How does HostingConfiguration limit the available text-embedding models

[zhaobingshuang]

Self Checks

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

1. Is this request related to a challenge you're experiencing? Tell me about your story.

class HostingConfiguration:
    ...
    
    @staticmethod
    def init_wenxin() -> HostingProvider:
        quota_unit = QuotaUnit.TOKENS
        if dify_config.HOSTED_WENXIN_ENABLED:
            hosted_quota_limit = dify_config.HOSTED_WENXIN_QUOTA_LIMIT
            quotas = [TrialHostingQuota(quota_limit=hosted_quota_limit, restrict_models=[
                RestrictModel(model="embedding-v1", model_type=ModelType.TEXT_EMBEDDING)
            ])]

            credentials = {
                "api_key": dify_config.HOSTED_WENXIN_API_KEY,
                "secret_key": dify_config.HOSTED_WENXIN_SECRET_KEY,
            }

            return HostingProvider(
                enabled=True,
                credentials=credentials,  # use credentials from the provider
                quota_unit=quota_unit,
                quotas=quotas,
            )

        return HostingProvider(
            enabled=False,
            quota_unit=quota_unit,
        )

I want to add a custom wenxin HostingProvider to the hosting_configuration.py file, as shown in the code snippet above. A RestrictModel of the text-embedding type named embedding v1 is added.

However, the Model Provider page shows that there are five available models, not just one embedding v1 as I expected.

How can I solve this problem?

2. Additional context or comments

Changing line 869 of the provider_configuration.py to if m.model not in restrict_model_names: suits my needs.
Is this the right thing to do?
Is there a better solution?

Modified code

class ProviderConfiguration(BaseModel):
...

    def _get_system_provider_models(
        self,
        model_types: list[ModelType],
        provider_instance: ModelProvider,
        model_setting_map: dict[ModelType, dict[str, ModelSettings]],
    ) -> list[ModelWithProviderEntity]:
    ...
            
            # if llm name not in restricted llm list, remove it
            restrict_model_names = [rm.model for rm in restrict_models]
            for m in provider_models:
                # if m.model_type == ModelType.LLM and m.model not in restrict_model_names:
                if m.model not in restrict_model_names:
                    m.status = ModelStatus.NO_PERMISSION
                elif not quota_configuration.is_valid:
                    m.status = ModelStatus.QUOTA_EXCEEDED

Modified effect