huggingface/transformers

Pull Requests Commits

add distributed config (#46705)

3outeille committed 5 days ago

Verified 2e2855f0

[Offloading] [Bugfix] Fix disk offloading of models with explicit tensor dtypes (#46849)

kylesayrs committed 5 days ago

Verified 176a17bd

Streamable chat parsing (#45847)

Rocketknight1 committed 5 days ago

Verified 4ab0479e

Round the ue8m0 FP8 scale before quantizing so dequant matches the stored inverse (#46763)

Incheonkirin committed 5 days ago

Verified e32afc72

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

jiqing-feng committed 5 days ago

Verified 22d94091

Fix typos in code (#46579)

cyyever committed 5 days ago

Verified 19d179f2

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

jiqing-feng committed 5 days ago

Verified 6b1baa93

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

Incheonkirin committed 5 days ago

Verified d8c23549

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

remi-or committed 5 days ago

Verified d87f670d

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

Cyrilvallez committed 5 days ago

Verified 8dbac377

Switch decorator order (#46853)

Cyrilvallez committed 6 days ago

Verified 32ec18bb

Fix beam search for mamba models (#46819)

Cyrilvallez committed 6 days ago

Verified 42d7b005

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

efazal committed 6 days ago

Verified c5deba28

Import diffusion_gemma in models init (#46841)

boringcrypto committed 6 days ago

Verified 612c371a

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

stevhliu committed 6 days ago

Verified cd09a2cd

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

Ramshankar07 committed 6 days ago

Verified 892be11b

[Offloading] Support full disk offloading (#46749)

kylesayrs committed 6 days ago

Verified 49f6b78f

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

sharmax-vikas committed 6 days ago

Verified b7c2cca6

Fix mask when generating beyond sliding window (#46839)

zucchini-nlp committed 6 days ago

Verified 5bbed98a

Fix VideoPrismForVideoClassification returning last_hidden_state as h… (#46830)

sharmax-vikas committed 6 days ago

Verified c2fafabb

Add MiniCPM3 (#41116)

bzantium committed 6 days ago

Verified 8964b4b8

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

qgallouedec committed 6 days ago

Verified decfe6a7

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

McPatate committed 6 days ago

Verified 1c75d06e

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

KoichiYasuoka committed 6 days ago

Verified f9bbe64e

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

Anai-Guo committed 6 days ago

Verified 4d47b061

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

vasqu committed 7 days ago

Verified 43639365

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

Abdennacer-Badaoui committed 7 days ago

Verified 8fea7d39

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

Sunt-ing committed 7 days ago

Verified 123f5dd7

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

jiqing-feng committed 7 days ago

Verified f007617e

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

Sunt-ing committed 7 days ago

Verified bfd3604d

Newer Older