qwen_1.8_feedback_dirty / train_results.json
terry69's picture
Model save
935f3da verified
raw
history blame
231 Bytes
{
"epoch": 1.0,
"total_flos": 161851279147008.0,
"train_loss": 0.8951905027058812,
"train_runtime": 4028.5705,
"train_samples": 98952,
"train_samples_per_second": 6.596,
"train_steps_per_second": 0.412
}