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 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.0013 | 4.0 | 100 | 2.0768 | -1.5902 | -1.8351 | 0.4800 | 0.2449 | -251.4758 | -226.9239 | -0.8311 | -0.8228 |
| 0.1427 | 8.0 | 200 | 3.3444 | -6.8511 | -7.5916 | 0.5800 | 0.7405 | -257.8718 | -232.7693 | -0.7115 | -0.7215 |
| 0.0029 | 12.0 | 300 | 4.2892 | -2.5538 | -3.3508 | 0.5200 | 0.7970 | -253.1599 | -227.9944 | -0.9768 | -0.9928 |
| 0.0 | 16.0 | 400 | 5.0370 | -14.7509 | -15.7815 | 0.5300 | 1.0306 | -266.9717 | -241.5469 | -0.8694 | -0.9104 |
| 0.0 | 20.0 | 500 | 5.0695 | -14.7352 | -15.7245 | 0.5400 | 0.9894 | -266.9084 | -241.5294 | -0.8698 | -0.9112 |
| 0.0 | 24.0 | 600 | 5.0615 | -14.7459 | -15.7542 | 0.5500 | 1.0083 | -266.9414 | -241.5412 | -0.8694 | -0.9109 |
| 0.0 | 28.0 | 700 | 5.0597 | -14.7540 | -15.7286 | 0.5300 | 0.9747 | -266.9130 | -241.5502 | -0.8700 | -0.9110 |
| 0.0 | 32.0 | 800 | 5.0420 | -14.7242 | -15.7535 | 0.5400 | 1.0293 | -266.9406 | -241.5171 | -0.8695 | -0.9111 |
| 0.0 | 36.0 | 900 | 5.0441 | -14.7134 | -15.7250 | 0.5400 | 1.0116 | -266.9089 | -241.5051 | -0.8697 | -0.9110 |
| 0.0 | 40.0 | 1000 | 5.0567 | -14.7249 | -15.7125 | 0.5500 | 0.9876 | -266.8951 | -241.5179 | -0.8700 | -0.9109 |