Language Modelling On Wiki 40B
Metrics
Perplexity
Results
Performance results of various models on this benchmark
Comparison Table
| Model Name | Perplexity |
|---|---|
| transformer-quality-in-linear-time | 14.998 |
| combiner-full-attention-transformer-with | 16.49 |
| combiner-full-attention-transformer-with | 16.60 |