Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3902 commits
Select commit Hold shift + click to select a range
11a6b95
Oupsy (#40544)
ArthurZucker Aug 29, 2025
d888bd4
Fix typos (#40511)
cyyever Aug 29, 2025
f68eb5f
Redundant code removal (#40534)
piyushK52 Aug 29, 2025
f0e7781
Clean-up kernel loading and dispatch (#40542)
Cyrilvallez Aug 29, 2025
ffdd10f
Allow compression on meta device (#39039)
shanjiaz Aug 29, 2025
74a2421
Add bfloat16 support detection for MPS in is_torch_bf16_gpu_available…
andrerom Aug 29, 2025
828a27f
Fix gpt-oss rope warning (#40550)
SunMarc Aug 29, 2025
7efb4c8
Flaky CI is annoying (#40543)
zucchini-nlp Aug 29, 2025
1067577
fix gpt-oss out shape (#40535)
jiqing-feng Aug 29, 2025
155fd92
Fix `GptOssModelTest::test_assisted_decoding_matches_greedy_search_1_…
ydshieh Aug 29, 2025
ce48e9c
Dev version
LysandreJik Aug 29, 2025
2d3b886
Fix collated reports upload filename (#40556)
ivarflakstad Aug 30, 2025
36fddeb
pin `pytest-rerunfailures<16.0` (#40561)
ydshieh Aug 30, 2025
1363fce
remove the redundant non maintained jieba and use rjieba instead (#40…
divyanshsinghvi Aug 30, 2025
b0db5a0
Set `test_all_params_have_gradient=False` for `DeepseekV2ModelTest` (…
ydshieh Aug 30, 2025
0b24507
processor tests - use dummy videos (#40537)
zucchini-nlp Sep 1, 2025
1d74264
[qwen-vl] fix position ids (#40490)
zucchini-nlp Sep 1, 2025
cfed99d
Fix `test_eager_matches_sdpa_inference` not run for `CLIP` (#40581)
ydshieh Sep 1, 2025
6546f28
Fix CircleCI step passes in the case of pytest worker crash at test c…
ydshieh Sep 1, 2025
db6821b
Allow `remi-or` to `run-slow` (#40590)
ydshieh Sep 1, 2025
48ebae9
Fix llava image processor (#40588)
zucchini-nlp Sep 1, 2025
2537ed4
Update `get_*_features` methods + update doc snippets (#40555)
qubvel Sep 1, 2025
90953d5
Fix custom generate relative imports (#40480)
manueldeprada Sep 1, 2025
3bccb02
🚨 Remove Group Beam Search decoding strategy (#40495)
manueldeprada Sep 1, 2025
564be6d
Support batch size > 1 image-text inference (#36682)
hiyouga Sep 1, 2025
8564e21
🚨 Remove Constrained Beam Search decoding strategy (#40518)
manueldeprada Sep 1, 2025
a543095
Fix typos (#40585)
cyyever Sep 1, 2025
8e0b2c8
Skip `TvpImageProcessingTest::test_slow_fast_equivalence` (#40593)
ydshieh Sep 1, 2025
2b71c5b
Fix inexistent imports (#40580)
cyyever Sep 1, 2025
3c3dac3
Add Copilot instructions (#40432)
Rocketknight1 Sep 1, 2025
c99d43e
Fix `siglip` flaky `test_eager_matches_sdpa_inference` (#40584)
ydshieh Sep 1, 2025
21e708c
Fix for missing default values in encoder decoder (#40517)
remi-or Sep 1, 2025
3beac9c
Fix quite a lot of FA tests (#40548)
Cyrilvallez Sep 1, 2025
abf5900
[`Tests`] Fixup duplicated mrope logic (#40592)
vasqu Sep 1, 2025
4da03d7
Reduce more test data fetch (#40595)
ydshieh Sep 1, 2025
b365550
Pin torchcodec to 0.5 in AMD docker (#40598)
remi-or Sep 1, 2025
514b3e8
Multiple fixes to FA tests in AMD (#40498)
remi-or Sep 1, 2025
8a091cc
Disable cache for `TokenizerTesterMixin` temporarily (#40611)
ydshieh Sep 2, 2025
b2b1c30
fix: continuous batching in `transformers serve` (#40479)
McPatate Sep 2, 2025
bbd8085
Fix processor chat template (#40613)
zucchini-nlp Sep 2, 2025
91be12b
Avoid `too many request` caused by `AutoModelTest::test_dynamic_savin…
ydshieh Sep 2, 2025
97266df
Fix flaky `JambaModelTest.test_load_balancing_loss` (#40617)
ydshieh Sep 2, 2025
8c60a7c
Add collated reports job to Nvidia CI (#40470)
ahadnagy Sep 2, 2025
4f54205
Remove unnecessary pillow version check (#40604)
cyyever Sep 2, 2025
37103d6
Fix invalid typing (#40612)
cyyever Sep 2, 2025
a470f21
Enable more ruff UP rules (#40579)
cyyever Sep 2, 2025
d47ad91
Support TF32 flag for MUSA backend (#33187)
fmo-mt Sep 2, 2025
ed49376
Remove random flag (#40629)
Cyrilvallez Sep 2, 2025
00a8364
🌐 [i18n-KO] Translated `deepseek_v3.md` to Korean (#39649)
ssum21 Sep 2, 2025
e690fe6
Fix `too many requests` in `TestMistralCommonTokenizer` (#40623)
ydshieh Sep 3, 2025
0f5e45a
fix: gas for gemma fixed (#40591)
yevvonlim Sep 3, 2025
cb54ce4
[auto-model] propagate kwargs (#40491)
zucchini-nlp Sep 3, 2025
acc968c
[CP] Add attention_mask to the buffer when the mask is causal (#40619)
kashif Sep 3, 2025
