Llama-2-7b-relevant-dpo / train_results.json
AmberYifan's picture
Model save
d87f2ac verified
{
"epoch": 3.0,
"total_flos": 0.0,
"train_loss": 0.48406885540674605,
"train_runtime": 1618.971,
"train_samples": 2000,
"train_samples_per_second": 3.706,
"train_steps_per_second": 0.117
}