chatglm add API URL

[carllisicau]

Self Checks

• This is only for bug report, if you would like to ask a question, please head to Discussions.
• 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).
• Please do not modify this template :) and fill in all the required fields.

Dify version

0.6.8

Cloud or Self Hosted

Self Hosted (Docker)

Steps to reproduce

i want to add API URL to chatglm
but occur a 404
log show
HTTP Request: POST http://10.15.12.xxx:18504/v1/chat/completions "HTTP/1.1 404 Not Found"
i just input http://10.15.12.xxx:18504
where add the suffix URL like /v1/chat/completions?
our local chatglm llm api like /v2/chat/completions
so i must find the code replace /v1/chat/completions to /v2/chat/completions
where the code?

✔️ Expected Behavior

none

❌ Actual Behavior

none

[dosubot[bot]]

To address the issue of changing the API URL path from /v1/chat/completions to /v2/chat/completions in your self-hosted Docker setup for Dify 0.6.8, you'll need to modify the code where the endpoint URL is constructed. This adjustment is necessary for the OAIAPICompatLargeLanguageModel class within the llm.py file, specifically in the _generate method that handles API requests.

Here's a concise guide on what to do:

1. Locate the llm.py file in your Dify installation.
2. Find the _generate method within the OAIAPICompatLargeLanguageModel class.
3. Look for the section of code that constructs the endpoint_url for chat completions. It will look something like this:

if completion_type is LLMMode.CHAT:
    endpoint_url = urljoin(endpoint_url, 'chat/completions')

4. Modify this line to point to the new v2 endpoint:

if completion_type is LLMMode.CHAT:
    endpoint_url = urljoin(endpoint_url, 'v2/chat/completions')

This change ensures that all chat completion requests are directed to the correct /v2/chat/completions endpoint, resolving the 404 error you've encountered.

Make sure your chatglm setup is configured to handle requests at this new endpoint. After making this modification, your self-hosted Dify application should correctly interact with the chatglm API at the updated path.

Details

Sources

• en/user-guide/launching-dify-apps/developing-with-apis/README.md
• api/core/model_runtime/model_providers/openai_api_compatible/llm/llm.py
• en/getting-started/install-self-hosted/environments.md
• api/tests/unit_tests/libs/test_yarl.py
• api/core/model_runtime/model_providers/deepseek/llm/llm.py
• sdks/nodejs-client/index.test.js
• api/controllers/console/app/error.py
• sdks/nodejs-client/index.js
• sdks/nodejs-client/index.d.ts
• sdks/php-client/dify-client.php
• zh_CN/guides/application-publishing/developing-with-apis.md

About Dosu
This response is meant to be useful and save you time. It is not meant to be a precise solution, but rather a starting point for your own research.

Help me be more useful! Please leave a 👍 if this is helpful and 👎 if it is irrelevant.

To continue the conversation, mention @dosu.