huggingface/transformers

Pull Requests Commits

update kernel name

MekkCyber committed 67 days ago

9580359b

fix prepare_config_and_inputs_for_common bug in llava test (#41942)

yao-matrix committed 67 days ago

Verified 90d1b67d

Fix: Gemma3TextConfig rope scaling assignments (#41934)

RyanMullins committed 68 days ago

Verified 02c324f4

Fix rope_parameters for gemma3 weights conversion script (#41922)

douglas-reid committed 68 days ago

Verified b47b3563

fix some ut failures on XPU w/ torch 2.9 (#41941)

yao-matrix committed 68 days ago

Verified e7e7eca0

Minor fix in docker image build workflow (#41949)

ydshieh committed 68 days ago

Verified cad7eeeb

Cache latest pytorch amd image locally on mi325 CI runner cluster (#41926)

jitesh-gupta committed 69 days ago

Verified 76fc50a1

fix some ut failures on XPU w/ torch 2.9 (#41923)

yao-matrix committed 69 days ago

Verified a43b36cf

Update some workflow files (#41892)

ydshieh committed 69 days ago

Verified 10d55712

Fix Florence2 conversion script model_type KeyError (#41866)

i3hz committed 69 days ago

Verified 259d174e

Allow parse_response to accept token IDs (#41849)

Rocketknight1 committed 69 days ago

Verified 38df1e94

Fix invalid examples in QwenVL model docstrings and add Qwen3VL example (#41812)

Xqle committed 69 days ago

Verified 5462376a

Add 6 huggingface notebooks on AMD dev cloud (#41883)

fan-amd committed 69 days ago

Verified e6142ad8

evaluate>=0.4.6 is needed (#41920)

stas00 committed 69 days ago

Verified 21dfd6e7

speed up loading checkpoints for zero stage 3 (#41850)

ri938 committed 69 days ago

Verified b22d0d07

Fix: avoid duplicate token in maybe_load_adapters (#41903)

luaenrique committed 70 days ago

Verified 4d0b6758

make lfm2_moe integration test pass on XPU (#41796)

yao-matrix committed 70 days ago

Verified 2f9e3ae7

revert changes in _is_package_available (#41891)

MekkCyber committed 71 days ago

Verified 1f0b490a

Fix installation cmds in docs (#41887)

yaswanth19 committed 71 days ago

Verified 8472ac68

Fix torch.no_grad decorator in VLMS (#41888)

yaswanth19 committed 71 days ago

Verified bf917156

Adds Universal Intelligence to awesome transformers documentation

victor-bluera committed 73 days ago

77e8b9f8

CI workflow for Flash Attn (#41857)

ydshieh committed 73 days ago

Verified e2e8dbed

:rotating_light: [`Clip`] Fix masking and enable flash attention on all model types (#41750)

vasqu committed 74 days ago

Verified 7a833d1c

Fix TypeError: find_adapter_config_file() got an unexpected keyword argument '_adapter_model_path' (#41604)

albertvillanova committed 74 days ago

Verified 8bde822a

Share embedding modules in BART, not only weights (#41821)

githubnemo committed 74 days ago

Verified 9bb51b31

Fix const parsing for dict inputs in chat schemas (#41824)

Rocketknight1 committed 74 days ago

Verified 090a8946

Fix Qwen2Audio flash attention mask format for generation (#41843)

Abdennacer-Badaoui committed 74 days ago

Verified 4faf6752

Fix MXFP4 quantizer to support variable num_local_experts and hidden_size (#41795)

marksverdhei committed 74 days ago

Verified bb6028cb

Remove redundant code from Qwen3VLProcessor (#41836)

Xqle committed 74 days ago

Verified 7935b869

further reducing flakiness in `utils/check_bad_commit.py` (#41658) (#41815)

ydshieh committed 74 days ago

Verified c27efe6e

Older