BookCorpus
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Language Modeling | BookCorpus (test) | PPL10.83 | 18 | |
| Conditional Text Generation | BookCorpus | BLEU-190.4 | 9 | |
| Language Modeling | BookCorpus (val) | Perplexity17.81 | 7 | |
| Autoencoding | BookCorpus | BLEU90.58 | 5 | |
| LLM-as-judge evaluation | BookCorpus | Diversity Rank2.4 | 3 | |
| Open-ended text generation | BookCorpus Story Project Gutenberg | MAUVE0.91 | 2 |
Showing 6 of 6 rows