qwen2.5-0.5b-expo-DPO-ES-0.1 / all_results.json
hZzy's picture
End of training
ee26d68 verified
raw
history blame contribute delete
795 Bytes
{
"epoch": 1.5588096362777515,
"eval_dpo_loss": 0.6892624497413635,
"eval_logits": -2.120311975479126,
"eval_logps": -91.03094482421875,
"eval_loss": 0.6920701265335083,
"eval_objective": 0.6892624497413635,
"eval_ranking_idealized": 0.6030020713806152,
"eval_ranking_idealized_expo": 0.5222567319869995,
"eval_ranking_simple": 0.5559006333351135,
"eval_regularize": 0.6892624497413635,
"eval_runtime": 309.0026,
"eval_samples": 5790,
"eval_samples_per_second": 18.738,
"eval_steps_per_second": 1.563,
"eval_wo_beta": 7.426205635070801,
"total_flos": 0.0,
"train_loss": 0.4484944924441251,
"train_runtime": 15200.6621,
"train_samples": 50802,
"train_samples_per_second": 16.71,
"train_steps_per_second": 0.116
}