huggingface/transformers

Pull Requests Commits

Fix BitNet packed-weight unpacking dtype (`F.linear` dtype mismatch) (#46808)

jiqing-feng committed 45 days ago

Verified 22d94091

Fix typos in code (#46579)

cyyever committed 45 days ago

Verified 19d179f2

Fix Evolla eager attention for the GQA text decoder (#46860)

jiqing-feng committed 45 days ago

Verified 6b1baa93

Fix Moonshine training-loss double-shift (train against labels, not labels[..., 1:]) (#46784)

Incheonkirin committed 45 days ago

Verified d8c23549

[CB] Fix issues with FA read / writes (#46765)

remi-or committed 45 days ago

Verified d87f670d

Fix expanded dim in the cache's cumulative length (#46856)

Cyrilvallez committed 45 days ago

Verified 8dbac377

Switch decorator order (#46853)

Cyrilvallez committed 45 days ago

Verified 32ec18bb

Fix beam search for mamba models (#46819)

Cyrilvallez committed 45 days ago

Verified 42d7b005

docs(trainer): add JIT checkpointing to trainer recipes (#46826)

efazal committed 45 days ago

Verified c5deba28

Import diffusion_gemma in models init (#46841)

bartjellema committed 46 days ago

Verified 612c371a

[skills] help your agent get started (#45732)

stevhliu committed 46 days ago

Verified cd09a2cd

Fix use_cache with seq_len > 1 ( #46032) (#46084)

Ramshankar07 committed 46 days ago

Verified 892be11b

[Offloading] Support full disk offloading (#46749)

kylesayrs committed 46 days ago

Verified 49f6b78f

fix: raise `ValueError` for empty conversation in `apply_chat_template` (#46753)

sharmax-vikas committed 46 days ago

Verified b7c2cca6

Fix mask when generating beyond sliding window (#46839)

zucchini-nlp committed 46 days ago

Verified 5bbed98a

Fix VideoPrismForVideoClassification returning last_hidden_state as h… (#46830)

sharmax-vikas committed 46 days ago

Verified c2fafabb

Add MiniCPM3 (#41116)

bzantium committed 46 days ago

Verified 8964b4b8

Avoid NumPy 2.0 `__array__` copy-keyword deprecation in `create_mm_token_type_ids` (#46827)

qgallouedec committed 46 days ago

Verified decfe6a7

docs: update apple silicon doc with safetensors `0.8.0` benefits (#46744)

McPatate committed 46 days ago

Verified 1c75d06e

[DiffusionGemma] return_dict_in_generate=False for image-text-to-text… (#46724)

KoichiYasuoka committed 46 days ago

Verified f9bbe64e

fix(cli/serve): import serve handlers lazily so the CLI works without the `serve` extra (#46473)

Anai-Guo committed 46 days ago

Verified 4d47b061

[`CB`] Add FA2 to the fast path (#46729)

vasqu committed 46 days ago

Verified 43639365

update kernels-community/aiter-rope version (#46810)

Abdennacer-Badaoui committed 46 days ago

Verified 8fea7d39

Fix prompt lookup decoding crash when no EOS token is configured (#46790)

Sunt-ing committed 47 days ago

Verified 123f5dd7

Fix flex_attention block mask creation when `get_seq_length` returns a tensor (#46802)

jiqing-feng committed 47 days ago

Verified f007617e

Fix left-padding token selection in `BioGptForSequenceClassification` (#46782)

Sunt-ing committed 47 days ago

Verified bfd3604d

Fix broken internal links in model documentation (#46807)

ShamSaleem committed 47 days ago

Verified be326dbd

DiffusionGemma: mask layout and CI (#46654)

zucchini-nlp committed 47 days ago

Verified 14917fc5

Use cached added-token dicts in per-token decode loops (#46535)

ishan-1010 committed 47 days ago

Verified 7f61d468

Fix offloaded cache device mismatch on hybrid models (#46748)

Sunt-ing committed 47 days ago

Verified 6446f1fc

Newer Older