huggingface/transformers

Pull Requests Commits

NVFP4: read MXFP8 checkpoints with NVFP4 experts, readers translate formats

IlyasMoutawwakil committed 1 hour ago

005b75d6

NVFP4: one swizzle op on row-major scale params, no core change

IlyasMoutawwakil committed 2 hours ago

12833c57

NVFP4: plan scales instead of special-casing the shard op, review cleanups

IlyasMoutawwakil committed 3 hours ago

9f1f68e7

NVFP4: review cleanups

IlyasMoutawwakil committed 3 hours ago

f016cd4e

NVFP4: build on the quantizer API, port the stack's layout ops, fix TP

IlyasMoutawwakil committed 3 hours ago

6d1360fc

NVFP4: save_pretrained support, parameterized tests

IlyasMoutawwakil committed 4 hours ago

e39ee4b9

NVFP4: MoE experts and ModelOpt exports on finegrained-kernels

IlyasMoutawwakil committed 5 hours ago

35b8576e

[Docs] Add Marathi (मराठी) translation of README (#48746)

shreeyadav06 committed 1 day ago

Verified 6a994bc4

Fix multi-zoom crops in the SAM mask-generation pipeline and related bugs (Fixes #35530) (#49689)

stlukyanenko committed 1 day ago

Verified 0b5cfc26

fix(audio): Route the ALM conversion scripts through from_pretrained (#48525)

harshaljanjani committed 1 day ago

Verified 72021435

fix model parallel bug for molmo2 model (#49816)

kaixuanliu committed 1 day ago

Verified 481e371f

Add Flint model support. (#49824)

lovishmadaan committed 1 day ago

Verified 0ea108da

Remove sdpa support for models using softcapping (#49811)

qgallouedec committed 1 day ago

Verified 7d3ae42e

Fix CLVP `c_proj` residual scaled-init silently never applied (#47478)

ShamSaleem committed 1 day ago

Verified 5fd67cf1

fix failed test cases for unlimited_ocr model (#49806)

kaixuanliu committed 1 day ago

Verified 82307147

Fix dtype cast in TimesFM 2.5 ResidualBlock for quantized weights (#46942)

JOhnsonKC201 committed 1 day ago

Verified 0d65c9cd

Fix transformers env when requests is not installed (#48412)

panshularora committed 1 day ago

Verified e2fec23a

Fix `base_model_prefix` in the PP-OCR detectors, CLIPSeg, TimesFM, CLIP and VideoPrism heads so base checkpoints load (#48871)

blipbyte committed 1 day ago

Verified c2abdef3

Apply EXIF orientation in load_image_as_tensor with torchvision (#49794)

digit50 committed 1 day ago

Verified d38514c6

Fix Fuyu processor test (#49821)

zucchini-nlp committed 1 day ago

Verified a762035c

Fix background_color handling across DeepSeek-VL and Janus image processors (#49767)

tritsystem committed 1 day ago

Verified 3f91ebf8

use torchcodec-xpu and upgrade torch to 2.14.1 (#49814)

sywangyi committed 1 day ago

Verified 6b75aeea

Move the MI300 docker image to ROCm 10.1 and update test expectations (#49460)

Abdennacer-Badaoui committed 1 day ago

Verified 09a71e5b

update some case to add xpu expectation (#49807)

sywangyi committed 1 day ago

Verified 4f71a891

[serge] Fix 2 integration tests regressed by commit c48824e84024 (PR #49404) (#49705)

sergereview[bot] committed 1 day ago

Verified f8900f99

docs: fix OOM/hang for Qwen dynamic resolution models by setting min_pixels and max_pixels (#46007)

poojansatani committed 1 day ago

Verified c20011b1

docs: fix 'allows to' grammar in Speech Encoder Decoder docs (#48727)

mrchatam committed 1 day ago

Verified 12ba8315

docs: fix 'allows to' grammar in DPT docs (#48728)

mrchatam committed 1 day ago

Verified 07cb294b

docs: fix 'allows to' grammar in LiLT docs (#48731)

mrchatam committed 1 day ago

Verified 2f627755

docs: fix prompt formatting and code alignment in Gemma model doc (#48775)

Rattan344 committed 1 day ago

Verified a6a5b4a9

Older