stackexchange_math / all_results.json
sedrickkeh's picture
End of training
db054de verified
raw
history blame contribute delete
359 Bytes
{
"epoch": 2.9822485207100593,
"eval_loss": 0.6119207143783569,
"eval_runtime": 28.673,
"eval_samples_per_second": 39.689,
"eval_steps_per_second": 0.628,
"total_flos": 210845313269760.0,
"train_loss": 0.6075608276185536,
"train_runtime": 4413.1086,
"train_samples_per_second": 14.697,
"train_steps_per_second": 0.029
}