Best model so far - eval_loss: 0.7097 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.7265 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.7521 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.7739 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.7890 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.8000 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.8412 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9192 -------------------------------------------------------- (Trained with Unsloth)

Initial model state ------------------------------------------------------------------------ (Trained with Unsloth)

Best model so far - eval_loss: 1.0718 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 1.0744 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 1.0813 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 1.0984 -------------------------------------------------------- (Trained with Unsloth)

Initial model state ------------------------------------------------------------------------ (Trained with Unsloth)

Best model so far - eval_loss: 1.1487 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 1.1727 -------------------------------------------------------- (Trained with Unsloth)

Initial model state ------------------------------------------------------------------------ (Trained with Unsloth)

Best model so far - eval_loss: 1.5643 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 1.5651 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 1.5654 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 1.5671 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 1.6000 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 1.6312 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 1.7003 -------------------------------------------------------- (Trained with Unsloth)

Initial model state ------------------------------------------------------------------------ (Trained with Unsloth)

Best model so far - eval_loss: 1.5922 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 1.6187 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 1.6529 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 1.7244 -------------------------------------------------------- (Trained with Unsloth)

Initial model state ------------------------------------------------------------------------ (Trained with Unsloth)

Best model so far - eval_loss: 0.9467 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9469 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9476 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9477 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9480 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9481 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9493 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9503 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9515 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9516 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9524 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9529 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9540 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9577 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9580 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9582 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9584 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9591 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9599 -------------------------------------------------------- (Trained with Unsloth)

Best model so far - eval_loss: 0.9625 -------------------------------------------------------- (Trained with Unsloth)