Pull Requests huggingface/transformers

[DeepSeekV2] Fix integration tests OOM: use device_map=auto instead of 8-bit quantization

#47991 by ydshieh was merged 2026-08-15 19:59

[OLMoE] Update expected logits for A10G and add torch.no_grad()

#47989 by ydshieh was merged 2026-08-15 17:56

[GPTNeoX] Fix post_processor not overridden when loading from pretrained (OLMo garbage generation)

#47988 by ydshieh was merged 2026-08-15 12:38

[OLMo] Fix OOM in logits tests by adding torch.no_grad()

#47986 by ydshieh was merged 2026-08-15 13:02

[AXK1] Fix expected logits for CUDA A10G

#47980 by ydshieh was merged 2026-08-14 21:03

[Gemma] Update expected values for A10G

#47976 by ydshieh was merged 2026-08-14 20:23

Remove stale (None, None) fallback in qwen2_5_vl batch_different_resolutions test

#47972 by ydshieh was merged 2026-08-14 12:42

[serge] Fix 2 integration tests for model `emu3` failing with `import_or_config` (other (2))

#47969 by sergereview[bot] was closed 2026-08-14 12:43

[tests] Fix expected output for Qwen2.5-VL batch_wo_image on CUDA

#47968 by ydshieh was merged 2026-08-14 10:54

Fix unquoted-key JSON repair skipping keys preceded by whitespace

#47967 by NishchayMahor was closed 2026-08-14 12:07

Fix Qwen3.5 BF16 gated delta initialization Code agent slop

#47962 by devXyi was closed 2026-08-14 11:15

[Generate] Return the distribution candidates were drawn from (speculative decoding losslessness) Code agent slop

#47961 by peterlodri-sec was closed 2026-08-14 11:12

Fix Jinja block endings in CHAT WITH MODELS' Writing a chat template …

#47960 by ak1for2business-prog was merged 2026-08-14 11:26

[Init] Guard nn.Embedding branch in generic _init_weights against packed weights Code agent slop

#47959 by peterlodri-sec was closed 2026-08-14 11:13

[Trainer] Exempt PEFT-wrapped models from quant-method training guard Code agent slop

#47958 by peterlodri-sec was closed 2026-08-14 11:14

[bitnet] Support use_sub_norms=False (weight-quant-only BitLinear checkpoints) Code agent slop

#47955 by peterlodri-sec was closed 2026-08-14 11:11

Potential fix for code scanning alert no. 267: Artifact poisoning

#47949 by tarekziade was merged 2026-08-14 12:11

[emu3] 🦮 Black Labrador is back! Fix image generation broken since #37033

#47948 by ydshieh was merged 2026-08-14 12:54

[emu3] 🦮 Black Labrador is back! Fix image generation broken since #37033

#47947 by ydshieh was closed 2026-08-13 15:21

Scan a diff in trufflehog, not the whole repo history

#47945 by tarekziade was merged 2026-08-13 17:15

Fix GatedDeltaNet A_log dtype to prevent -inf under bfloat16 init

#47944 by Nkluge-correa was merged 2026-08-14 11:52

Fix Gemma `sliding_window` being halved on every config save/reload

#47940 by Bluear7878 was merged 2026-08-13 13:41

Declare sdpa support in `TimmWrapper`

#47939 by jiqing-feng was merged 2026-08-14 09:46

[serge] Fix 1 integration test for model `emu3` failing with `output_mismatch` (list output differs (1))

#47938 by sergereview[bot] was closed 2026-08-14 10:02

[serge] Fix 2 integration tests for model `got_ocr2` failing with `other` (other (2))

#47937 by sergereview[bot] was merged 2026-08-14 11:35

[docs] torchcodec + vision processor features

#47935 by stevhliu was merged 2026-08-14 16:25

CI: add vllm-test-init and vllm-test-transformers jobs on dedicated runners

#47934 by ydshieh was merged 2026-08-12 18:57

Add support for Lumma Model

#47929 by ramsha-khann was closed 2026-08-12 20:25

docs: fix incorrect PEFT anchor link in fine-tuning section

#47927 by dsulot was merged 2026-08-12 21:48

Fix CLIP _init_weights when a child module carries quantized weights

#47921 by Bluear7878 was merged 2026-08-14 12:21