Commits

Commits on Sep 5, 2025

refactor(serve): move `request_id` to headers (#40722)
Show description for bb45d36

McPatate
and
gante
authored
bb45d36
Copy full SHA for bb45d36
Skip `VitMatteImageProcessingTest::test_fast_is_faster_than_slow` (#40713)
Show description for 12b8e10
ydshieh
authored
12b8e10
Copy full SHA for 12b8e10
Keypoint matching docs (#40541)
Show description for 6b23261

authored
6b23261
Copy full SHA for 6b23261
[`Gemma Embedding`] Fix SWA (#40700)
Show description for 948bc0f
vasqu
authored
948bc0f
Copy full SHA for 948bc0f
Add Optional typing (#40686)
Show description for 828044c
cyyever
authored
828044c
Copy full SHA for 828044c
[tests] remove overwrites of removed test (#40720)
Show description for e9d6a69
gante
authored
e9d6a69
Copy full SHA for e9d6a69
[serve] re-enable tests (#40717)
Show description for 96a5774
gante
authored
96a5774
Copy full SHA for 96a5774
Fix arguments (#40605)
Show description for c76387e
cyyever
authored
c76387e
Copy full SHA for c76387e
🔴 Update Glm4V to use config values (#40712)
Show description for 21f0903
zucchini-nlp
authored
21f0903
Copy full SHA for 21f0903
Fix parent classes of AllKwargsForChatTemplate (#40685)
Show description for b62e5b6
cyyever
authored
b62e5b6
Copy full SHA for b62e5b6
[onnx] use logical `or` for grounding dino mask (#40625)
Show description for 313effa

lmarshall12
and
Lewis Marshall
authored
313effa
Copy full SHA for 313effa
[moduar] Add missing `self` in post-process methods (#40711)
framonmar7
authored
f3211b5
Copy full SHA for f3211b5
[tests] fix blip2 edge case (#40699)
gante
authored
a2a8a3c
Copy full SHA for a2a8a3c
🚨 Allow `check_model_inputs` in core VLMs (#40342)
Show description for 4e195f1
zucchini-nlp
authored
4e195f1
Copy full SHA for 4e195f1
Fix parent classes of ProcessingKwargs (#40676)
Show description for 93df343
cyyever
authored
93df343
Copy full SHA for 93df343
feat(serve): add healthcheck test (#40697)
McPatate
authored
89e103c
Copy full SHA for 89e103c
Fetch more test data with `hf_hub_download` (#40710)
Show description for a2fffa5
ydshieh
authored
a2fffa5
Copy full SHA for a2fffa5

Commits on Sep 4, 2025

Add Fast Image Processor for ImageGPT (#39592)
Show description for 4a88e81

authored
4a88e81
Copy full SHA for 4a88e81
Fetch one missing test data (#40703)
Show description for 9db11b7
ydshieh
authored
9db11b7
Copy full SHA for 9db11b7
Align assisted generate for unified signature in decoding methods (#40657)
Show description for acd8205
manueldeprada
authored
acd8205
Copy full SHA for acd8205
Avoid `T5GemmaModelTest::test_eager_matches_sdpa_inference` being flaky (#40702)
Show description for 16b821c
ydshieh
authored
16b821c
Copy full SHA for 16b821c
Fix broken Llama4 accuracy in MoE part (#40609)
Show description for 519c252

nvpohanh
and
Cyrilvallez
authored
519c252
Copy full SHA for 519c252
[Glm4.5V] fix vLLM support (#40696)
Show description for 586dc5d
zucchini-nlp
authored
586dc5d
Copy full SHA for 586dc5d
Fix self.dropout_p is not defined for SamAttention/Sam2Attention (#40667)
Show description for ad2da3e
yonigozlan
authored
ad2da3e
Copy full SHA for ad2da3e
Fix backward compatibility with accelerate in Trainer (#40668)
qgallouedec
authored
e39f222
Copy full SHA for e39f222
Change docker image to preview for the MI355 CI (#40693)
Show description for d8f6705
ahadnagy
authored
d8f6705
Copy full SHA for d8f6705
Fixing bug in Voxtral when merging text and audio embeddings (#40671)
Show description for 4cbca0d

rcogill
and
eustlb
authored
4cbca0d
Copy full SHA for 4cbca0d
feat: support request cancellation (#40599)
Show description for 9a6c656
McPatate
authored
9a6c656
Copy full SHA for 9a6c656
add: embedding model (#40694)
Show description for 87f38db

RyanMullins
and
SindhuRaghuram97
authored
87f38db
Copy full SHA for 87f38db
Final test data cache - inside CI docker images (#40689)
Show description for 5b0c01b
ydshieh
authored
5b0c01b
Copy full SHA for 5b0c01b
Load a tiny video to make CI faster (#40684)
Show description for 1f3cc93
zucchini-nlp
authored
1f3cc93
Copy full SHA for 1f3cc93
fix broken offline mode when loading tokenizer from hub (#40669)
Show description for 669230a
winglian
authored
669230a
Copy full SHA for 669230a
Add codebook_dim attribute to DacVectorQuantize for DacResidualVectorQuantize.from_latents() (#40665)
Show description for 91b34be
flavioialongo
authored
91b34be
Copy full SHA for 91b34be
Add sequence classification support for small Gemma 3 text models (#40562)
Show description for 25b4a0d

abdokaseb
and
zucchini-nlp
authored
25b4a0d
Copy full SHA for 25b4a0d
CircleCI docker images cleanup / update / fix (#40681)
Show description for 30a4b87
ydshieh
authored
30a4b87
Copy full SHA for 30a4b87