2bbf98a
Fix: PIL image load in Processing utils apply_chat_template (#40622)
abdokaseb Sep 3, 2025
c485c52
Skip `test_prompt_lookup_decoding_matches_greedy_search` for `voxtral…
ydshieh Sep 3, 2025
03708cc
add DeepseekV3ForTokenClassification (#40641)
bzantium Sep 3, 2025
5e2e496
fix MetaCLIP 2 wrong link & wrong model names in the docstrings (#40565)
voidism Sep 3, 2025
fe1a9e0
Remove TF/Flax examples (#40654)
Rocketknight1 Sep 3, 2025
5ea5c81
Mark `LongformerModelTest::test_attention_outputs` as flaky (#40655)
ydshieh Sep 3, 2025
f2416b4
fix pipeline dtype (#40638)
jiqing-feng Sep 3, 2025
238a827
feat(serving): add healthcheck (#40653)
McPatate Sep 3, 2025
57f5668
Fix Metaclip modular conversion (#40660)
Rocketknight1 Sep 3, 2025
a8f4003
Avoid attention_mask copy in qwen2.5 (#40658)
cyyever Sep 3, 2025
3f40ebf
Allow custom args in `custom_generate` Callables and unify generation…
manueldeprada Sep 3, 2025
30d66dc
Update `check_determinism` inside `test_determinism` (#40661)
ydshieh Sep 3, 2025
b67608f
Skip `test_fast_is_faster_than_slow` for `Owlv2ImageProcessingTest` (…
ydshieh Sep 3, 2025
55676d7
Fix warning for output_attentions=True (#40597)
qubvel Sep 3, 2025
68b9cbb
Skip `test_prompt_lookup_decoding_matches_greedy_search` for `qwen2_a…
ydshieh Sep 3, 2025
afd1393
Remove overwritten `GitModelTest::test_beam_search_generate` (#40666)
ydshieh Sep 3, 2025
459c1fa
refactor: use `tolist` instead of list comprehension calling `.item()…
McPatate Sep 3, 2025
f22ec7f
Benchmarking V2: framework impl (#40486)
ahadnagy Sep 3, 2025
34595cf
Even more test data cached (#40636)
ydshieh Sep 3, 2025
d40e7ea
Skip more fast v.s slow image processor tests (#40675)
ydshieh Sep 4, 2025
ca9b36a
Avoid night torch CI not run because of irrelevant docker image faili…
ydshieh Sep 4, 2025
7f92e1f
Mark `Aimv2ModelTest::test_eager_matches_sdpa_inference_04_fp16_pad_r…
ydshieh Sep 4, 2025
30a4b87
CircleCI docker images cleanup / update / fix (#40681)
ydshieh Sep 4, 2025
25b4a0d
Add sequence classification support for small Gemma 3 text models (#4…
abdokaseb Sep 4, 2025
91b34be
Add codebook_dim attribute to DacVectorQuantize for DacResidualVector…
flavioialongo Sep 4, 2025
669230a
fix broken offline mode when loading tokenizer from hub (#40669)
winglian Sep 4, 2025
1f3cc93
Load a tiny video to make CI faster (#40684)
zucchini-nlp Sep 4, 2025
5b0c01b
Final test data cache - inside CI docker images (#40689)
ydshieh Sep 4, 2025
87f38db
add: embedding model (#40694)
RyanMullins Sep 4, 2025
9a6c656
feat: support request cancellation (#40599)
McPatate Sep 4, 2025
4cbca0d
Fixing bug in Voxtral when merging text and audio embeddings (#40671)
rcogill Sep 4, 2025
d8f6705
Change docker image to preview for the MI355 CI (#40693)
ahadnagy Sep 4, 2025
e39f222
Fix backward compatibility with accelerate in Trainer (#40668)
qgallouedec Sep 4, 2025
ad2da3e
Fix self.dropout_p is not defined for SamAttention/Sam2Attention (#40…
yonigozlan Sep 4, 2025
586dc5d
[Glm4.5V] fix vLLM support (#40696)
zucchini-nlp Sep 4, 2025
519c252
Fix broken Llama4 accuracy in MoE part (#40609)
nvpohanh Sep 4, 2025
16b821c
Avoid `T5GemmaModelTest::test_eager_matches_sdpa_inference` being fla…
ydshieh Sep 4, 2025
acd8205
Align assisted generate for unified signature in decoding methods (#4…
manueldeprada Sep 4, 2025
9db11b7
Fetch one missing test data (#40703)
ydshieh Sep 4, 2025
4a88e81
Add Fast Image Processor for ImageGPT (#39592)
agamjots05 Sep 4, 2025
a2fffa5
Fetch more test data with `hf_hub_download` (#40710)
ydshieh Sep 5, 2025
89e103c
feat(serve): add healthcheck test (#40697)
McPatate Sep 5, 2025
93df343
Fix parent classes of ProcessingKwargs (#40676)
cyyever Sep 5, 2025
4e195f1
🚨 Allow `check_model_inputs` in core VLMs (#40342)
zucchini-nlp Sep 5, 2025
a2a8a3c
[tests] fix blip2 edge case (#40699)
gante Sep 5, 2025
f3211b5
[moduar] Add missing `self` in post-process methods (#40711)
framonmar7 Sep 5, 2025
313effa
[onnx] use logical `or` for grounding dino mask (#40625)
lmarshall12 Sep 5, 2025
b62e5b6
Fix parent classes of AllKwargsForChatTemplate (#40685)
cyyever Sep 5, 2025
21f0903
🔴 Update Glm4V to use config values (#40712)
zucchini-nlp Sep 5, 2025
c76387e
