Do not hardcode gptel model

armindarvish / consult-web

Powerful Web and Omni Search inside Emacs

https://github.com/armindarvish/consult-web

86 stars 6 forks source link

Do not hardcode gptel model #21

Closed benthamite closed 6 months ago

benthamite commented 6 months ago

Currently, when no model parameter is passed to consult-web--chatgpt-fetch-results, "gpt-3.5-turbo" is used. This pull request makes the function instead use the value of gptel-model, allowing the user to set the model to whichever one they have set in their gptel configuration.

armindarvish commented 6 months ago

@benthamite, Read the section about contributing: https://github.com/armindarvish/consult-web?tab=readme-ov-file#contributions

I am using literate programming with one org file as the single source of truth. I cannot merge your pull request because it is not changing the org file! I implemented this in #22 and merged it in develop and main, and am closing your pull request. Next time, please change the org file and tangle it instead of directly editing the *.el files.