LLama-8B-Instruct-v0.1-MI-2e-5 / train_results.json
tengxiao1
TX
ae60326
{
"epoch": 0.9982631930527722,
"total_flos": 0.0,
"train_loss": 1.2796013329728513,
"train_runtime": 21025.6961,
"train_samples": 59876,
"train_samples_per_second": 2.848,
"train_steps_per_second": 0.022
}