LLama-8B-Instruct-v0.1-MI-6e-7 / train_results.json
tengxiao1
TX
9969e95
raw
history blame contribute delete
233 Bytes
{
"epoch": 0.9982631930527722,
"total_flos": 0.0,
"train_loss": 1.2480301234145237,
"train_runtime": 21322.7535,
"train_samples": 59876,
"train_samples_per_second": 2.808,
"train_steps_per_second": 0.022
}