[docs] README update #1411
[docs] README update #1411
Conversation
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
There was a problem hiding this comment.
Thanks so much @stevhliu for this!
I left few suggestions to reflect the README with recent resources such as DPO, etc. Please let me know what do you think
There was a problem hiding this comment.
Thanks a lot for this update of the README. It makes the README much more approachable than it was before.
Apart from the merge conflict, I wonder if it would make sense to link to our docs instead of/in addition to the papers in the section about supported methods. I could imagine that many users are more interested in the PEFT docs than the paper, and the docs link to the papers too. But this is just a suggestion, so no need to do it now.
|
Good idea, I think it's better to link users to existing resources instead of maintaining another list :) |
* update * feedback * feedback
This PR cleans up the
README.mdas a whole:➕ adds a tip to promote and watch the PEFT Hub organization more for updates on the latest supported PEFT methods and notebooks
➕ expands quickstart a bit with installation, training, and inference code snippets
➕ creates two new sections, Why you should use PEFT (this focuses on the benefits and use cases) and PEFT integrations (this focuses on how PEFT fits into the 🤗 ecosystem)
➕ adds the Space we have in the docs for discovering supporting models (more compact than the current model matrix)
➖ removes specific code snippets of using Accelerate and DreamBooth training, caveats, injecting adapter types, and mixing different adapter types (all of this info can already be found in the docs)