huggingface/transformers

Pull Requests Commits

Remove setting that should be inherited. Formatting.

ivarflakstad committed 160 days ago

Verified ccedf5be

Merge branch 'main' into set-supports-static-cache-false-on-moes

ivarflakstad committed 160 days ago

Verified ee85f488

Fix cache-related tests (#39676)

zucchini-nlp committed 160 days ago

Verified 1c6b4745

Fix Layer device placement in Caches (#39732)

Cyrilvallez committed 160 days ago

Verified fc2bd1ea

Fix `Qwen2AudioForConditionalGeneration.forward()` and `test_flash_attn_kernels_inference_equivalence` (#39503)

ebezzam committed 160 days ago

Verified 7623aa3e

skip `Glm4MoeModelTest::test_torch_compile_for_training` (#39670)

ydshieh committed 160 days ago

Verified 28f26198

Update `QAPipelineTests::test_large_model_course` after #39193 (#39666)

ydshieh committed 160 days ago

Verified 88aed92b

mllama outputs refactor (#39643)

itazap committed 160 days ago

Verified da823fc0

Remove all expired deprecation cycles (#39725)

Cyrilvallez committed 160 days ago

Verified 686bb3b0

[`CI`] Add Eric to comment slow ci (#39601)

vasqu committed 160 days ago

Verified a0fa500a

PATCH: add back n-dim device-mesh + fix tp trainer saving (#39693)

S1ro1 committed 160 days ago

Verified 4c7da9fe

Add self-hosted runner scale set workflow for mi325 CI (#39651)

jitesh-gupta committed 161 days ago

Verified cbede296

[configuration] remove redundant `classmethod` (#38812)

zucchini-nlp committed 161 days ago

Verified b56d7213

update ernie model card (#39657)

jzhang533 committed 161 days ago

Verified 02ea23cb

[processors] add tests for helper fn (#39629)

zucchini-nlp committed 161 days ago

Verified 8b237b86

xpu optimization for generation case (#39573)

sywangyi committed 161 days ago

Verified 6638b364

fix(tokenization): check token.content for trie (#39587)

pjo256 committed 161 days ago

Verified 5c15eb55

Fix missing initialization of `FastSpeech2Conformer` (#39689)

bvantuan committed 161 days ago

Verified 6a61e166

fix missing model._tp_size from ep refactor (#39688)

winglian committed 163 days ago

Verified a6393e7d

More robust tied weight test (#39681)

Cyrilvallez committed 163 days ago

Verified 18a7c29f

dev version 4.55

ArthurZucker committed 163 days ago

c3401d6f

Add padding-free to Granite hybrid moe models (#39677)

garrett361 committed 163 days ago

Verified 97f8c71f

Fix tied weight test (#39680)

Cyrilvallez committed 163 days ago

Verified d6e9f71a

fix break for ckpt without _tp_plan (#39658)

MoyanZitto committed 163 days ago

Verified 5da6ad27

Add EXAONE 4.0 model (#39129)

lgai-exaone committed 163 days ago

Verified c06d4cd6

Support `typing.Literal` as type of tool parameters or return value (#39633)

grf53 committed 163 days ago

Verified 3e4d584a

Add ep (#39501)

ArthurZucker committed 163 days ago

Verified 300d42a4

bad_words_ids no longer slow on mps (#39556)

DWarez committed 163 days ago

Verified abaa043d

Add xlstm model (#39665)

Cyrilvallez committed 163 days ago

Verified 6630c5b7

Use auto_docstring for perception_lm fast image processor (#39679)

yonigozlan committed 163 days ago

Verified ed9a96bc

Older