max token问题导致标题生成错误

[jiangying000]

有个必现的bug， 我如果把模型设置为：o1-mini max token 设置为 32768

压缩模型设置为 gpt-4o-mini

聊天2次，标题生成错误：{ "error": { "message": "max_tokens is too large: 32768. This model supports at most 16384 completion tokens, whereas you provided 32768. (request id: )", "type": "invalid_request_error", "param": "max_tokens", "code": null }}

Originally posted by @jiangying000 in https://github.com/ChatGPTNextWeb/ChatGPT-Next-Web/issues/5646#issuecomment-2408624801

[nextchat-manager[bot]]

Please follow the issue template to update title and description of your issue.

[Issues-translate-bot]

Bot detected the issue body's language is not English, translate it automatically.

---

Title: max token problem causes title generation error

[Dogtiti]

也就是说本来设置max_token是为了用于o1-mini 结果gpt-4o-mini也报错了

[Issues-translate-bot]

Bot detected the issue body's language is not English, translate it automatically.

---

In other words, max_token was originally set to be used for gpt-4o-mini. As a result, o1-mini also reported an error.

[Dogtiti]

@lloydzhou 这里由于gpt-4o-mini进入了这个判断导致带上了max_tokens 看看有没有好的办法来处理各个模型的参数问题

[Issues-translate-bot]

Bot detected the issue body's language is not English, translate it automatically.

---

@lloydzhou Here, because gpt-4o-mini entered this judgment, max_tokens was brought in See if there is a good way to deal with the parameter issues of each model

[lloydzhou]

这里把max_token去掉可以吗？

[Issues-translate-bot]

Bot detected the issue body's language is not English, translate it automatically.

---

Can I remove max_token here?

[q246]

环境变量中max_token可以直接设置吗 默认4000太小了