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
- Add support for BoRA
- Astra preprocessing scales activations by max value, not max abs value, can overflow fp16 covariance
- DoRA hotswap raises a bare KeyError for compatible adapters
- DeFT / TinyLoRA: adding a second adapter to a Conv1D layer breaks the layer (TinyLoRA forward crashes, DeFT merge crashes)
- PeftModelForQuestionAnswering silently drops `token_type_ids` and `position_ids`
- aLoRA: speed up forward pass
- delete_adapter skips adapter layers whose module name contains the tuner prefix (e.g. lora_foobar)
- OFT module_dropout has had no effect since #2575: the MultiplicativeDropoutLayer is created but never called
- FSDP2 with `reshard_after_forward=False`: `disable_adapter()` leaves LoRA frozen and `merge_adapter()` is lost at reshard
- `load_adapter` infers `cuda` for adapter weights even when the base model is on CPU — breaks ZeroGPU Spaces with safetensors >= 0.8
- Docs
- not yet supported