e8efbe1ca9
Model: W-61/llama-3-8b-base-new-dpo-ultrafeedback-4xh200-batch-128-q_t-0.43-s_star-0.4-20260429-230725 Source: Original Platform
21 lines
825 B
JSON
21 lines
825 B
JSON
{
|
|
"epoch": 0.9989528795811519,
|
|
"eval_KL/chosen_KL_mean": -301.3088073730469,
|
|
"eval_KL/mean": -359.5681457519531,
|
|
"eval_KL/rejected_KL_mean": -417.8275146484375,
|
|
"eval_KL/std": 183.18824768066406,
|
|
"eval_fcm_dpo/beta": 0.0021828499156981707,
|
|
"eval_logits/chosen": -0.846419095993042,
|
|
"eval_logits/rejected": -0.8267433047294617,
|
|
"eval_logps/chosen": -589.1356201171875,
|
|
"eval_logps/ref_chosen": -287.8268127441406,
|
|
"eval_logps/ref_rejected": -266.9300231933594,
|
|
"eval_logps/rejected": -684.7575073242188,
|
|
"eval_loss": 0.6045041084289551,
|
|
"eval_margin_dpo/margin_mean": 116.51864624023438,
|
|
"eval_margin_dpo/margin_std": 204.82022094726562,
|
|
"eval_runtime": 78.19,
|
|
"eval_samples": 2000,
|
|
"eval_samples_per_second": 25.579,
|
|
"eval_steps_per_second": 3.197
|
|
} |