downstream_gemma / all_results.json
terry69's picture
End of training
68862cb verified
raw
history blame contribute delete
363 Bytes
{
"epoch": 1.0,
"eval_runtime": 3.7697,
"eval_samples": 10,
"eval_samples_per_second": 2.653,
"eval_steps_per_second": 0.796,
"total_flos": 1.3011067827388416e+16,
"train_loss": 1.2484874595598594,
"train_runtime": 22644.1998,
"train_samples": 103932,
"train_samples_per_second": 2.726,
"train_steps_per_second": 0.17
}