Fix arguments (#40605)
cyyever Sep 5, 2025
96a5774
[serve] re-enable tests (#40717)
gante Sep 5, 2025
e9d6a69
[tests] remove overwrites of removed test (#40720)
gante Sep 5, 2025
828044c
Add Optional typing (#40686)
cyyever Sep 5, 2025
948bc0f
[`Gemma Embedding`] Fix SWA (#40700)
vasqu Sep 5, 2025
6b23261
Keypoint matching docs (#40541)
merveenoyan Sep 5, 2025
12b8e10
Skip `VitMatteImageProcessingTest::test_fast_is_faster_than_slow` (#4…
ydshieh Sep 5, 2025
bb45d36
refactor(serve): move `request_id` to headers (#40722)
McPatate Sep 5, 2025
3f7bda4
[Continous Batching] fix do_Sample=True in continuous batching (#40692)
kashif Sep 8, 2025
44b3888
Fix order of mask functions when using `and/or_mask_function` (#40753)
Cyrilvallez Sep 8, 2025
def7558
Fix np array typing (#40741)
cyyever Sep 8, 2025
7b8d40e
Set accepts_loss_kwargs to False for ConvNext(|V2)ForImageClassificat…
clinty Sep 8, 2025
2a1eb5b
Add BF16 support check for MUSA backend (#40576)
fmo-mt Sep 8, 2025
9ab6078
remove gemmas eager training warning (#40744)
August-murr Sep 8, 2025
dc262ee
remove FSDP prefix when using save_pretrained with FSDP2 (#40207)
winglian Sep 8, 2025
2077f17
feat: err when unsupported attn impl is set w/ `--continuous_batching…
McPatate Sep 8, 2025
a9b313a
docs: add continuous batching to serving (#40758)
McPatate Sep 8, 2025
bb8e9cd
Remove unnecessary tildes from documentation (#40748)
st81 Sep 8, 2025
fd2a29d
Fix more typos (#40627)
cyyever Sep 8, 2025
e8db153
Fix inconsistency in SeamlessM4T and SeamlessM4Tv2 docs (#39364)
clinty Sep 8, 2025
5a468e5
Fix `continue_final_message` in `apply_chat_template` to prevent subs…
abdokaseb Sep 8, 2025
126264d
🌐 [i18n-KO] Translated 'xclip.md' to Korean (#39594)
ssum21 Sep 8, 2025
aee5000
Fix Bark failing tests (#39478)
ebezzam Sep 8, 2025
a871f6f
Add EfficientLoFTRImageProcessorFast for GPU-accelerated image proces…
LawJarp-A Sep 8, 2025
82d66e5
Fix: swanlab `public.cloud.experiment_url` api error (#40763)
Zeyi-Lin Sep 9, 2025
ed10021
[generate] `PromptLookupCandidateGenerator` won't generate forbidden …
gante Sep 9, 2025
1cdbbb3
Support sliding window in CB (#40688)
remi-or Sep 9, 2025
de5cbe8
[deprecations] Remove generate-related deprecations up to v4.56 (#40729)
gante Sep 9, 2025
7aaef98
rm src/transformers/convert_pytorch_checkpoint_to_tf2.py (#40718)
gante Sep 9, 2025
71ac7ea
[tests] update `test_past_key_values_format` and delete overwrites (#…
gante Sep 9, 2025
d33c189
[RoPE] run RoPE tests when the model uses RoPE (#40630)
gante Sep 9, 2025
d09fdf5
Fix crash when executing MambaCache sample code (#40557)
torotoki Sep 9, 2025
37c1443
[pipeline] ASR pipeline kwargs are forwared to `generate` (#40375)
gante Sep 9, 2025
79fdbf2
[docs] CPU install (#40631)
stevhliu Sep 9, 2025
b928235
Adding Support for Qwen3-Next (#40771)
bozheng-hit Sep 9, 2025
3340ccb
Fix gpt-oss router_indices in EP (#40545)
jiqing-feng Sep 10, 2025
c52889b
Remove reference of video_load_backend and video_fps for processor (#…
cyyever Sep 10, 2025
08edec9
[processors] Unbloating simple processors (#40377)
zucchini-nlp Sep 10, 2025
a5ecd94
Enable ruff on benchmark and scripts (#40634)
cyyever Sep 10, 2025
a72e5a4
🚨 Fix Inconsistant `input_feature` length and `attention_mask` length…
BakerBunker Sep 10, 2025
0997c2f
Fix doc for PerceptionLMForConditionalGeneration forward. (#40733)
shuminghu Sep 10, 2025
6c9f412
Fix typos in tests and util (#40780)
cyyever Sep 10, 2025
8ab2448
Fix invalid PipelineParallel member (#40789)
cyyever Sep 10, 2025
7401cfa
Use functools.cached_property (#40607)
cyyever Sep 10, 2025
75202b0
Read config pattern for Qwen3Next (#40792)
Cyrilvallez Sep 10, 2025
abbed70
Fix dotted model names (#40745)
August-murr Sep 10, 2025
e5ecb03
Fix the issue that csm model cannot work with pipeline mode. (#39349)
yuanwu2017 Sep 10, 2025
3378e7d
Move num_items_in_batch to correct device before accelerator.gather (…
ssharpe42 Sep 10, 2025
7d57b31
Remove use_ipex option from Trainer (#40784)
cyyever Sep 10, 2025
dae1ccf
fix_image_processing_fast_for_glm4v (#40483)
lambertwjh Sep 10, 2025
549ba5b
[Docs] Add missing class documentation for optimizer_schedules (#3187…
jijihuny Sep 10, 2025
df67cd3
