mirror of
https://github.com/tencentmusic/supersonic.git
synced 2025-12-31 14:28:27 +08:00
c398ac1a84
* [improvement] llm supports all models * [improvement] alias convert to SemanticParseInfo * [improvement] support join * [improvement] add evaluation.py * [improvement] add text2sql_evalution.py * [improvement] add text2sql_evalution.py * [improvement] add evalution * [improvement] add evalution * [improvement] add evalution --------- Co-authored-by: zuopengge <hwzuopengge@tencent.com>
438 B
438 B
评测流程
1、正常启动项目(必须包括LLM服务) 2、将要评测问题放到evalution/data目录下，如：internet.txt；将要评测问题对应的SQL也放到evalution/data目录下，如：gold_example_dusql.txt。 3、执行evalution.sh脚本，主要包括构建表数据、获取模型预测结果，执行对比逻辑。可以在命令行看到执行准确率，错误case会写到同目录的eval.json文件中。