AdvSoft (+ 4 layer QRNN + dynamic eval)

Reported on 2 benchmarks across 1 task · 1 paper

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Medical2 results

Language ModellingonWikiText-103
Test perplexity· 2019-06-10
28
best: 2.4 (RETRO (7.5B))
Improving Neural Language Modeling via Adversarial Training arXiv:1906.03805
Language ModellingonWikiText-103
Validation perplexity· 2019-06-10
27.2
best: 13.11 (Ensemble of All)
Improving Neural Language Modeling via Adversarial Training arXiv:1906.03805