Fix DeepSpeed mixed precision precedence over Accelerate defaults (#3…
notkisk Sep 11, 2025
ec532f2
feature: Add robust token counting with padding exclusion (#40416)
PrathmeshAdsod Sep 11, 2025
de01a22
Fix edge case for tokenize (#36277) (#36555)
Sep 11, 2025
02f1d7c
Fix config dtype parsing for Emu3 edge case (#40766)
Isotr0py Sep 11, 2025
0f1b128
⚠️ 🔴 Add ministral model (#40247)
manueldeprada Sep 11, 2025
6d36912
Align torch implementation of Gated DeltaNet in Qwen3-Next with fla l…
bozheng-hit Sep 11, 2025
895b3eb
Fix typos in src (#40782)
cyyever Sep 11, 2025
c264c0e
add general hub test for Fast Image Processors in test_image_processi…
namgyu-youn Sep 11, 2025
dfae7dd
Push generation config along with checkpoints (#40804)
qgallouedec Sep 11, 2025
cf084f5
[`Jetmoe`] Fix RoPE (#40819)
vasqu Sep 11, 2025
c81f426
🌐 [i18n-KO] Translated clipseg.md to Korean (#39903)
HyunZ118 Sep 12, 2025
5e2e77f
Improve torch_dtype checks (#40808)
cyyever Sep 12, 2025
827b65c
Add VideoProcessors to auto-backend requirements (#40843)
Cyrilvallez Sep 12, 2025
6e69b60
Adds Causal Conv 1D kernel for mamba models (#40765)
MekkCyber Sep 12, 2025
797859c
Update no split modules in T5Gemma model (#40810)
npuichigo Sep 12, 2025
77aa35e
Replace image classification loss functions to `self.loss_function` (…
qubvel Sep 12, 2025
98a8078
Fix the misalignment between the l2norm in GDN of Qwen3-Next and the …
bozheng-hit Sep 12, 2025
ef05393
Fixes for continuous batching (#40828)
remi-or Sep 12, 2025
4cb41ad
[tests] re-enable aria fast tests (#40846)
gante Sep 12, 2025
f384bb8
[SAM2] Fix inconsistent results with original implementation with inp…
yonigozlan Sep 12, 2025
8502b41
[Sam2Video] Fix video inference with batched boxes and add test (#40797)
yonigozlan Sep 12, 2025
291772b
add: differential privacy research model (#40851)
RyanMullins Sep 12, 2025
8d84591
[test] Fix test_eager_matches_sdpa incorrectly skipped (#40852)
eustlb Sep 12, 2025
e682f90
[tests] move generative tests away from `test_modeling_common.py` (#4…
gante Sep 12, 2025
6eb3255
[generate] Always use decoder config to init cache (#40772)
gante Sep 12, 2025
d42e96a
Use checkpoint in auto_class_docstring (#40844)
cyyever Sep 13, 2025
02ea2b3
Fix TrainingArguments.parallelism_config NameError with accelerate<1.…
albertvillanova Sep 14, 2025
9c804f7
Redirect MI355 CI results to dummy dataset (#40862)
ahadnagy Sep 14, 2025
c65fea0
[Bug fix #40813] Fix base_model_tp_plan of Starcoder2 model. (#40814)
greg-kwasniewski1 Sep 15, 2025
93f810e
[docstrings / type hints] Update outdated annotations for `past_key_v…
gante Sep 15, 2025
ada64ce
fix florence kwargs (#40826)
SunMarc Sep 15, 2025
f5e1641
fix: XIELU act parameters not being casted to correct dtype (#40812)
NanoCode012 Sep 15, 2025
96d3795
Update model tags and integration references in bug report (#40881)
ArthurZucker Sep 15, 2025
fc5f910
[Qwen3 Next] Use numerically stable `rsqrt` (#40848)
Sep 15, 2025
c0dbe09
Adding Support for Qwen3-VL Series (#40795)
JJJYmmm Sep 15, 2025
3442b2f
[`VaultGemma`] Update expectations in integration tests (#40855)
vasqu Sep 15, 2025
5e9ec59
Fix modular consistency (#40883)
Cyrilvallez Sep 15, 2025
0957999
🔴 Move variable output controls to `_prepare_generation_config ` (#40…
manueldeprada Sep 15, 2025
e674e9d
Clarify passing is_causal in sdpa_attention_paged_forward (#40838)
cyyever Sep 15, 2025
6254bb4
Use torch.expm1 and torch.log1p for better numerical results (#40860)
cyyever Sep 15, 2025
ff26fe8
Add Fast PromptDepthAnything Processor (#40602)
SamuelBarryCS Sep 15, 2025
04d1c8f
Fix deta loading & dataclass (#40878)
Cyrilvallez Sep 15, 2025
2a83792
Remove dict branch of attention_mask in sdpa_attention_paged_forward …
cyyever Sep 15, 2025
2141a5b
🌐 [i18n-KO] Translated smolvlm.md to Korean (#40414)
HyunZ118 Sep 15, 2025
20ee3a7
🌐 [i18n-KO] Translated `imageprocessor.md` to Korean (#39557)
HyunZ118 Sep 15, 2025
5af248b
[generate] remove docs of a feature that no longer exists (#40895)
gante Sep 15, 2025
21c8379
Make debugging failing tests (check and update expect output values) …
ydshieh Sep 16, 2025
8e1a12b
Fixing the call to kernelize (#40628)
MekkCyber Sep 16, 2025
65adc3a
