huggingface/transformers

Pull Requests Commits

use `TokenizersBackend` (#42894)

ArthurZucker committed 138 days ago

Verified 9daee2e8

Fix failing `DebertaV2ModelIntegrationTest` (#43136)

Sai-Suraj-27 committed 138 days ago

Verified 69ec61f7

Fixed failing `BioGPT` batch generation test (#43128)

Sai-Suraj-27 committed 138 days ago

Verified 6d7b10d5

Use timm-side buffers initialization (#43124)

Cyrilvallez committed 138 days ago

Verified 8dd9c999

Optimize MoEs for decoding using batched_mm (#43126)

IlyasMoutawwakil committed 138 days ago

Verified 4520b549

SAM3 Video: Remove call to pin_memory if the target device is not CUDA (#43061)

atultw committed 139 days ago

Verified 3f73610f

[`Ernie 4.5 VL Moe`] Fix non contiguous params (#43134)

vasqu committed 139 days ago

Verified 828c039e

[`Moe`] Post interface fixes (#43129)

vasqu committed 139 days ago

Verified 5cf0f5be

Fix buffer offloading (#43131)

Cyrilvallez committed 139 days ago

Verified e5b9fbd1

Fix test_all_tensors_are_parameter_or_buffer (#43111)

remi-or committed 139 days ago

Verified 02ddf5d4

[Quantization] torchao serialization (#43127)

MekkCyber committed 139 days ago

Verified c8208185

[`Ernie 4.5 VL Moe`] Post merge adjustments (#43117)

vasqu committed 139 days ago

Verified 52e9d05f

[`GPT OSS`] Fix false flag (#43120)

vasqu committed 139 days ago

Verified 5c68832e

Fix shapes in modular_gpt_oss.py (#42737)

tanmay2004 committed 139 days ago

Verified 68dcd13b

[CB] Support the `num_return_sequences` argument (#42921)

remi-or committed 139 days ago

Verified accb698c

[CI] Fixing some AMD failures (#42879)

Abdennacer-Badaoui committed 139 days ago

Verified c154b021

Fix inits in modernbert (#43115)

remi-or committed 139 days ago

Verified 491e0cd2

🚨Fix MobileViT image processor default interpolation to BICUBIC (#43024)

lukepayyapilli committed 140 days ago

Verified a34922d6

Add fast version of `convert_segmentation_map_to_binary_masks` to EoMT (#43073)

simonreise committed 140 days ago

Verified 98578fc6

🚨 Fix EfficientNet image processor default interpolation to BICUBIC (#42956)

lukepayyapilli committed 140 days ago

Verified 1b743cd9

[SAM3] Fix MPS race condition in add_point_inputs (#43042)

vykhovanets committed 140 days ago

Verified 8fe3ec21

gpt-oss is not working with flash-attention (#42736)

pushkar-hue committed 140 days ago

Verified e8c51d18

Support block_size and bf16_stochastic_round keyword arguments for torchao optimizers (#42972)

klae01 committed 140 days ago

Verified c9b0498f

Silence pytest warnings due to sentencepiece version (#43118)

Cyrilvallez committed 140 days ago

Verified e8d60a7f

Fix warnings poping up with `fixup` and `pytest` (#43109)

Cyrilvallez committed 140 days ago

Verified a616b914

Add expected outputs to DetrForSegmentation docstring example (#43035)

raimbekovm committed 140 days ago

Verified 1802c84f

Added import torch in distilbert pipeline docs (#43053)

PredictiveManish committed 140 days ago

Verified 761d66e4

Fix duplicate words typos in comments and docstrings (#43038)

raimbekovm committed 140 days ago

Verified fc15b322

skip test_sdpa_can_dispatch_on_flash in tests/models/pe_audio/test_mo… (#43014)

sywangyi committed 140 days ago

Verified 37cd3240

Fix InvalidVersion error when flash_attn version cannot be determined (#43004)

lukepayyapilli committed 140 days ago

Verified 82c60184

Newer Older