Better ranking between models

shmsw25 / FActScore

A package to evaluate factuality of long-form generation. Original implementation of our EMNLP 2023 paper "FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long Form Text Generation"

https://arxiv.org/abs/2305.14251

MIT License

275 stars 40 forks source link

Better ranking between models #13

Closed martiansideofthemoon closed 1 year ago

martiansideofthemoon commented 1 year ago

Yizhong: clear tradeoff between number of facts and precision

Possible solution: graph between number of facts and precision

shmsw25 commented 1 year ago

PR #21