peft
https://github.com/huggingface/peft
Triage Issues!
When you volunteer to triage issues, you'll receive an email each day with a link to an open issue that needs help in this project. You'll also receive instructions on how to triage issues.
Triage Docs!
Receive a documented method or class from your favorite GitHub repos in your inbox every day. If you're really pro, receive undocumented methods or classes and supercharge your commit history.
not yet supported1 Subscribers
Add a CodeTriage badge to peft
Help out
- Issues
- merge_adapter()/unmerge_adapter() should raise a helpful error for prompt learning and adaption prompt (follow-up to #4077)
- feat(peft_model): raise informative error on merge/unmerge for prompt learning and adaption prompt
- delete_adapter breaks the model when the active adapter targets different modules than the remaining one
- FIX layers_to_transform without layers_pattern for layers at the model root
- FEAT Check zero embedding
- Feat - healthcheck: warn when trainable parameters are not saved with the adapter
- LoRA on MoE experts (target_parameters) fails under FSDP + expert parallelism
- get_additive_delta / convert_to_lora ignore the DoRA magnitude after training
- ties and dare_ties elect the majority sign before applying adapter weights
- convert_to_lora silently converts variants that cannot be merged (aLoRA), producing a non-equivalent adapter
- Docs
- not yet supported