OPENAI_MODEL_LIST参数似乎未按文档描述的方式执行

[LittlePeach996]

📦 部署环境

Docker

📌 软件版本

v1.35.12

💻 系统环境

Other Linux

🌐 浏览器

Edge, Safari

🐛 问题描述

使用第三方 API 接口时的模型配置问题

问题描述

在使用第三方 API 接口时出现了问题，该第三方在 /v1/chat/completions 下给出了来自多个 provider 的 API。

实际操作

我尝试通过 .env 文件设置 OPENAI_MODEL_LIST 参数，以实现模型扩展配置的修改。根据以下文档的描述：

• 环境变量 - 模型提供商
• 高级配置 - 模型列表

我设置了以下参数：

OPENAI_MODEL_LIST=+all,-gpt-4o,gpt-4o-all=gpt-4o<128000:fc:vision:file>

预期操作结果

1. 在所有模型列表中删除原有的 gpt-4o 模型。
2. 将 gpt-4o-all 重命名为 gpt-4o，并配置其最大 token 为 128000，启用视觉识别、函数调用和文件上传功能。
3. 返回剩余的所有模型。

实际操作结果

模型列表中包含以下两个模型，结果与预期不符：

all
gpt-4o-all

all是一个不可用的模型，gpt-4o按照我的预期修改了模型扩展配置。

问题定位

请问以下几点是否可能是问题的原因：

1. 这是 OPENAI_MODEL_LIST 参数的 bug 吗？
2. 文档中的描述是否有误？
3. 或者是我的理解存在偏差？

希望能够得到进一步的说明或修复建议。

---

感谢您的支持！

📷 复现步骤

复现流程

1. 设置 .env 文件中的 OPENAI_MODEL_LIST 参数：

   OPENAI_MODEL_LIST=+all,-gpt-4o,gpt-4o-all=gpt-4o<128000:fc:vision:file>

2. 保存 .env 文件。

3. 重构 LobeChat 项目。

4. 打开 LobeChat 面板，查看模型列表。

🚦 期望结果

希望得到的答复

针对上述内容，期望能够得到以下问题的明确解答：

1. 这是 OPENAI_MODEL_LIST 参数的 bug 吗？
2. 文档中的描述是否有误？
3. 或者是我的理解存在偏差？

📝 补充信息

无

[lobehubbot]

👀 @LittlePeach996

Thank you for raising an issue. We will investigate into the matter and get back to you as soon as possible.
Please make sure you have given us as much context as possible.
非常感谢您提交 issue。我们会尽快调查此事，并尽快回复您。 请确保您已经提供了尽可能多的背景信息。

[dosubot[bot]]

Hey there, @LittlePeach996! I'm here to help you with any bugs, questions, or contributions you might have. Let's tackle this together!

The issue you're experiencing with the OPENAI_MODEL_LIST parameter might be due to a few reasons:

1. Bug in the Parameter Handling: It's possible that there is a bug in how the OPENAI_MODEL_LIST parameter is processed in version v1.35.12. The behavior you're describing, where the all model appears as an unusable option, suggests that the +all operation might not be functioning as expected. However, I couldn't find specific information about known bugs related to this parameter in the current version ^[1].

2. Documentation Discrepancy: There might be a discrepancy between the documentation and the actual implementation. The documentation should ideally provide clear guidance on how the +all and other operations should work, but if there's a mismatch, it could lead to confusion.

3. Misunderstanding of the Parameter Usage: Ensure that the syntax and operations you're using are supported as per the documentation. The +all operation is intended to enable all models, but if it's not working as expected, it might be due to a misunderstanding or an undocumented limitation.

To address this issue, you might want to:

• Double-check the documentation to ensure that the operations you're using are correctly implemented.
• Test with a simpler configuration to isolate the problem, such as removing the +all operation and manually specifying models to see if the issue persists.
• Consider reaching out to the maintainers or checking for any updates or patches that might address this issue.

If the problem continues, it might be worth reporting it as a potential bug to the maintainers for further investigation.

To continue talking to Dosu, mention @dosu.

[SpeedupMaster]

没有+all这个，只有-all

[lobehubbot]

Bot detected the issue body's language is not English, translate it automatically. 👯👭🏻🧑‍🤝‍🧑👫🧑🏿‍🤝‍🧑🏻👩🏾‍🤝‍👨🏿👬🏿

---

There is no +all, only -all