Views
No views yet
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.065 | 2.67 | 100 | 0.8697 | -2.4636 | -3.2853 | 0.6100 | 0.8218 | -143.5742 | -139.3904 | -1.1274 | -1.0559 |
| 0.0012 | 5.33 | 200 | 1.3102 | -7.4077 | -9.2027 | 0.5900 | 1.7950 | -202.7473 | -188.8313 | -1.0694 | -1.0590 |
| 0.0001 | 8.0 | 300 | 1.3943 | -7.6543 | -9.2750 | 0.6000 | 1.6207 | -203.4705 | -191.2975 | -1.0784 | -1.0808 |
| 0.0001 | 10.67 | 400 | 1.4180 | -7.8775 | -9.5214 | 0.6000 | 1.6439 | -205.9348 | -193.5301 | -1.0719 | -1.0792 |
| 0.0001 | 13.33 | 500 | 1.4336 | -7.9994 | -9.6559 | 0.5900 | 1.6566 | -207.2803 | -194.7483 | -1.0664 | -1.0765 |
| 0.0 | 16.0 | 600 | 1.4429 | -8.0690 | -9.7339 | 0.5900 | 1.6649 | -208.0597 | -195.4449 | -1.0648 | -1.0756 |
| 0.0001 | 18.67 | 700 | 1.4431 | -8.1002 | -9.7716 | 0.5900 | 1.6714 | -208.4371 | -195.7567 | -1.0628 | -1.0746 |
| 0.0001 | 21.33 | 800 | 1.4493 | -8.1216 | -9.7892 | 0.5900 | 1.6676 | -208.6128 | -195.9707 | -1.0626 | -1.0745 |
| 0.0 | 24.0 | 900 | 1.4508 | -8.1223 | -9.7864 | 0.5900 | 1.6640 | -208.5843 | -195.9782 | -1.0626 | -1.0745 |
| 0.0001 | 26.67 | 1000 | 1.4508 | -8.1181 | -9.7860 | 0.6000 | 1.6680 | -208.5809 | -195.9354 | -1.0629 | -1.0749 |