pasta-shapes / all_results.json
nateraw's picture
End of training
a8a1a12
raw
history blame contribute delete
187 Bytes
{
"epoch": 4.0,
"total_flos": 0.0,
"train_loss": 0.7672463009754816,
"train_runtime": 44.747,
"train_samples_per_second": 34.058,
"train_steps_per_second": 2.145
}