2.4343,
"step": 25
},
{
"epoch": 0.48,
"learning_rate": 7.372648442002871e-05,
"loss": 2.7558,
"step": 30
},
{
"epoch": 0.56,
"learning_rate": 5.668773501204858e-05,
"loss": 2.5089,
"step": 35
},
{
"epoch": 0.63,
"learning_rate": 4.0385704725240065e-05,
"loss": 2.4147,
"step": 40
},
{
"epoch": 0.71,
"learning_rate": 2.5828599592490882e-05,
"loss": 2.123,
"step": 45
},
{
"epoch": 0.79,
"learning_rate": 1.3916710004507539e-05,
"loss": 2.2562,
"step": 50
},
{
"epoch": 0.87,
"learning_rate": 5.38673186569003e-06,
"loss": 2.4784,
"step": 55
},
{
"epoch": 0.95,
"learning_rate": 7.662053209561833e-07,
"loss": 2.5457,
"step": 60
},
{
"epoch": 1.0,
"eval_loss": 2.059675931930542,
"eval_runtime": 3.6409,
"eval_samples_per_second": 22.522,
"eval_steps_per_second": 3.021,
"step": 63
},
{
"epoch": 1.03,
"learning_rate": 3.408888099334633e-07,
"loss": 2.2316,
"step": 65
},
{
"epoch": 1.11,
"learning_rate": 4.137086214086682e-06,
"loss": 2.3402,
"step": 70
},
{
"epoch": 1.19,
"learning_rate": 1.1920020081922749e-05,
"loss": 2.2531,
"step": 75
},
{
"epoch": 1.27,
"learning_rate": 2.320835154085542e-05,
"loss": 2.3415,
"step": 80
},
{
"epoch": 1.35,
"learning_rate": 3.7303948905573005e-05,
"loss": 2.1674,
"step": 85
},
{
"epoch": 1.43,
"learning_rate": 5.333506393059682e-05,
"loss": 2.2066,
"step": 90
},
{
"epoch": 1.51,
"learning_rate": 7.031024545323179e-05,
"loss": 1.9122,
"step": 95
},
{
"epoch": 1.59,
"learning_rate": 8.71796561146101e-05,
"loss": 2.1474,
"step": 100
},
{
"epoch": 1.67,
"learning_rate": 0.00010290000000000001,
"loss": 2.0603,
"step": 105
},
{
"epoch": 1.75,
"learning_rate": 0.0001164990457207046,
"loss": 1.8847,
"step": 110
},
{
"epoch": 1.83,
"learning_rate": 0.00012713575447996587,
"loss": 1.881,
"step": 115
},
{
"epoch": 1.9,
"learning_rate": 0.00013415229447692924,
"loss": 2.2478,
"step": 120
},
{
"epoch": 1.98,
"learning_rate": 0.00013711472479561806,
"loss": 2.2805,
"step": 125
},
{
"epoch": 2.0,
"eval_loss": 1.8867709636688232,
"eval_runtime": 3.3707,
"eval_samples_per_second": 22.547,
"eval_steps_per_second": 2.967,
"step": 126
}
],
"max_steps": 126,
"num_train_epochs": 2,
"total_flos": 130646016000000.0,
"trial_name": null,
"trial_params": null
}