Fix getter regression (#40824)
molbap Sep 16, 2025
1aff033
Fix flaky `Gemma3nAudioFeatureExtractionTest::test_dither` (#40902)
ydshieh Sep 16, 2025
087775d
[cache] Merge static sliding and static chunked layer (#40893)
Cyrilvallez Sep 16, 2025
3688a97
Harmonize CacheLayer names (#40892)
Cyrilvallez Sep 16, 2025
0c1839d
[cache] Only use scalars in `get_mask_sizes` (#40907)
Cyrilvallez Sep 16, 2025
65f9ede
Set seed for `Glm4vIntegrationTest` (#40905)
ydshieh Sep 16, 2025
d0af426
Add Olmo3 model (#40778)
2015aroras Sep 16, 2025
96bc19b
remove dummy EncodingFast (#40864)
cyyever Sep 16, 2025
df03fc1
Improve module name handling for local custom code (#40809)
XuehaiPan Sep 16, 2025
270da89
Remove `runner_map` (#40880)
ydshieh Sep 16, 2025
88ba0f1
disable `test_fast_is_faster_than_slow` (#40909)
ydshieh Sep 16, 2025
3b3f6cd
[gemma3] `Gemma3ForConditionalGeneration` compatible with assisted ge…
gante Sep 16, 2025
d4af0d9
[generate] misc fixes (#40906)
gante Sep 16, 2025
beb09cb
🔴Make `center_crop` fast equivalent to slow (#40856)
yonigozlan Sep 16, 2025
cccef4b
Fix dtype in Paligemma (#40912)
zucchini-nlp Sep 16, 2025
6e50a8a
[Docs] Adding documentation of MXFP4 Quantization (#40885)
ariG23498 Sep 16, 2025
b82cd1c
Processor load with multi-processing (#40786)
zucchini-nlp Sep 17, 2025
ddd4caf
[Llama4] Remove `image_sizes` arg and deprecate `vision_feature_layer…
yaswanth19 Sep 17, 2025
8428c7b
Fix #40067: Add dedicated UMT5 support to GGUF loader (config, tokeni…
akshay-babbar Sep 17, 2025
f6999b0
[torchao safetensors] renaming get_state_dict function (#40774)
liangel-02 Sep 17, 2025
c830fc1
Adding activation kernels (#40890)
MekkCyber Sep 17, 2025
89949c5
Minor fix for #40727 (#40929)
ydshieh Sep 17, 2025
48a5565
Add support for Florence-2 training (#40914)
ducviet00 Sep 17, 2025
6cade29
Add LongCat-Flash (#40730)
molbap Sep 17, 2025
3853bfe
[DOC] Add missing dates in model cards (#40922)
yonigozlan Sep 17, 2025
66f97d3
[models] remove unused `import torch.utils.checkpoint` (#40934)
gante Sep 17, 2025
26b65fb
Intel CPU dockerfile (#40806)
jiqing-feng Sep 17, 2025
14f01ae
docs(i18n): Correct the descriptive text in the README_zh-hans.md (#4…
lilin-1 Sep 17, 2025
c5553b4
Fix trainer tests (#40823)
SunMarc Sep 17, 2025
ecc1d77
Fix `Glm4vMoeIntegrationTest` (#40930)
ydshieh Sep 17, 2025
eb04363
Raise error instead of warning when using meta device in from_pretrai…
Cyrilvallez Sep 17, 2025
8e837f6
Consistent naming for images kwargs (#40834)
zucchini-nlp Sep 17, 2025
a2ac4de
Remove nested import logic for torchvision (#40940)
yonigozlan Sep 17, 2025
529d3a2
Fix `Glm4vModelTest::test_eager_matches_fa2_generate` (#40947)
ydshieh Sep 17, 2025
58e13b9
Update expected values for some `test_speculative_generation` (#40949)
ydshieh Sep 17, 2025
3bb1b48
Standardize audio embedding function name for audio multimodal models…
jackzhxng Sep 18, 2025
449da6b
Add FlexOlmo model (#40921)
2015aroras Sep 18, 2025
438343d
Don't list dropout in eager_paged_attention_forward (#40924)
cyyever Sep 18, 2025
2ccc6ca
v4.57.0 Branch (#41310)
ArthurZucker Oct 3, 2025
8ac2b91
Release: v4.57.0
LysandreJik Oct 3, 2025
0645c9e
fix optax dep issue
Cyrilvallez Oct 14, 2025
a92b1e8
remove offload_state_dict from kwargs
Cyrilvallez Oct 14, 2025
2fbd25c
Fix bnb fsdp loading for pre-quantized checkpoint (#41415)
SunMarc Oct 9, 2025
e0c6038
Fix tests fsdp (#41422)
SunMarc Oct 9, 2025
c6ae19e
Fix trainer for py3.9 (#41359)
SunMarc Oct 6, 2025
8cb5963
Release: v4.57.1
Cyrilvallez Oct 14, 2025
c77bef9
Merge tag 'v4.57.1' into whisper-transformers-4.57
mavibirdesmi Oct 20, 2025
5ef4152
fix: add #39938
mavibirdesmi Oct 28, 2025
8c3fd53
chore: possible fix by indent
mavibirdesmi Oct 28, 2025
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
The diff you're trying to view is too large. We only load the first 3000 changed files.
173 changes: 103 additions & 70 deletions .circleci/config.yml
Original file line number Diff line number Diff line change
Expand Up @@ -7,12 +7,25 @@ parameters:
nightly:
type: boolean
default: false
GHA_Actor:
type: string
default: ""
GHA_Action:
type: string
default: ""
GHA_Event:
type: string
default: ""
GHA_Meta:
type: string
default: ""

