Text Generation on Cornell IMDB (test)
22.47ROUGE-LSECOND THOUGHTS (AEM + VM default)
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SECOND THOUGHTS (AEM + VM default)variant=AEM + VM (default)2023.01 | 22.47 | 8.84 | |
| SECOND THOUGHTS (AEM + AIL)variant=AEM + AIL2023.01 | 19.6 | 7.31 | |
| SECOND THOUGHTS (AEM Only)variant=AEM Only2023.01 | 16.37 | 7.01 | |
| InstructGPTService=Huge LM API service2023.01 | 12.53 | 8.78 | |
| GPT-3Service=Huge LM API service2023.01 | 11.22 | 8.43 |