huggingface/transformers

Pull Requests Commits

Actually fix in the modular file

pcuenca committed 1 year ago

0652d891

Fix Gemma2 sliding window attention

pcuenca committed 1 year ago

6564e152

Fix : add require_read_token for gemma2 gated model (#35687)

MekkCyber committed 1 year ago

Verified a11041ff

Fix expected output for ggml test (#35686)

MekkCyber committed 1 year ago

Verified df2a812e

Fix : HQQ config when hqq not available (#35655)

MekkCyber committed 1 year ago

Verified 05063651

Update torchao.md: use auto-compilation (#35490)

martin0258 committed 1 year ago

Verified 715fdd64

Fix : adding einops lib in the CI docker for some bitsandbytes tests (#35652)

MekkCyber committed 1 year ago

Verified 4b8d1f7f

Fix `zero_shot_image_classification` documentation guide link in SigLIP (#35671)

aretrace committed 1 year ago

Verified 34f76bb6

Add-helium (#35669)

ArthurZucker committed 1 year ago

Verified c23a1c19

[i18n-ar] Translated file : docs/source/ar/tasks/token_classification.md into Arabic (#35193)

AhmedAlmaghz committed 1 year ago

Verified a3f82328

[tests] make cuda-only tests device-agnostic (#35607)

faaany committed 1 year ago

Verified 2fa876d2

[`Compile`] Only test compiling model forward pass (#35658)

ArthurZucker committed 1 year ago

Verified e6f9b034

Enable different torch dtype in sub models (#34873)

zucchini-nlp committed 1 year ago

Verified 84a67891

[`Phi`] bias should be True (#35650)

ArthurZucker committed 1 year ago

Verified 87089176

Removed some duplicated code (#35637)

Sai-Suraj-27 committed 1 year ago

Verified 91f14f1f

Fix whisper compile (#35413)

jiqing-feng committed 1 year ago

Verified b8c34d97

Fix device in rope module when using dynamic updates (#35608)

Cyrilvallez committed 1 year ago

Verified cd44bdb4

Update codeowners with individual model owners (#35595)

Rocketknight1 committed 1 year ago

Verified 15bd3e61

Skip `MobileNetV1ModelTest::test_batching_equivalence` for now (#35614)

ydshieh committed 1 year ago

Verified 1e3c6c1f

Fix flaky `test_beam_search_low_memory` (#35611)

ydshieh committed 1 year ago

Verified 04eae987

Let `EarlyStoppingCallback` not require `load_best_model_at_end` (#35101)

muellerzr committed 1 year ago

Verified b02828e4

Added error when sequence length is bigger than max_position_embeddings (#32156)

Taha1506 committed 1 year ago

Verified 0aaf124f

Use inherit tempdir makers for tests + fix failing DS tests (#35600)

muellerzr committed 1 year ago

Verified 1211e616

Fix flaky `test_custom_4d_attention_mask` (#35606)

ydshieh committed 1 year ago

Verified bbc00046

ArthurZucker committed 1 year ago

f63829c8

[WIP] Emu3: add model (#33770)

zucchini-nlp committed 1 year ago

Verified 52e1f87c

Fix flex_attention in training mode (#35605)

Cyrilvallez committed 1 year ago

Verified ccc0381d

Remove `benchmark.py` after #34275

ArthurZucker committed 1 year ago

a9bd1e62

Chat template: return vectorized output in processors (#34275)

zucchini-nlp committed 1 year ago

Verified e0646f3d

Add Moonshine (#34784)

eustlb committed 1 year ago

Verified 5f087d13

Older