|
{ |
|
"epoch": 2.999098751126561, |
|
"eval_kl/chosen": 0.0002518190012779087, |
|
"eval_kl/rejected": -9.245327964890748e-05, |
|
"eval_logits/chosen": -1.375, |
|
"eval_logits/rejected": -1.3359375, |
|
"eval_logps/chosen": -322.0, |
|
"eval_logps/rejected": -298.0, |
|
"eval_loss": 0.7700312733650208, |
|
"eval_nll_loss": 1.046875, |
|
"eval_ref_probs/chosen": 0.0038271306548267603, |
|
"eval_ref_probs/rejected": 0.0034569373819977045, |
|
"eval_rejected_term/max": -9.245327964890748e-05, |
|
"eval_rejected_term/min": -9.245327964890748e-05, |
|
"eval_rewards/accuracies": 0.46000000834465027, |
|
"eval_rewards/chosen": 0.016357421875, |
|
"eval_rewards/margins": 0.048828125, |
|
"eval_rewards/rejected": -0.032470703125, |
|
"eval_runtime": 25.3986, |
|
"eval_samples_per_second": 39.372, |
|
"eval_steps_per_second": 9.843 |
|
} |