huggingface/transformers

Pull Requests Commits

Revert "not needed"

tarekziade committed 164 days ago

3c10a25a

tarekziade committed 164 days ago

d9b3dd54

add timers in check and revert some changes

tarekziade committed 164 days ago

8d9af741

reduce threshold

tarekziade committed 164 days ago

5e74df4c

qa: more agressive lazy loading - fixes #44273

tarekziade committed 164 days ago

31b71156

Gemma3n and Gemma4 cannot use rotary kernel (#45564)

Cyrilvallez committed 165 days ago

Verified bc4b3304

do not index past decoded chars with special tokens (#45435)

itazap committed 165 days ago

Verified 74a2a4d0

Update dev version (#45583)

vasqu committed 165 days ago

Verified c71125f1

Add expert parallelism (EP) config support for Qwen3 MoE (#45436)

AmineDiro committed 165 days ago

Verified 7cf4241f

Update torchao usage for XPU and CPU (#45560)

jiqing-feng committed 165 days ago

Verified b00b7c08

[docs] per-request sampling params (#45553)

stevhliu committed 165 days ago

Verified decdbdd0

[`Privacy Filter`] Add model (#45580)

vasqu committed 165 days ago

Verified 7187177f

Add ForSequenceClassification heads for the OLMo family (#45551)

earino committed 165 days ago

Verified 864db665

Add IndexCache support for GLM5 DSA (#45424)

louzongzhi committed 165 days ago

Verified bca7eee6

Fix redundant logic in video processing SmolVLM (#45272)

yonigozlan committed 165 days ago

Verified 56175617

Add /v1/completions endpoint (OpenAI legacy completions API) to `transformers serve` (#44558)

rain-1 committed 165 days ago

Verified 0ea540ef

Fix typos (#45574)

vasqu committed 165 days ago

Verified 8a3b7b04

Updated the image cache for Paddle models according to the latest API (#45562)

zhang-prog committed 165 days ago

Verified 71bb903d

[Model] Add SLANet Model Support (#45532)

zhang-prog committed 165 days ago

Verified 3d779723

refactor(Dots1): drop Dots1MoE override to `pass` (inherits from DSV3 MoE) (#45572)

casinca committed 165 days ago

Verified 8ca32875

perf: avoid recomputing rotary_emb for each layer in some Google and ModernBERT models (#45555)

casinca committed 165 days ago

Verified 628c54c2

Gemma4 training with text-only samples (#45454)

zucchini-nlp committed 165 days ago

Verified 6979d696

[nemotron_h] Add support for MLP mixers (#44763)

xenova committed 165 days ago

Verified 77c0e6e7

add expert parallelism for gemma-4-26B-A4B-it (#45279)

sywangyi committed 166 days ago

Verified e737fb86

Add full GGUF loading support for GPT‑OSS (fixes #43366, supersedes #43757) latest (#45506)

sirzechs66 committed 166 days ago

Verified 5e238684

Update Gemma4 weight conversion script (#45328)

RyanMullins committed 166 days ago

Verified 2da596b0

Move some conversion mappings to PrefixChange (#45567)

Cyrilvallez committed 166 days ago

Verified 71d1a7b7

fix table update versions (#45544)

tarekziade committed 166 days ago

Verified f51cf506

Add disable_mmap kwarg to from_pretrained with hf-mount auto-detection (#45547)

rtrompier committed 166 days ago

Verified 8fb7c7e5

fix(DSV3): parity between native `DeepseekV3MoE` and remote official implementation (#45441)

casinca committed 166 days ago

Verified cb0adddc

Older