Adding New model Success But can't List and Select the added model.

[AlwaysBluer]

Self Checks

• [X] This is only for bug report, if you would like to ask a question, please head to Discussions.
• [X] I have searched for existing issues search for existing issues, including closed ones.
• [X] I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
• [X] [FOR CHINESE USERS] 请务必使用英文提交 Issue，否则会被关闭。谢谢！:）
• [X] Please do not modify this template :) and fill in all the required fields.

Dify version

0.11.1

Cloud or Self Hosted

Self Hosted (Docker)

Steps to reproduce

I'm integrating a new model provider and I have completed the majority of the work. But when I start docker and test, I found it could validate the credentials in a success, but can't store the model information. The module is a custom model integration.

(Ignore the X inference picture please, the uderlying code is for another model provider)

this is the console information

and these are the files added

the lindormai.yaml is written according to xinference.yaml

I'm sure this validate is success since I check it in the backend. So what's my be the reason of my failure? I would appreciate it if you can give me a hand. My

✔️ Expected Behavior

The rerank models should be listed after I added the model

❌ Actual Behavior

no models listed

[dosubot[bot]]

I found a similar unsolved discussion titled add self deploy llm model, add error which might be related to your issue ^[1]. Additionally, here are some potential reasons and solutions for your issue:

1. Model Configuration: Ensure that the model configuration files are correctly set up according to the AI Model Entity schema. Any discrepancies in the configuration might prevent the model from being listed.

2. Provider YAML File: Verify that the provider YAML file is correctly defined, including the provider's identifier, display name, supported model types, and credential rules. This file is crucial for the integration of new models ^[2].

3. Frontend Retrieval: Check if the frontend is correctly retrieving the list of available models. The frontend should access the backend's configuration and credential information to display the models. Ensure that the backend is providing the necessary data for the frontend to list the models ^[3].

4. Backend Validation: Since the credentials are validated successfully, ensure that the backend is also correctly storing the model information after validation. There might be an issue in the database operations or the logic that handles the storage of model information.

5. Logs and Debugging: Review the backend logs for any errors or warnings that might indicate why the model information is not being stored. This can provide insights into any issues during the storage process.

By checking these areas, you might be able to identify the cause of the issue and resolve it.

To continue talking to Dosu, mention @dosu.

[AlwaysBluer]

@dosu, do you know what is the entrypoint to persist the model information in dify?

[dosubot[bot]]

The entry point for persisting model information in Dify is primarily handled in the api/core/model_manager.py file. This file contains classes such as ModelInstance, ModelManager, and LBModelManager, which are responsible for managing model instances, provider configurations, and load balancing configurations. These classes handle the storage and retrieval of model information, including credentials and load balancing configurations ^[1].

To continue talking to Dosu, mention @dosu.

[crazywoola]

Please fresh the page to see if this persists.

[AlwaysBluer]

I found the problem, it's the missing implementation of function get_customizable_model_schema.