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
- Silently slower delta-rule fallback when training a Qwen3.8-27B LoRA with examples/sft under FSDP on 8x B200
- docs: fix 'allows to' grammar in low-level API guide
- PeftModel.set_adapter raises opaque TypeError: unhashable type: 'list' instead of a clear error
- FEAT Add Astra initialization for LoRA
- target_parameters docs: broken example and a wrong regex claim
- LoRA + torchao: merging several adapters in one call raises UnboundLocalError
- [perf] Qwen3.8-27B LoRA SFT with FSDP (examples/sft) on 8x B200: 52.8 s to 34.7 s per optimizer step, mostly from the pure-torch delta-rule fallback
- Failed adapter injection leaves orphaned adapter layers behind
- Pooled thread: safe merging not working properly
- Transformers v5 MoE conversion incorrectly rewrites fully qualified shared-expert LoRA targets
- Docs
- not yet supported