huggingface/peft

Pull Requests Commits

CI: Add FSDP tests on multi GPU machine (#2856)

BenjaminBossan committed 13 hours ago

Verified 4f62d700

Refactor layer initialization to use PEFT config directly (#2960)

BenjaminBossan committed 5 days ago

Verified 91aad930

TST Lower tol for regression test for A100 (#2996)

jiqing-feng committed 5 days ago

Verified da1b6d18

FIX Don't warn about unknown layer type when using target parameters (#2997)

BenjaminBossan committed 6 days ago

Verified 622c2821

FEAT Add LoRA-GA (#2926)

sambhavnoobcoder committed 6 days ago

Verified 5e5a1b7d

Add OSF continual learning example (#2897)

NikhilNayak-debug committed 7 days ago

Verified 720d7abd

ENH Caching option for DoRA inferrence (#2661)

BenjaminBossan committed 7 days ago

Verified a33b1789

TST Small clean up regarding weight initialization (#2961)

BenjaminBossan committed 7 days ago

Verified 5b198fd6

TST Fix tolerance issue with GPT-OSS and trf v5 (#2982)

BenjaminBossan committed 7 days ago

Verified bf258301

TST Remove tests that are completely skipped (#2965)

BenjaminBossan committed 7 days ago

Verified 10f92c78

DOC Prefix tuning for encoder-decoder models (#2989)

BenjaminBossan committed 7 days ago

Verified 4f619225

Bump version to 0.18.2.dev0 after release (#2985)

BenjaminBossan committed 7 days ago

Verified d1162c59

DOC Remove useless evals from example (#2994)

maerory committed 7 days ago

Verified f9ed5253

ENH Add Conv1D support to CorDA (#2992)

sambhavnoobcoder committed 8 days ago

Verified 389f5883

ENH XPU support for Cartridges example (#2990)

kaixuanliu committed 8 days ago

Verified bb199f4e

Updated MetaMathQA results (#2984)

githubnemo committed 8 days ago

Verified a9994269

TST Set device map for GPTQ tests (#2979)

jiqing-feng committed 9 days ago

Verified 2cbe6ab3

ENH Weight tying option for trainable tokens (#2870)

sambhavnoobcoder committed 9 days ago

Verified aaa6a50e

CI Upgrade GitHub Actions to latest versions (#2966)

salmanmkc committed 9 days ago

Verified c6be9b2d

FIX Bug in how forgetting metric treats padding (#2986)

BenjaminBossan committed 12 days ago

Verified 8be1a16f

FEAT Convert non-LoRA PEFT adapters to LoRA (#2939)

BenjaminBossan committed 13 days ago

Verified edaab7f2

FIX Don't implicitly require transformers v4.52 (#2976)

BenjaminBossan committed 13 days ago

Verified 9bb8947e

FIX OFT GPTQ forward (#2978)

jiqing-feng committed 13 days ago

Verified 0f3e9f9b

FEAT Add cartridges to PEFT (#2953)

kashif committed 14 days ago

Verified b4750c71

CI For transformers main tests, clear disk space (#2956)

BenjaminBossan committed 34 days ago

Verified 261366de

FIX: Correct adapter dtype with bnb weights (#2893)

BenjaminBossan committed 35 days ago

Verified 89a75aa1

FIX: Inject from state dict into compiled model (#2962)

BenjaminBossan committed 35 days ago

Verified 66ca79d3

FIX Detect if torch.distributed is available (#2963)

vladmandic committed 35 days ago

Verified c65c8861

TST Remove unnecessary prefix tuning dtype test (#2955)

BenjaminBossan committed 37 days ago

Verified e7054226

TST Remove unnecessary PREFIXES constant (#2942)

BenjaminBossan committed 42 days ago

Verified c5a905d0

Older