vit-base-mnist / trainer_state.json
farleyknight's picture
End of training
1632508
raw
history blame
3.58 kB
{
"best_metric": 0.689139723777771,
"best_model_checkpoint": "/storage/img-cls-data/roman_numeral_outputs/checkpoint-1445",
"epoch": 5.0,
"global_step": 1445,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.35,
"learning_rate": 1.8615916955017305e-05,
"loss": 2.2093,
"step": 100
},
{
"epoch": 0.69,
"learning_rate": 1.7231833910034604e-05,
"loss": 1.9053,
"step": 200
},
{
"epoch": 1.0,
"eval_accuracy": 0.7107843137254902,
"eval_loss": 1.324064016342163,
"eval_runtime": 2.1357,
"eval_samples_per_second": 191.039,
"eval_steps_per_second": 23.88,
"step": 289
},
{
"epoch": 1.04,
"learning_rate": 1.5847750865051904e-05,
"loss": 1.6347,
"step": 300
},
{
"epoch": 1.38,
"learning_rate": 1.4463667820069205e-05,
"loss": 1.4087,
"step": 400
},
{
"epoch": 1.73,
"learning_rate": 1.3079584775086506e-05,
"loss": 1.3293,
"step": 500
},
{
"epoch": 2.0,
"eval_accuracy": 0.7892156862745098,
"eval_loss": 0.933335542678833,
"eval_runtime": 1.9774,
"eval_samples_per_second": 206.327,
"eval_steps_per_second": 25.791,
"step": 578
},
{
"epoch": 2.08,
"learning_rate": 1.1695501730103806e-05,
"loss": 1.2134,
"step": 600
},
{
"epoch": 2.42,
"learning_rate": 1.0311418685121109e-05,
"loss": 1.1315,
"step": 700
},
{
"epoch": 2.77,
"learning_rate": 8.92733564013841e-06,
"loss": 1.1251,
"step": 800
},
{
"epoch": 3.0,
"eval_accuracy": 0.7843137254901961,
"eval_loss": 0.7988855838775635,
"eval_runtime": 2.1384,
"eval_samples_per_second": 190.795,
"eval_steps_per_second": 23.849,
"step": 867
},
{
"epoch": 3.11,
"learning_rate": 7.5432525951557104e-06,
"loss": 1.0903,
"step": 900
},
{
"epoch": 3.46,
"learning_rate": 6.159169550173011e-06,
"loss": 1.05,
"step": 1000
},
{
"epoch": 3.81,
"learning_rate": 4.775086505190312e-06,
"loss": 0.9837,
"step": 1100
},
{
"epoch": 4.0,
"eval_accuracy": 0.8186274509803921,
"eval_loss": 0.6956034898757935,
"eval_runtime": 2.0398,
"eval_samples_per_second": 200.024,
"eval_steps_per_second": 25.003,
"step": 1156
},
{
"epoch": 4.15,
"learning_rate": 3.3910034602076125e-06,
"loss": 1.0053,
"step": 1200
},
{
"epoch": 4.5,
"learning_rate": 2.0069204152249138e-06,
"loss": 0.9802,
"step": 1300
},
{
"epoch": 4.84,
"learning_rate": 6.228373702422146e-07,
"loss": 0.999,
"step": 1400
},
{
"epoch": 5.0,
"eval_accuracy": 0.8308823529411765,
"eval_loss": 0.689139723777771,
"eval_runtime": 1.9257,
"eval_samples_per_second": 211.874,
"eval_steps_per_second": 26.484,
"step": 1445
},
{
"epoch": 5.0,
"step": 1445,
"total_flos": 8.947091674631578e+17,
"train_loss": 1.2808735091793495,
"train_runtime": 453.7657,
"train_samples_per_second": 25.443,
"train_steps_per_second": 3.184
}
],
"max_steps": 1445,
"num_train_epochs": 5,
"total_flos": 8.947091674631578e+17,
"trial_name": null,
"trial_params": null
}