Pull Requests huggingface/peft

Fix two issues introduced in AutoGPTQ deprecation

#3014 by githubnemo was merged 2026-02-02 10:53

fix: eliminate cross-terms bug when combining adapters with different weights

#3013 by wingding12 was closed 2026-03-18 13:25

Ignore BPE errors in tests

#3011 by githubnemo was merged 2026-01-29 12:48

Fix fbgemm exception in nightly CI

#3010 by githubnemo was merged 2026-01-29 10:45

Upgrade GitHub Actions for Node 24 compatibility

#3008 by salmanmkc was merged 2026-02-12 11:46

[LoRA] Document support for effective rank for LoRA on MOE experts

#3007 by kashif was merged 2026-01-27 16:30

Fix initialization bug introduced in #2960

#3006 by githubnemo was merged 2026-01-29 12:48

[feature] Tiny modification to enable OFT for finetuning embedding layers

#3005 by zqiu24 was merged 2026-03-10 10:08

Bugfix turned into restructuring

#3003 by githubnemo was merged 2026-01-22 18:10

Fix error of PEFT with disable adapters and FSDP

#3001 by Isalia20 was merged 2026-02-19 12:36

Intruder dimension reduction for LoRA

#2999 by githubnemo was merged 2026-01-26 17:16

FIX Don't warn about unknown layer type when using target parameters

#2997 by BenjaminBossan was merged 2026-01-15 14:52

lower tol for specific test

#2996 by jiqing-feng was merged 2026-01-16 10:48

[WIP] Support transformers weight conversion

#2995 by BenjaminBossan was closed 2026-02-27 13:35

Update prompt_based_methods.md - remove eval_preds

#2994 by maerory was merged 2026-01-14 10:20

Add Conv1D support to CoRDA for GPT-2 compatibility #2991

#2992 by sambhavnoobcoder was merged 2026-01-13 17:27

add Intel XPU platform support for cartridge_self_study example

#2990 by kaixuanliu was merged 2026-01-13 15:39

DOC Prefix tuning for encoder-decoder models

#2989 by BenjaminBossan was merged 2026-01-14 15:44

FIX Bug in how forgetting metric treats padding

#2986 by BenjaminBossan was merged 2026-01-09 17:01

Bump version to 0.18.2.dev0 after release

#2985 by BenjaminBossan was merged 2026-01-14 15:43

Updated MetaMathQA results

#2984 by githubnemo was merged 2026-01-13 12:21

Patch release 0.18.1

#2983 by BenjaminBossan was merged 2026-01-09 12:42

TST Fix tolerance issue with GPT-OSS and transformers v5

#2982 by BenjaminBossan was merged 2026-01-14 15:45

Testing Transformers v5 RC 2

#2981 by BenjaminBossan was closed 2026-01-08 11:06

fix device map check

#2979 by jiqing-feng was merged 2026-01-12 11:20

fix oft gptq forward

#2978 by jiqing-feng was merged 2026-01-08 12:39

FIX Don't implicitly require transformers v4.52

#2976 by BenjaminBossan was merged 2026-01-08 15:19

fix: auto-untie word embeddings on merge_and_unload when both are adapted

#2972 by www-spam was closed 2026-03-17 15:24

Implements EWoRA in PEFT

#2971 by HarshKohli was closed 2026-03-16 15:25

intorduce AdaDoRA implementation using both DoRA and AdaLoRa together

#2969 by OrMullerHahitti was closed 2026-03-22 15:06