FLAN 137B (few-shot, k=16)

Reported on 3 benchmarks across 2 tasks · 1 paper

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

Natural Language Processing3 results

Question AnsweringonCOPA
Accuracy· 2021-09-03
87
best: 100 (PaLM 540B (finetuned) )
Finetuned Language Models Are Zero-Shot Learners arXiv:2109.01652
Question AnsweringonOBQA
Accuracy· 2021-09-03
78.2
best: 78.4 (FLAN 137B (zero-shot))
Finetuned Language Models Are Zero-Shot Learners arXiv:2109.01652
Common Sense ReasoningonWinoGrande
Accuracy· 2021-09-03
72.8
best: 96.1 (ST-MoE-32B 269B (fine-tuned))
Finetuned Language Models Are Zero-Shot Learners arXiv:2109.01652