VisualStudio Code installation and own llm

[flefevre]

Dear Pythagora team,
thanks for the amazing work.

I want to bring your tools to our french public laboratory by installing gptpilot inside a visualstudiocode and by referencing our internal LLM which is served by liteLLM.

I have got the following error:
There was a problem with request to openai API: API responded with status code: 401. Request token size: 23 tokens. Response text: {"error":{"message":"Missing Authentication header or invalid API key","code":401}}

Setup
I Have tried the following setup without success.

• install visual studio code
• install pythogora plugin
• go to the plugin directory and edit manual the .env file

my .env file
`ENDPOINT=OPENROUTER
OPENAI_ENDPOINT=https://litellm.dev.localhost/chat/completions
OPENAI_API_KEY=dummy
OPENROUTER_API_KEY=dummy

MODEL_NAME=openai/gemma-2b
MAX_TOKENS=8192
`

I have tried with MODEL_NAME=openai/gemma-2b and MODEL_NAME=gemma-2b
When i start vsc, and create a new pythogora project I got the following exception

There was a problem with request to openai API: API responded with status code: 401. Request token size: 23 tokens. Response text: {"error":{"message":"Missing Authentication header or invalid API key","code":401}}

I do not know how to activate more logs to debug it.
I have deactivated the litellm master-key which was only use this if you to require all calls to contain this key (Authorization: Bearer sk-1234). I should at term reactivate it.
I am able to go to https://litellm.dev.localhost/chat/completions and to make a curl request. so it do not come from litellm which seems to be well configured.

curl -X POST https://litellm.dev.localhost/chat/completions -H 'Content-Type: application/json' -d '{ "model": "gemma-2b", "stream": False, "messages": [{ "content": "Hello, how are you?", "role": "user" }]}' {"id":"cmpl-c8f808e876684d62bf0dcb83e694eebf","choices":[{"finish_reason":"stop","index":0,"message":{"content":"I am an AI language model, so I do not have personal feelings or the capacity to experience emotions. I am functioning well and am here to assist you with any questions or tasks you may have. How can I help you today?","role":"assistant"}}],"created":1714913340,"model":"google/gemma-1.1-2b-it","object":"chat.completion","system_fingerprint":null,"usage":{"completion_tokens":48,"prompt_tokens":16,"total_tokens":64}}f

[krrishdholakia]

Hey @flefevre - litellm maintainer here -

Missing Authentication header or invalid API key

does your proxy require an api key? it might be getting rejected b/c of that

---

P.S.: Would love to learn how you're using litellm internally. Attaching my calendly if that helps - https://calendly.com/d/4mp-gd3-k5k/litellm-1-1-onboarding-chat

[mercury64]

I think you only need:

OPENAI_ENDPOINT=https://litellm.dev.localhost

Without the ~/chat/completions