API Request Fails with LM Studio's LLM (CORS Enabled) and Unexpected Ollama Errors
Closed this issue · 3 comments
Describe the bug
I encountered the following issues while using LM Studio and would like some assistance. I am running bolt.diy directly on WSL without using Docker.
CORS設定を有効化
As shown in the attached image, I enabled the "CORS" option in LM Studio's server settings. After doing so, I was able to select the LLM (Large Language Model) from the dropdown menu.
APIリクエストエラー
When I selected LM Studio's LLM and attempted to generate code, I received the following error message:
`
There was an error processing your request: An error occurred.
`
Ollama関連の警告
Despite not using the Ollama server, I see these warning messages in the terminal:
‘
WARN Constants Failed to get Ollama models: fetch failed
WARN Constants Failed to get Ollama models: fetch failed
‘
Environment Details:
WSL (Windows Subsystem for Linux): Running bolt.diy directly (without Docker).
I would appreciate any guidance on:
Why the API request to LM Studio's server fails.
Why Ollama warnings appear even though I am not using it.
Thank you for your support!
Link to the Bolt URL that caused the error
Steps to reproduce
I entered the prompt, but I get an error
Expected behavior
I noticed in the screenshot of the debug info that the URL for the chat API in LMStudio seemed different from what LMStudio shows.
Screen Recording / Screenshot
Platform
- OS: [ Linux]
- Browser: [Chrome]
- Version: [eb6d435(v0.0.3) - stable]
Provider Used
No response
Model Used
No response
Additional context
I am getting the following error on the LMStudio server.
2024-12-18 21:50:37 [ERROR] Unexpected endpoint or method. (GET /api/health). Returning 200 anyway
I think the URL that bolt.diy requests to LMStudio's API is different? I thought so.
2024-12-18 20:52:33 [INFO] [LM STUDIO SERVER] Success! HTTP server listening on port 1234
2024-12-18 20:52:33 [INFO]
2024-12-18 20:52:33 [INFO] [LM STUDIO SERVER] Supported endpoints:
2024-12-18 20:52:33 [INFO] [LM STUDIO SERVER] -> GET http://localhost:1234/v1/models
2024-12-18 20:52:33 [INFO] [LM STUDIO SERVER] -> POST http://localhost:1234/v1/chat/completions
2024-12-18 20:52:33 [INFO] [LM STUDIO SERVER] -> POST http://localhost:1234/v1/completions
2024-12-18 20:52:33 [INFO] [LM STUDIO SERVER] -> POST http://localhost:1234/v1/embeddings
in lm studio try to enable: Serve on Local Network
if that dose not work than also try the local ip address of the computer that LM Stuido is running on
In this video I used docker for bolt.diy and ollama but it might help: https://youtu.be/TMvA10zwTbI
Thank you very much.
I have turned on serving on the local network of the LMStudio server.
But the error did not improve.
With the Docker method, I was able to use the Model of LMStudio.
Thank you very much for teaching me how to do this.
Sorry.
I re-cloned the repository this morning.
And I also turned on serving the LM Studio URL on the local network.
I set this URL in env.local and started it up, and it worked with WSL.
I was able to use it with WSL without using Docker.
