config.json file not working properly

[hiofficial4477]

Before submitting your bug report

• [X] I believe this is a bug. I'll try to join the Continue Discord for questions
• [X] I'm not able to find an open issue that reports the same bug
• [X] I've seen the troubleshooting guide on the Continue Docs

Relevant environment info

Error streaming response: Cannot read properties of undefined (reading 'model')

Description

config.json file I accidentally deleted, and now after reinstalling the plugin also config.json file is coming blank/empty. Can I get this file

To reproduce

No response

Log output

No response

[hiofficial4477]

it got fixed you need to delete the config.json file from the installed location

and then connect with the preferred API key

so default file is

{ "models": [ { "model": "claude-3-5-sonnet-latest", "provider": "anthropic", "apiKey": "", "title": "Claude 3.5 Sonnet" }, { "model": "gpt-3.5-turbo", "title": "GPT-3.5-Turbo", "apiKey": Your API key", "provider": "openai" } ], "tabAutocompleteModel": { "title": "Codestral", "provider": "mistral", "model": "codestral-latest", "apiKey": "" }, "customCommands": [ { "name": "test", "prompt": "{{{ input }}}\n\nWrite a comprehensive set of unit tests for the selected code. It should setup, run tests that check for correctness including important edge cases, and teardown. Ensure that the tests are complete and sophisticated. Give the tests just as chat output, don't edit any file.", "description": "Write unit tests for highlighted code" } ], "contextProviders": [ { "name": "code", "params": {} }, { "name": "docs", "params": {} }, { "name": "diff", "params": {} }, { "name": "terminal", "params": {} }, { "name": "problems", "params": {} }, { "name": "folder", "params": {} }, { "name": "codebase", "params": {} } ], "slashCommands": [ { "name": "share", "description": "Export the current chat session to markdown" }, { "name": "cmd", "description": "Generate a shell command" }, { "name": "commit", "description": "Generate a git commit message" } ] }

[hiofficial4477]

Now New error its showing

HTTP 401 Unauthorized from https://api.openai.com/v1/chat/completions

[hiofficial4477]

it is also fixed just by changing the open ai key , and setting up ollama locally with local gpt

[micuentadecasa]

I'm having the same issue, I was adding lm studio and started giving errors.

[ducphamle2]

In my case, Continue can't detect my custom self-hosted models

how to fix it guys?

[ducphamle2]

It just doesnt show anything, even after I "Add Chat model"

[micuentadecasa]

I have tried a lot of things, as @ducphamle2 say to add chat model, rewriting the config file (I saved the config for lm studio that was working), delete the file, use the content of the file referred on this thread, etc, and it doesn't works, in my case I did a update of vs code, but I'm not totally sure if the error started there.

[hiofficial4477]

I got it fixed first remove everything from the config.json file

Just leave it as { } >> save this config.json file >> uninstall your extension >> restart your VS code editor >> install again the continue dev extension >> then first try to add the open ai or any other LLM key first before opening your config file >> it will generate config.json file automatically according to your provided input

[hiofficial4477]

After everything works I will give you an error if you have added an open API key and not the other key mentioned in the config.json file just remove other models from there andwill work fine + in the autocomplete section of the config file you have to install the ollama model + model and change the name according to ollama in autocomplete section would look like this

"tabAutocompleteModel": { "title": "qwen2.5-coder", "provider": "ollama", "model": "qwen2.5-coder:32b", },

I am using qwen2.5 coder 32 billion parameter model that is better than Gpt4 for auto complete coding sections

Hope it help's :)

[micuentadecasa]

I got it fixed first remove everything from the config.json file

Just leave it as { } >> save this config.json file >> uninstall your extension >> restart your VS code editor >> install again the continue dev extension >> then first try to add the open ai or any other LLM key first before opening your config file >> it will generate config.json file automatically according to your provided input

This solution didn't work for me

[ducphamle2]

I have tried a lot of things, as @ducphamle2 say to add chat model, rewriting the config file (I saved the config for lm studio that was working), delete the file, use the content of the file referred on this thread, etc, and it doesn't works, in my case I did a update of vs code, but I'm not totally sure if the error started there.

same here. I have tried everything as well. Sometimes it works sometimes it just doesn't. It's pure luck afak.

Also, it only works for one VS Code project, not multiple ones.

Another thing is that the @codebase indexing feature is so slow.

