gemma-7b-sft-qlora-1 / all_results.json
chansung's picture
End of training
4671ebd verified
raw
history blame
354 Bytes
{
"epoch": 22.73,
"eval_loss": 2.209489345550537,
"eval_runtime": 0.6001,
"eval_samples": 16,
"eval_samples_per_second": 3.333,
"eval_steps_per_second": 1.667,
"train_loss": 3.2900945229530336,
"train_runtime": 472.0874,
"train_samples": 926,
"train_samples_per_second": 4.66,
"train_steps_per_second": 0.265
}