Error decoding response body: expected value at line 1 column 1

[jalalirs]

I am trying to use llm-vscode with a locally deployed Text Generation Inference (TGI) server but I keep getting the following error:

Error decoding response body: expected value at line 1 column 1

My setting is the following where and correspond to my server path. I tried both with /generate and without it

{
    "editor.accessibilitySupport": "off",
    "workbench.colorTheme": "Default Dark+",
    "git.openRepositoryInParentFolders": "always",
    "diffEditor.codeLens": true,
    "llm.attributionEndpoint": "http:https://<host>:<port>/generate",
    "llm.configTemplate": "Custom",
    "llm.modelIdOrEndpoint": "http:https:// <host>:<port>/generate",
    "llm.fillInTheMiddle.enabled": true,
    "llm.fillInTheMiddle.prefix": "<PRE> ",
    "llm.fillInTheMiddle.middle": " <MID>",
    "llm.fillInTheMiddle.suffix": " <SUF>",
    "llm.temperature": 0.2,
    "llm.contextWindow": 4096,
    "llm.tokensToClear": [
        "<EOT>"
    ],
    "llm.enableAutoSuggest": true,
    "llm.documentFilter": {
 

    },
    "llm.tlsSkipVerifyInsecure": true
}

[github-actions]

This issue is stale because it has been open for 30 days with no activity.

[KeyvNari]

Hello, I also get the same error. Did you find a solution for it?

[jalalirs]

Unfortunately not yet and I am still waiting for an answer or an update

[McPatate]

Hello, did you check the TGI logs? I assume the response body is not formatted correctly, there may be an issue with the way the response is parsed.

[jalalirs]

The TGI output is fine and can be consumed both by langchain and chat-ui. TGI with codellama-34 can be consumed fine with python requests call.

[github-actions]

This issue is stale because it has been open for 30 days with no activity.