FRAGE + AWD-LSTM-MoS + dynamic eval
Reported on 4 benchmarks across 1 task · 1 paper · 2 SOTA
Note: results are matched by exact model name. Different papers may use the same name for different model variants.
Medical4 results
- Test perplexity· 2018-09-18SOTA39.14best: 8.21 (SparseGPT (175B, 50% Sparsity))
- Validation perplexity· 2018-09-18SOTA40.85best: 15.69 (GPT-2 (fine-tuned))
- Test perplexity· 2018-09-1846.54best: 20.5 (GPT-3 (Zero-Shot))
- Validation perplexity· 2018-09-1847.38best: 36.1 (BERT-Large-CAS)