stackexchange_mathematica / train_results.json
sedrickkeh's picture
End of training
1b8db2d verified
raw
history blame contribute delete
203 Bytes
{
"epoch": 3.0,
"total_flos": 200795089797120.0,
"train_loss": 0.7261142373085022,
"train_runtime": 4198.995,
"train_samples_per_second": 14.629,
"train_steps_per_second": 0.029
}