Model_folder / train_results.json
JLB-JLB's picture
🍻 cheers
d4d6a87
raw
history blame
208 Bytes
{
"epoch": 4.0,
"total_flos": 3.205097416476426e+17,
"train_loss": 0.15786233500522726,
"train_runtime": 149.2116,
"train_samples_per_second": 27.719,
"train_steps_per_second": 0.456
}