huggingface/transformers

Pull Requests Commits

[tests] Stricter generate + compilation test -- no recompilations allowed (#37629)

gante committed 237 days ago

Verified 85665a42

[test] update `test_past_key_values_format` (#37614)

gante committed 237 days ago

Verified 362fa37d

Add test to ensure unknown exceptions reraising in utils/hub.py::cached_files() (#37651)

manueldeprada committed 237 days ago

Verified 1cd110c6

Support loading Gemma3 QAT GGUF models (#37649)

Isotr0py committed 237 days ago

Verified c69e2345

Restructure torchao quantization examples (#37592)

jerryzh168 committed 237 days ago

Verified 7eb1107c

[fix gemma] Set default value for output_attentions parameter in Gemma2 and Gemma… (#37633)

chenin-wang committed 237 days ago

Verified 006530d2

[fix] make legacy bnb code work (#37331)

cyr0930 committed 237 days ago

Verified 31ea547b

Fix Qwen2.5-Omni get_chunked_index chunking functionality (#37631)

imkero committed 237 days ago

Verified 5f791281

Refactor phi doc (#37583)

JihadHammoud02 committed 238 days ago

Verified fee11906

Update longformer.md (#37622)

JihadHammoud02 committed 238 days ago

Verified b2db54f6

fix link in kv_cache.md (#37652)

manueldeprada committed 238 days ago

Verified 2c60a442

Allow Exclusion of Input IDs from RepetitionPenaltyLogitsProcessor (#37625)

alex-jw-brooks committed 238 days ago

Verified a42ba80f

Remove torchvision requirement from AutoImageProcessor (#37457)

LysandreJik committed 238 days ago

Verified 10776034

[kernels] use original forward at compile time (#37604)

gante committed 238 days ago

Verified 1930e750

Fix InternVL attention when using qk_norm (38B and 78B) (#37620)

yonigozlan committed 240 days ago

Verified 6daa3eeb

chore: update model card for SigLIP (#37585)

saswatmeher committed 241 days ago

Verified 27a25bee

Fixing the example in generation strategy doc (#37598)

jeasinema committed 241 days ago

Verified e1f379bb

Deprecate modeling_utils.py classes (#37298)

qubvel committed 241 days ago

Verified 4f58fc9c

Add InternVL (2.5 MPO) (#35968)

yonigozlan committed 241 days ago

Verified a2450112

fix issue that some example with no trainer use accelerator.end_train… (#37435)

we1559 committed 241 days ago

Verified b0c6ff5e

fix 2 encoder_decoder issues on XPU (#37572)

yao-matrix committed 241 days ago

Verified 6f5014ac

[VLMs] use only `xxx_token_id` for multimodal tokens (#37573)

zucchini-nlp committed 241 days ago

Verified 2ba6b92a

Model debugger upgrades (#37391)

molbap committed 241 days ago

Verified 4afd3f48

[Gemma3] compile ✨ (#37447)

gante committed 241 days ago

Verified e5ac2308

enable 6 modeling cases on XPU (#37571)

yao-matrix committed 241 days ago

Verified a1b82563

enable 6 gemma2 cases on XPU (#37564)

yao-matrix committed 241 days ago

Verified 3cd6627c

Flag SpeechT5 flaky test (#37587)

molbap committed 241 days ago

Verified 049b75ea

[Bugfix] Fix flash-attention func param mismatch and softmax_scale default value mistake on Ascend NPU (#37575)

FightingZhen committed 241 days ago

Verified aa17cfb4

remove _run_third_party_device_tests (#37445)

jiqing-feng committed 241 days ago

Verified 14b3dbcf

Fix some GPU OOM after #37553 (#37591)

ydshieh committed 241 days ago

Verified f9742143

Newer Older