05ce659608
Model: W-61/llama-3-8b-base-r-dpo-ultrafeedback-4xh200-batch-128-20260426-105614 Source: Original Platform
24 lines
873 B
JSON
24 lines
873 B
JSON
{
|
|
"epoch": 0.9989528795811519,
|
|
"eval_logits/chosen": -0.8717418909072876,
|
|
"eval_logits/rejected": -0.8549114465713501,
|
|
"eval_logps/chosen": -415.0469055175781,
|
|
"eval_logps/ref_chosen": -288.6414794921875,
|
|
"eval_logps/ref_rejected": -265.96160888671875,
|
|
"eval_logps/rejected": -453.6796875,
|
|
"eval_loss": 0.5324965119361877,
|
|
"eval_r_dpo/chosen_len": 286.97601318359375,
|
|
"eval_r_dpo/length_delta": 40.88800048828125,
|
|
"eval_r_dpo/regularization_term": 0.0,
|
|
"eval_r_dpo/rejected_len": 246.08799743652344,
|
|
"eval_runtime": 77.9812,
|
|
"eval_samples": 2000,
|
|
"eval_samples_per_second": 25.647,
|
|
"eval_steps_per_second": 3.206,
|
|
"total_flos": 0.0,
|
|
"train_loss": 4.588983364824979,
|
|
"train_runtime": 6140.251,
|
|
"train_samples": 61135,
|
|
"train_samples_per_second": 9.956,
|
|
"train_steps_per_second": 0.078
|
|
} |