Generate Commit Message "OpenAI GPT-4 Turbo" Support

[4itech-paul]

Is it possible to incorporate support for OpenAI GPT-4 Turbo into the "Generate Commit Message" feature? This model was recently released.

[ivanscorral]

You can manually set the model to "gpt-4-1106-preview" in VS Code's settings.json like this:

"gitlens.ai.experimental.openai.model": "gpt-4-1106-preview"

I can confirm this works, even if VSCode displays a warning saying that the value is not accepted. You can check this by setting the model to some model that does not exist. This will return a 404 error, meaning that the extension correctly reads the model you manually set here.

Also, around ~10 minutes after executing some requests using this model you should be able to see the usage popping up in https://platform.openai.com/usage.

In my experimentation, GPT-4 Turbo yields much better results than GPT-4, while being much faster and, although more costly than 3.5 Turbo, its quality is far beyond.

Hope this is helpful.

[4itech-paul]

thanks a lot

[eamodio]

In tomorrow's pre-release this will be a supported option as well as refined prompts for better outputs

[github-actions]

This issue has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.