flan-t5-large-stacked-samsum-1024 / train_results.json
pszemraj's picture
add fp32 chk
6fe1cc3
raw
history blame
196 Bytes
{
"epoch": 1.0,
"train_loss": 0.10453087175669877,
"train_runtime": 27358.0249,
"train_samples": 29441,
"train_samples_per_second": 1.076,
"train_steps_per_second": 0.004
}