dpo-model-lora / all_results.json
lewtun's picture
lewtun HF staff
End of training
aefb0c8 verified
raw
history blame contribute delete
553 Bytes
{
"epoch": 0.9989701338825953,
"eval_logits/chosen": -2.3471405506134033,
"eval_logits/rejected": -2.2715110778808594,
"eval_logps/chosen": -378.4928283691406,
"eval_logps/rejected": -359.0921325683594,
"eval_loss": 0.6533845067024231,
"eval_rewards/accuracies": 0.6171875,
"eval_rewards/chosen": -0.7319744229316711,
"eval_rewards/margins": 0.09831348061561584,
"eval_rewards/rejected": -0.8302878737449646,
"eval_runtime": 2.1287,
"eval_samples_per_second": 469.776,
"eval_steps_per_second": 7.516
}