jobs:
# Ensure running with CircleCI/huggingface
check_circleci_user:
docker:
- image: python:3.10-slim
resource_class: small
parallelism: 1
steps:
- run: echo $CIRCLE_PROJECT_USERNAME
Expand All @@ -34,64 +47,44 @@ jobs:
- run: echo 'export "GIT_COMMIT_MESSAGE=$(git show -s --format=%s)"' >> "$BASH_ENV" && source "$BASH_ENV"
- run: mkdir -p test_preparation
- run: python utils/tests_fetcher.py | tee tests_fetched_summary.txt
- store_artifacts:
path: ~/transformers/tests_fetched_summary.txt
- run: |
if [ -f test_list.txt ]; then
cp test_list.txt test_preparation/test_list.txt
else
touch test_preparation/test_list.txt
fi
- run: |
if [ -f examples_test_list.txt ]; then
mv examples_test_list.txt test_preparation/examples_test_list.txt
else
touch test_preparation/examples_test_list.txt
fi
- run: |
if [ -f filtered_test_list_cross_tests.txt ]; then
mv filtered_test_list_cross_tests.txt test_preparation/filtered_test_list_cross_tests.txt
else
touch test_preparation/filtered_test_list_cross_tests.txt
fi
- run: |
if [ -f doctest_list.txt ]; then
cp doctest_list.txt test_preparation/doctest_list.txt
else
touch test_preparation/doctest_list.txt
fi
- run: |
if [ -f test_repo_utils.txt ]; then
mv test_repo_utils.txt test_preparation/test_repo_utils.txt
else
touch test_preparation/test_repo_utils.txt
fi
- run: python utils/tests_fetcher.py --filter_tests
- run: export "GIT_COMMIT_MESSAGE=$(git show -s --format=%s)" && echo $GIT_COMMIT_MESSAGE && python .circleci/create_circleci_config.py --fetcher_folder test_preparation
- run: |
if [ -f test_list.txt ]; then
mv test_list.txt test_preparation/filtered_test_list.txt
else
touch test_preparation/filtered_test_list.txt
if [ ! -s test_preparation/generated_config.yml ]; then
echo "No tests to run, exiting early!"
circleci-agent step halt
fi

