Sentiment Review Generation on Sentiment Review 100 instances (test)
2.7894Avg Preference ScorePFM
Evaluation Results
| Method | Links | |
|---|---|---|
| PFMBase Policy=PPO, Method Adaptation=PFM, Number of PFM iterations=52024.05 | 2.7894 | |
| PFMBase Policy=Reference, Method Adaptation=PFM, Number of PFM iterations=52024.05 | 2.7469 | |
| PFMBase Policy=PPO, Method Adaptation=PFM, Number of PFM iterations=12024.05 | 2.178 | |
| pi_PPOBase Policy=PPO, Method Adaptation=None, Number of PFM iterations=02024.05 | 2.0156 | |
| PFMBase Policy=Reference, Method Adaptation=PFM, Number of PFM iterations=12024.05 | 0.6399 | |
| pi_refBase Policy=Reference, Method Adaptation=None, Number of PFM iterations=02024.05 | -0.3607 |