[Bug] Github 和 OpenAI 的同名模型冲突

[CaiJingLong]

📦 部署环境

Docker

📌 软件版本

v1.21.8

💻 系统环境

macOS

🌐 浏览器

Edge

🐛 问题描述

配置了 Github 的 gpt4o 模型，则选择 OpenAI 的 gpt4o 模型时，会被错误的识别为 Github 的 gpt4o 模型
必须删除 Github 的同名模型，才可以识别到 OpenAI 的模型，否则会错误识别为 Github 的 gpt4o 模型。

暂时解决方案，删除 Github 中的 OpenAI 同名模型

📷 复现步骤

1. 删除语言模型中的 OpenAI 模型列表中的 gpt4o
2. 配置 Github 模型，并加入 gpt4o
3. 添加 OpenAI 中的 4o 模型
4. 打开默认助手的模型设置，选择 OpenAI 下的 gpt4o

这个影响所有位置的模型选择，而不仅仅是默认助手
哪怕关闭 Github，只要不删除 Github 模型中的 gpt4o，就无法使用 OpenAI 模型

🚦 期望结果

正确选择 OpenAI 的 gpt4o，而不是 github 的 gpt4o

📝 补充信息

Kapture.2024-10-10.at.14.29.07.mp4

[lobehubbot]

👀 @CaiJingLong

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。

[lobehubbot]

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

---

📦 Deployment environment

Docker

📌 Software version

v1.21.8

💻 System environment

macOS

🌐 Browser

Edge

🐛 Problem description

If Github's gpt4o model is configured, then when you select OpenAI's gpt4o model, it will be mistakenly recognized as Github's gpt4o model.
The model with the same name in Github must be deleted before the OpenAI model can be recognized, otherwise it will be mistakenly recognized as Github's gpt4o model.

Temporary solution, delete the OpenAI model with the same name in Github

📷 Steps to reproduce

1. Delete gpt4o in the OpenAI model list in the language model
2. Configure the Github model and add gpt4o
3. Open the model settings of the default assistant and select gpt4o under OpenAI

This affects model selection in all locations, not just the default helper
Even if Github is closed, the OpenAI model cannot be used as long as gpt4o in the Github model is not deleted.

🚦 Expected results

Correctly choose OpenAI's gpt4o instead of github's gpt4o

📝 Supplementary information

Kapture.2024-10-10.at.14.29.07.mp4

[CaiJingLong]

I upgrade the steps to reproduce, but the translate bot not follow change:

1. Delete gpt4o in the OpenAI model list in the language model
2. Configure the Github model and add gpt4o
3. Add gpt4o to OpenAI model list
4. Open the model settings of the default assistant and select gpt4o under OpenAI

[dosubot]

@CaiJingLong 你好！我是 Dosu，一个可以帮助解决 bug、回答问题，并成为贡献者的机器人。在等待人类维护者的期间，我会尽力帮助你解决问题。

The issue you're experiencing is due to a conflict between models with the same name from different providers. The system should differentiate models using unique identifiers, but it seems there's a problem with this process in your setup.

To resolve this temporarily, you can remove the conflicting model from GitHub, as you've mentioned. However, a more permanent solution would involve ensuring that each model has a unique identifier in the system, even if they share the same name. This is typically handled by the getModelCardsById function, which uses unique IDs to distinguish models ^[1].

You might want to check if the unique identifiers are correctly implemented and used in your configuration. If the issue persists, reviewing the logic in the model selection and configuration components might help identify where the conflict arises.

To continue talking to Dosu, mention @dosu.

[lobehubbot]

✅ @CaiJingLong

This issue is closed, If you have any questions, you can comment and reply.
此问题已经关闭。如果您有任何问题，可以留言并回复。