deepseek-8b-orpo-lora / train_results.json
zfz1's picture
Model save
9cb3e8d verified
raw
history blame contribute delete
No virus
232 Bytes
{
"epoch": 1.9936102236421727,
"total_flos": 0.0,
"train_loss": 0.7846963420892373,
"train_runtime": 5577.3844,
"train_samples": 20000,
"train_samples_per_second": 7.172,
"train_steps_per_second": 0.056
}