-
**例行检查**
[//]: # '方框内填 x 表示打钩'
- [ ✅] 我已确认目前没有类似 issue
- [✅ ] 我已完整查看过项目 README,以及[项目文档](https://doc.fastgpt.in/docs/intro/)
- [✅ ] 我使用了自己的 key,并确认我的 key 是可正常使用的
- [ ✅] 我理解并愿意跟进此 issue,协助测试和提供…
-
**例行检查**
[//]: # '方框内填 x 表示打钩'
- [x] 我已确认目前没有类似 issue
- [x] 我已完整查看过项目 README,以及[项目文档](https://doc.fastgpt.in/docs/intro/)
- [x] 我使用了自己的 key,并确认我的 key 是可正常使用的
- [x] 我理解并愿意跟进此 issue,协助测试和提供反馈
…
-
**例行检查**
[//]: # '方框内填 x 表示打钩'
- [x] 我已确认目前没有类似 features
- [x] 我已确认我已升级到最新版本
- [x] 我已完整查看过项目 README,已确定现有版本无法满足需求
- [x] 我理解并愿意跟进此 features,协助测试和提供反馈
- [x] 我理解并认可上述内容,并理解项目维护者精力有限,**不遵循规则的 fe…
-
**例行检查**
[//]: # '方框内填 x 表示打钩'
- [x ] 我已确认目前没有类似 issue
- [ x] 我已完整查看过项目 README,以及[项目文档](https://doc.fastgpt.in/docs/intro/)
- [ x] 我使用了自己的 key,并确认我的 key 是可正常使用的
- [x ] 我理解并愿意跟进此 issue,协助测试和提供…
-
**例行检查**
[//]: # '方框内填 x 表示打钩'
- [ ] 我已确认目前没有类似 issue
- [ ] 我已完整查看过项目 README,以及[项目文档](https://doc.fastgpt.in/docs/intro/)
- [ ] 我使用了自己的 key,并确认我的 key 是可正常使用的
- [ ] 我理解并愿意跟进此 issue,协助测试和提供反馈
…
-
请问调用阿里语音模型和文本模型方式一样吗?
参照文本模型的方式改了config文件和oneapi配置不生效。
本地whisper效果不好
我想对接阿里语音模型,没有思路,请指点下。
congfig.json怎么改 oneapi怎么设置,需要写python调用吗
fastgpt找不到相关文档
-
**例行检查**
[//]: # '方框内填 x 表示打钩'
- [x] 我已确认目前没有类似 features
- [x] 我已确认我已升级到最新版本
- [x] 我已完整查看过项目 README,已确定现有版本无法满足需求
- [x] 我理解并愿意跟进此 features,协助测试和提供反馈
- [x] 我理解并认可上述内容,并理解项目维护者精力有限,**不遵循规则的 fe…
-
一直都是用fastgpt+oneapi,看文档有描述xinference的功能,没看明白。
请问这个xinference是做什么的?跟oneapi什么区别?转了oneapi还需要这个吗,他的作用是什么
-
**例行检查**
[//]: # '方框内填 x 表示打钩'
- [ x] 我已确认目前没有类似 issue
- [x ] 我已完整查看过项目 README,以及[项目文档](https://doc.fastgpt.in/docs/intro/)
- [ x] 我使用了自己的 key,并确认我的 key 是可正常使用的
- [ x] 我理解并愿意跟进此 issue,协助测试和提供…
-
4.8.9
知识库+ai的场景下
同样的两段提示词,
一个提示词是让模型输出思考过程,然后回复答案。
另一个提示词是让模型不要输出思考过程,直接回复答案。
两种方式回答的效果不一样,输出思考过程的回答ok,不输出思考过程的回答不好。
例如知识库中有很多电话
用户输入问题:电话
输出提示词的ai回复有多个不同电话,需要进一步明确。。。。——这个效果还可以。
不输…