{ "epoch": 1.0, "eval_beta_dpo/beta": 0.11110328137874603, "eval_beta_dpo/beta_margin_grad_mean": -0.36612096428871155, "eval_beta_dpo/beta_margin_grad_std": 0.18751376867294312, "eval_beta_dpo/beta_margin_mean": 1.2664130926132202, "eval_beta_dpo/beta_margin_std": 1.7243142127990723, "eval_beta_dpo/beta_used": 0.11110328137874603, "eval_beta_dpo/beta_used_raw": 0.08256327360868454, "eval_beta_dpo/gap_mean": 9.540072441101074, "eval_beta_dpo/gap_std": 15.214235305786133, "eval_beta_dpo/loss_margin_mean": 9.249459266662598, "eval_beta_dpo/mask_keep_frac": 1.0, "eval_logits/chosen": 2.378512382507324, "eval_logits/rejected": 2.443873882293701, "eval_loss": 0.5857504606246948, "eval_runtime": 44.777, "eval_samples": 2339, "eval_samples_per_second": 52.237, "eval_steps_per_second": 1.653, "total_flos": 0.0, "train_loss": 1.1809129529850073, "train_runtime": 2340.6187, "train_samples": 43598, "train_samples_per_second": 18.627, "train_steps_per_second": 0.291 }