- store_artifacts:
path: test_preparation/test_list.txt
- store_artifacts:
path: test_preparation/doctest_list.txt
- store_artifacts:
path: ~/transformers/test_preparation/filtered_test_list.txt
- store_artifacts:
path: test_preparation/examples_test_list.txt
- run: export "GIT_COMMIT_MESSAGE=$(git show -s --format=%s)" && echo $GIT_COMMIT_MESSAGE && python .circleci/create_circleci_config.py --fetcher_folder test_preparation
- run: |
if [ ! -s test_preparation/generated_config.yml ]; then
echo "No tests to run, exiting early!"
circleci-agent step halt
fi
path: test_preparation

- run:
name: "Retrieve Artifact Paths"
# [reference] https://circleci.com/docs/api/v2/index.html#operation/getJobArtifacts
# `CIRCLE_TOKEN` is defined as an environment variables set within a context, see `https://circleci.com/docs/contexts/`
command: |
project_slug="gh/${CIRCLE_PROJECT_USERNAME}/${CIRCLE_PROJECT_REPONAME}"
job_number=${CIRCLE_BUILD_NUM}
url="https://circleci.com/api/v2/project/${project_slug}/${job_number}/artifacts"
curl -o test_preparation/artifacts.json ${url} --header "Circle-Token: $CIRCLE_TOKEN"
- run:
name: "Prepare pipeline parameters"
command: |
python utils/process_test_artifacts.py

