huggingface/transformers

Pull Requests Commits

AutoGrad support for grouped_mm fallback (#44152)

IlyasMoutawwakil committed 51 days ago

Verified 9a7d8178

Patch `__setitem__` on `ModelOutput` even if the parameter was previously `None` (#44080)

tomaarsen committed 51 days ago

Verified 00cc937c

[`simple`] Fix up `__repr__` whitespace/brackets (#44048)

tomaarsen committed 51 days ago

Verified e0b07d1b

[`chore`] Fix incorrect forward type hint for Gemma3n (#44051)

tomaarsen committed 51 days ago

Verified 64230197

Raise informative error when loading video processors (#44125)

zucchini-nlp committed 51 days ago

Verified 1f468a8b

fix(flaky): Different approach to make sure loss exists (#43804)

tarekziade committed 51 days ago

Verified 6b55285d

Add Higgs Audio V2 Model (#40294)

szhengac committed 51 days ago

Verified 1618d44b

[voxtral] fix voxtral proc (#44132)

eustlb committed 51 days ago

Verified 1e31876d

[docs] Fix typos in GenerationConfig docstring (#44143)

nightcityblade committed 52 days ago

Verified 35324377

Fix gemma3n `get_audio_features` (#44040)

zucchini-nlp committed 52 days ago

Verified 4aa550b4

Fix UMT5EncoderModel embedding weights not being tied after loading (#43880)

jiqing-feng committed 52 days ago

Verified a6ef2a6f

fix(testing): Update stale device override test in GraniteSpeech (#44113)

harshaljanjani committed 52 days ago

Verified b240061e

[Misc][vlms] Use text_config when initializing the fine-grained FP8Expert (#44032)

JJJYmmm committed 52 days ago

Verified 196216c3

docs: fix typo 'AuoQuant' → 'AutoQuant' and clarify FINEGRAINED_FP8 library column (#44131)

cluster2600 committed 52 days ago

Verified 556312cd

Update post proc (#44090)

itazap committed 52 days ago

Verified 549835e5

Fix: flaky `Kosmos2ModelTest` test (#44061)

tarekziade committed 52 days ago

Verified 6976f757

AutoTokenizer ignores config when model_type is None (#44127)

itazap committed 53 days ago

Verified 398780d3

Migrate GPT2 to standardized output capture decorators (#43983)

Aki-07 committed 53 days ago

Verified 3689287d

Simplify input preparation in generate (#44126)

Cyrilvallez committed 53 days ago

Verified d6e53e26

`grouped_mm` fallback (#44043)

IlyasMoutawwakil committed 53 days ago

Verified 39ddb88a

Bump dev version (#44099)

qgallouedec committed 53 days ago

Verified ddfe1c5f

audio text to text task guide (#43413)

Deep-unlearning committed 53 days ago

Verified ffbea2db

Fix loading logic issue (#44095)

Cyrilvallez committed 53 days ago

Verified a64996e3

[docs] tokenizer summary (#43965)

stevhliu committed 53 days ago

Verified 83eb94c1

[docs] refactor tokenizer docs (#43900)

stevhliu committed 53 days ago

Verified 0e5516d6

Fix percentage formatting in help messages for gradient checkpointing, Liger Kernel, and empty cache steps (#44100)

qgallouedec committed 53 days ago

Verified 4e0573b3

[docs] customizing tokenizers (#43929)

stevhliu committed 53 days ago

b19817bc

Merge test_keep_in_fp32_modules and test_keep_in_fp32_modules_strict (#44097)

Rocketknight1 committed 53 days ago

Verified 35c5d2a0

[voxtral-realtime] update runner expected values (#44096)

eustlb committed 53 days ago

Verified 1f76bd4d

Use torch.isfinite (#44069)

cyyever committed 53 days ago

Verified 66b40e69

Newer Older