From b358e9880de13993463a5039dac37a4a325df109 Mon Sep 17 00:00:00 2001 From: Yang Liu <pku7yang@gmail.com> Date: Fri, 16 Jun 2023 16:16:19 -0700 Subject: [PATCH] Create README.md --- README.md | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) create mode 100644 README.md diff --git a/README.md b/README.md new file mode 100644 index 0000000..86217f8 --- /dev/null +++ b/README.md @@ -0,0 +1,16 @@ +# Code for paper "G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment" [https://arxiv.org/abs/2303.16634] + +## Experiments on SummEval dataset +### Evaluate fluency on SummEval dataset +```python .\gpt4_eval.py --prompt .\prompts\summeval\flu_detailed.txt --save_fp .\results\gpt4_flu_detailed.json --summeval_fp .\data\summeval.json --key XXXXX``` + +### Meta Evaluate the G-Eval results + +```python .\meta_eval_summeval.py --input_fp .\results\gpt4_flu_detailed.json --dimension fluency``` + +## Prompts and Evaluation Results + +Prompts used to evaluate SummEval are in prompts/summeval + +G-eval results on SummEval are in results/summeval + -- GitLab