huggingface/transformers

Pull Requests Commits

Merge branch 'main' into fix_weight_tying

Cyrilvallez committed 16 days ago

Verified 587fef4b

Cyrilvallez committed 16 days ago

Verified b5537a39

hardcode llava onevision

Cyrilvallez committed 16 days ago

Verified f85d0f4f

revert unwanted changes

Cyrilvallez committed 16 days ago

Verified 65dca34d

Attempt to fix VLM gradient enabling (#41993)

molbap committed 16 days ago

Verified e2f08ea5

[quantization] Dequant fp8 when cuda or xpu not available (#42511)

MekkCyber committed 16 days ago

Verified afd039d1

Migration guide: tokenization (#42477)

LysandreJik committed 16 days ago

Verified fa3cf839

Delete irrelevant test that sometimes fails (#42515)

Rocketknight1 committed 16 days ago

Verified 305f82a1

[loading] `tie_weights` is finally symmetric (#42507)

Cyrilvallez committed 16 days ago

Verified 9f587100

perf: Remove implicit CPU-GPU syncs due to implicit .item() call (#42433)

Dhruv88 committed 16 days ago

Verified a3881a8d

Fix broken link to dataset food101 (#42495)

JuanFKurucz committed 16 days ago

Verified 297f5ae8

Small migration guide update (#42510)

ArthurZucker committed 16 days ago

Verified f78723fa

Fix a bug where compile_config gets deleted (#42474)

remi-or committed 16 days ago

Verified a0216a80

Patch Mistral Common Tokenizer 2 (#41962)

juliendenize committed 16 days ago

Verified aaa73254

Handle decorator with optional arguments better (#42512)

hmellor committed 16 days ago

Verified 2dba9725

logic to select tf32 API as per Pytorch version (#42428)

khushali9 committed 16 days ago

Verified 7f5c2094

[`CI`] Fix copies (#42487)

vasqu committed 16 days ago

Verified 01823d7e

remove tokenizer warning (#42483)

SunMarc committed 16 days ago

Verified e51e75e1

Fix `pr_build_doc_with_comment.yml` not reporting correctly (#42505)

ydshieh committed 16 days ago

Verified 0c05c3bd

Convenient default behavior for pipeline TTS usage. (#42473)

ebezzam committed 16 days ago

Verified 26dbe643

Add afmoe model (#42168)

alyosha-swamy committed 18 days ago

Verified cac0a28c

Fix broken models due to modeling/weight mapping (#42468)

Cyrilvallez committed 19 days ago

Verified 2a61590a

[`Docs`] Fix build on other langs (#42485)

vasqu committed 19 days ago

Verified a4a7008e

add rotary kernel support to Qwen3 model (#41147)

kaixuanliu committed 19 days ago

Verified 6587d77f

Fix minor bugs in ConversionOps and WeightConverter (#42479)

SunMarc committed 19 days ago

Verified 4abfd165

Fix gpt2 tokenizer `add_prefix_space` default value (#42481)

SunMarc committed 19 days ago

Verified c79ac071

[quantization] make torchao tests slow (#42482)

MekkCyber committed 19 days ago

Verified f8e29b5a

Skip failing irrelevant test for ColQwen2 (#42480)

Rocketknight1 committed 19 days ago

Verified 7b5f1db8

add require_deterministic_for_xpu to make the case pass in xpu (#42439)

sywangyi committed 19 days ago

Verified 98fc2a22

extend CwmIntegrationTest to xpu (#42314)

sywangyi committed 19 days ago

Verified 5dcda1b9

Older