# To avoid too long generated_config.yaml on the continuation orb, we pass the links to the artifacts as parameters.
# Otherwise the list of tests was just too big. Explicit is good but for that it was a limitation.
# We used:

# https://circleci.com/docs/api/v2/index.html#operation/getJobArtifacts : to get the job artifacts
# We could not pass a nested dict, which is why we create the test_file_... parameters for every single job

- store_artifacts:
path: test_preparation/generated_config.yml
path: test_preparation/transformed_artifacts.json
- store_artifacts:
path: test_preparation/filtered_test_list_cross_tests.txt
path: test_preparation/artifacts.json
- continuation/continue:
parameters: test_preparation/transformed_artifacts.json
configuration_path: test_preparation/generated_config.yml

# To run all tests for the nightly build
Expand All @@ -102,22 +95,47 @@ jobs:
parallelism: 1
steps:
- checkout
- run: uv pip install -e .
- run: |
mkdir test_preparation
echo -n "tests" > test_preparation/test_list.txt
echo -n "all" > test_preparation/examples_test_list.txt
echo -n "tests/repo_utils" > test_preparation/test_repo_utils.txt
- run: uv pip install -U -e .
- run: echo 'export "GIT_COMMIT_MESSAGE=$(git show -s --format=%s)"' >> "$BASH_ENV" && source "$BASH_ENV"
- run: mkdir -p test_preparation
- run: python utils/tests_fetcher.py --fetch_all | tee tests_fetched_summary.txt
- run: python utils/tests_fetcher.py --filter_tests
- run: export "GIT_COMMIT_MESSAGE=$(git show -s --format=%s)" && echo $GIT_COMMIT_MESSAGE && python .circleci/create_circleci_config.py --fetcher_folder test_preparation
- run: |
echo -n "tests" > test_list.txt
python utils/tests_fetcher.py --filter_tests
mv test_list.txt test_preparation/filtered_test_list.txt
- run: python .circleci/create_circleci_config.py --fetcher_folder test_preparation
- run: cp test_preparation/generated_config.yml test_preparation/generated_config.txt
if [ ! -s test_preparation/generated_config.yml ]; then
echo "No tests to run, exiting early!"
circleci-agent step halt
fi

- store_artifacts:
path: test_preparation

- run:
name: "Retrieve Artifact Paths"
command: |
project_slug="gh/${CIRCLE_PROJECT_USERNAME}/${CIRCLE_PROJECT_REPONAME}"
job_number=${CIRCLE_BUILD_NUM}
url="https://circleci.com/api/v2/project/${project_slug}/${job_number}/artifacts"
curl -o test_preparation/artifacts.json ${url}
- run:
name: "Prepare pipeline parameters"
command: |
python utils/process_test_artifacts.py

# To avoid too long generated_config.yaml on the continuation orb, we pass the links to the artifacts as parameters.
# Otherwise the list of tests was just too big. Explicit is good but for that it was a limitation.
# We used:

# https://circleci.com/docs/api/v2/index.html#operation/getJobArtifacts : to get the job artifacts
# We could not pass a nested dict, which is why we create the test_file_... parameters for every single job

- store_artifacts:
path: test_preparation/transformed_artifacts.json
- store_artifacts:
path: test_preparation/generated_config.txt
path: test_preparation/artifacts.json
- continuation/continue:
configuration_path: test_preparation/generated_config.yml
parameters: test_preparation/transformed_artifacts.json
configuration_path: test_preparation/generated_config.yml

check_code_quality:
working_directory: ~/transformers
Expand All @@ -130,15 +148,15 @@ jobs:
parallelism: 1
steps:
- checkout
- run: uv pip install -e .
- run: uv pip install -e ".[quality]"
- run:
name: Show installed libraries and their versions
command: pip freeze | tee installed.txt
- store_artifacts:
path: ~/transformers/installed.txt
- run: python -c "from transformers import *" || (echo '🚨 import failed, this means you introduced unprotected imports! 🚨'; exit 1)
- run: ruff check examples tests src utils
- run: ruff format tests src utils --check
- run: ruff format examples tests src utils --check
- run: python utils/custom_init_isort.py --check_only
- run: python utils/sort_auto_mappings.py --check_only
- run: python utils/check_doc_toc.py
Expand All @@ -155,36 +173,51 @@ jobs:
parallelism: 1
steps:
- checkout
- run: uv pip install -e .
- run: uv pip install -e ".[quality]"
- run:
name: Show installed libraries and their versions
command: pip freeze | tee installed.txt
- store_artifacts:
path: ~/transformers/installed.txt
- run: python utils/check_copies.py
- run: python utils/check_table.py
- run: python utils/check_modular_conversion.py
- run: python utils/check_dummies.py
- run: python utils/check_repo.py
- run: python utils/check_inits.py
- run: python utils/check_pipeline_typing.py
- run: python utils/check_config_docstrings.py
- run: python utils/check_config_attributes.py
- run: python utils/check_doctest_list.py
- run: make deps_table_check_updated
- run: python utils/update_metadata.py --check-only
- run: python utils/check_docstrings.py
- run: python utils/check_support_list.py

workflows:
version: 2
setup_and_quality:
when:
not: <<pipeline.parameters.nightly>>
and:
- equal: [<<pipeline.project.git_url>>, https://github.com/huggingface/transformers]
- not: <<pipeline.parameters.nightly>>
jobs:
- check_circleci_user
- check_code_quality
- check_repository_consistency
- fetch_tests

setup_and_quality_2:
when:
not:
equal: [<<pipeline.project.git_url>>, https://github.com/huggingface/transformers]
jobs:
- check_circleci_user
- check_code_quality
- check_repository_consistency
- fetch_tests:
# [reference] https://circleci.com/docs/contexts/
context:
- TRANSFORMERS_CONTEXT

nightly:
when: <<pipeline.parameters.nightly>>
jobs:
Expand Down
Loading