[hiofficial4477]

can you share some screenshots

[micuentadecasa]

the screenshot that @ducphamle2 shared is what I see, continue doesn´t load properly the config file therefore the "select model" is visible in the chat window, and even adding a new model doesn´t works

[tomasz-stefaniak]

I'm wondering if what @micuentadecasa is saying is correct - it could be an unrelated bug that's breaking Continue and preventing the config from loading correctly. Normally, if the config is misconfigured you should get an error like this in the UI:

If your problem is that the config.json file is empty and you don't know how to regenerate it, here is an example of a functional config.json:

{
  "models": [
    {
      "title": "Codegemma",
      "provider": "ollama",
      "model": "codegemma:2b"
    },
    {
      "model": "claude-3-5-sonnet-20240620",
      "provider": "anthropic",
      "apiKey": "",
      "title": "Claude 3.5 Sonnet"
    },
    {
      "title": "Starcoder",
      "provider": "ollama",
      "model": "starcoder2:3b"
    }
  ],
  "tabAutocompleteModel": {
    "title": "Starcoder",
    "provider": "ollama",
    "model": "starcoder2:3b"
  },
  "tabAutocompleteOptions": {
    "useCache": false
    // "transform": false
  },
  "customCommands": [
    {
      "name": "test",
      "prompt": "{{{ input }}}\n\nWrite a comprehensive set of unit tests for the selected code. It should setup, run tests that check for correctness including important edge cases, and teardown. Ensure that the tests are complete and sophisticated. Give the tests just as chat output, don't edit any file.",
      "description": "Write unit tests for highlighted code"
    }
  ],
  "contextProviders": [
    {
      "name": "code",
      "params": {}
    },
    {
      "name": "docs",
      "params": {}
    },
    {
      "name": "diff",
      "params": {}
    },
    {
      "name": "terminal",
      "params": {}
    },
    {
      "name": "problems",
      "params": {}
    },
    {
      "name": "folder",
      "params": {}
    },
    {
      "name": "codebase",
      "params": {}
    }
  ],
  "slashCommands": [
    {
      "name": "edit",
      "description": "Edit selected code"
    },
    {
      "name": "comment",
      "description": "Write comments for the selected code"
    },
    {
      "name": "share",
      "description": "Export the current chat session to markdown"
    },
    {
      "name": "cmd",
      "description": "Generate a shell command"
    },
    {
      "name": "commit",
      "description": "Generate a git commit message"
    }
  ],
  "docs": []
}

[hiofficial4477]

@micuentadecasa - I looked into the issue and to be honest if nothing works out for you.

I suggest cleaning and uninstalling the whole vs code. And its extension file from C > user directory also

and if you save some backup on Github or Mikroosft I would suggest taking the backup of it and also removing the resources from there.

But tbh, I personally tried all of these, and it doesn't work, for me.

So this is what works out for me...

You can try out the default config.json file given by @tomasz-stefaniak. Then, try to add 1 local model from oolama and auto-detect the model for it. But remember to provide the correct llama destination in your environment variable; otherwise, it won't work properly. It should look like this (image attached below) to save any llama model at any location in your PC/Mac. And remember after your continue dev extension works properly with your local LLM model, any other online LLM would also work perfectly it can any model like open ai or any other listed model on the extension website.

and in your config.json file add this line somewhere at the start of the code

{ "model": "AUTODETECT", "title": "Autodetect", "provider": "ollama" }

cheers...

[ducphamle2]

@micuentadecasa - I looked into the issue and to be honest if nothing works out for you.

I suggest cleaning and uninstalling the whole vs code. And its extension file from C > user directory also

and if you save some backup on Github or Mikroosft I would suggest taking the backup of it and also removing the resources from there.

But tbh, I personally tried all of these, and it doesn't work, for me.

So this is what works out for me...

You can try out the default config.json file given by @tomasz-stefaniak. Then, try to add 1 local model from oolama and auto-detect the model for it. But remember to provide the correct llama destination in your environment variable; otherwise, it won't work properly. It should look like this (image attached below) to save any llama model at any location in your PC/Mac. And remember after your continue dev extension works properly with your local LLM model, any other online LLM would also work perfectly it can any model like open ai or any other listed model on the extension website.

and in your config.json file add this line somewhere at the start of the code

{

  "model": "AUTODETECT",

  "title": "Autodetect",

  "provider": "ollama"

}

cheers...

I tried this on Mac, and it worked at first, but if I played with it by changing the models or adding new workspaces -> went back to the same problem. It's the only bug that's preventing me from using Continue

[tomasz-stefaniak]

@ducphamle2 could you try starting from scratch again and adding a single new model, then checking if it works? If the error occurs again, could you share your config.json with that one additional model added?

If changing the model breaks Continue somehow, could you attach a short video reproduction?

Thanks!

[micuentadecasa]

I tried everything including removing the plugin (removing manually the folder of the plugin, etc.) and nothing works. As @ducphamle2 commented this is stopping us to use it.

[ducphamle2]

@ducphamle2 could you try starting from scratch again and adding a single new model, then checking if it works? If the error occurs again, could you share your config.json with that one additional model added?

If changing the model breaks Continue somehow, could you attach a short video reproduction?

Thanks!

Sure lemme prepare a short video to demonstrate