Pull Requests huggingface/transformers

Add auto_docstring task model overrides

#47815 opened 2026-08-06 12:11 by guarin

fix: the _file_md5 function uses the md5 hashing alg... in checkers.py

#47813 opened 2026-08-06 09:56 by anupamme

Add KV replicate feature when kv_head<tp_size

#47811 opened 2026-08-06 07:35 by kaixuanliu

[CB] [Major] Upgrade the cache to support different attention types

#47809 opened 2026-08-06 06:24 by remi-or

[Fix] Export crashes on kernel-decorated function

#47808 opened 2026-08-06 05:56 by remi-or

[serge] Fix 1 integration test for model `qwen3_vl_moe` failing with `output_mismatch` (list output differs (1))

#47806 opened 2026-08-06 00:18 by sergereview[bot]

Update EXAONE MoE for K-EXAONE 2.0

#47802 opened 2026-08-05 19:17 by nuxlear

Add llava onevision 1.5

#47795 opened 2026-08-05 10:20 by chengzheng345

fix: decode() batch path respects self.clean_up_tokenization_spaces

#47793 opened 2026-08-05 09:53 by lorenzozanee

[serge] Fix 2 integration tests for model `qwen2_moe` failing with `other` (other (2))

#47784 opened 2026-08-04 22:48 by sergereview[bot]

Add tiny model

#47782 opened 2026-08-04 21:12 by sushmanthreddy

[docs] Update toctree

#47781 opened 2026-08-04 20:17 by stevhliu

Add RWKV-7 (Goose) New model

#47780 opened 2026-08-04 20:04 by Hakureirm

Refactor GGUF to speed-up inference

#47779 opened 2026-08-04 18:57 by SunMarc

Fix broken links in troubleshooting guide

#47778 opened 2026-08-04 18:20 by seedaylight

docs: use relative paths for README language menus and add fa/ro entries

#47777 opened 2026-08-04 17:52 by Priyans-Lathiya

Add minimax vl 01

#47776 opened 2026-08-04 17:24 by sushmanthreddy

[don't merge] check mlinter

#47774 opened 2026-08-04 16:17 by ydshieh

:rotating_light: Kernels=True ?!

#47773 opened 2026-08-04 16:07 by vasqu

[docs] Fix code snippets

#47772 opened 2026-08-04 15:42 by stevhliu

fix device-to-host sync in flash attention

#47771 opened 2026-08-04 15:14 by DhanushPillay

Fix auto_docstrings for recent models

#47768 opened 2026-08-04 14:04 by guarin

Fix HuBERT positional conv embedding padding sensitivity when conv_pos_batch_norm=True (#47739)

#47763 opened 2026-08-04 08:55 by loulanyue

Post the model linter's first pass as a PR review

#47760 opened 2026-08-04 08:29 by ArthurZucker

chore: update ssh-runner.yml

#47758 opened 2026-08-04 08:06 by hf-security-analysis[bot]

support for VoxCPM2 New model

#47756 opened 2026-08-04 05:23 by sushmanthreddy

[serge] Fix 2 integration tests for model `clvp` failing with `output_mismatch` (tensor values differ (2))

#47751 opened 2026-08-03 23:02 by sergereview[bot]

[serge] Fix 4 integration tests for model `pvt` failing with `output_mismatch` (tensor values differ (4))

#47750 opened 2026-08-03 22:47 by sergereview[bot]

[docs] Fix links and remove TokenizerFast

#47748 opened 2026-08-03 21:03 by stevhliu

Support quantization in Dynamo, ONNX and ET exporters

#47747 opened 2026-08-03 19:56 by IlyasMoutawwakil