Added azure api version patch

[hargunmujral]

Describe the changes you have made:

• Azure openai uses an "api version" parameter when running open interpreter. Thus, added this as a param in the setup file
• Fixed broken link to CONTRIBUTING.md in the README.md

Reference any relevant issue (Fixes #000)

• I have performed a self-review of my code:

I have tested the code on the following OS:

• Windows
• MacOS
• Linux

AI Language Model (if applicable)

• GPT4
• GPT3
• Llama 7B
• Llama 13B
• Llama 34B
• Huggingface model (Please specify which one)

[Notnaton]

Should this have a --api_version cli argument?

[KillianLucas]

Nice @hargunmujral. We should start moving in this direction of exposing more params via the Python package (and via CLI flags as @Notnaton mentioned, which I'll do after merging). Especially with the new config setup that @Notnaton is building, I think this takes us in the right direction. Merging!

[KillianLucas]

Also great catch on the CONTRIBUTING link!

[symonh]

Thanks everyone. I updated interpreter (pip install...) and then export AZURE_API_VERSION=2023-08-01-preview. Now when I run interpreter I get this error:

APIError(status_code=500, message=str(original_exception), llm_provider=custom_llm_provider, model=model)
litellm.exceptions.APIError: LLM Provider NOT provided. Pass in the LLM provider you are trying to call. E.g. For 'Huggingface' inference endpoints pass in `completion(model='huggingface/',..)` Learn more: https://docs.litellm.ai/docs/providers

Following the litellm docs, I tried running interpreter --model=azure/nameai-63k-tpm

This loads open interpreter, but when I pass a message to it I get

openai.error.InvalidRequestError: Invalid URL (POST /v1/openai/deployments/nameai-63k-tpm/chat/completions)

My config looks like this:

local: false
temperature: 1
context_window: 31000
max_tokens: 3000
OPENAI_API_KEY: 97-*******************84
API_BASE: https://nameai.openai.azure.com/
API_TYPE: azure
MODEL: azure/nameai-63k-tpm
AZURE_API_VERSION: 2023-08-01-preview

I've played with setting API_BASE to https://nameai.openai.azure.com/openai/deployments/nameai-63k-tpm — but I get the same openai.error.InvalidRequestError: Invalid URL (POST /v1/openai/deployments/nameai-63k-tpm/chat/completions).

Thank you for this project and for any help you can provide — this will be an absolute dream when I get it working!

[Notnaton]

@symonh use lower case in the config file and try again

[symonh]

@Notnaton Thank you very much — it's working!

When analyzing data, how good should I expect the performance to be compared to ChatGPT? Roughly on par, minus the limitations from packages and compute, or less capable?

I just asked it to reshape a dataset and it performed beautifully :)