HASH Layers 10B (0-shot)

Reported on 6 benchmarks across 3 tasks · 1 paper

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

Natural Language Processing6 results

Question AnsweringonCOPA
Accuracy· 2022-03-14
64
best: 100 (PaLM 540B (finetuned) )
Efficient Language Modeling with Sparse all-MLP arXiv:2203.06850
Question AnsweringonPIQA
Accuracy· 2022-03-14
63.8
best: 90.1 (Unicorn 11B (fine-tuned))
Efficient Language Modeling with Sparse all-MLP arXiv:2203.06850
Question AnsweringonStoryCloze
Accuracy· 2022-03-14
64.7
best: 96.3 (BLOOMZ)
Efficient Language Modeling with Sparse all-MLP arXiv:2203.06850
Common Sense ReasoningonWinoGrande
Accuracy· 2022-03-14
51.7
best: 96.1 (ST-MoE-32B 269B (fine-tuned))
Efficient Language Modeling with Sparse all-MLP arXiv:2203.06850
Common Sense ReasoningonReCoRD
EM· 2022-03-14
67.2
best: 95.9 (Turing NLR v5 XXL 5.4B (fine-tuned))
Efficient Language Modeling with Sparse all-MLP arXiv:2203.06850
Sentence CompletiononHellaSwag
Accuracy· 2022-03-14
33
best: 96.1 (CompassMTL 567M with Tailor)
Efficient Language Modeling with Sparse all-MLP arXiv:2203.06850