huggingface/transformers

Pull Requests Commits

:rotating_light: [`Attn`] New attn mask interface everywhere (#42848)

vasqu committed 26 days ago

Verified 4b8ba25a

Fix video interpolation in pe_audio_video (#43811)

Rocketknight1 committed 26 days ago

Verified 7d5681b3

[docs] Remove pipeline() examples from summarization/translation tasks (#43831)

Mr-Neutr0n committed 26 days ago

Verified 018af00b

Update ModelType for Unigram tokenizer (#43860)

pavel-esir committed 26 days ago

Verified 1f62d4ed

Add GlmMoeDsa (#43858)

Cyrilvallez committed 26 days ago

Verified 33a8e68d

delete unnecessary code to make moe compatible to full graph compile (#43855)

kaixuanliu committed 26 days ago

Verified 3b4a064d

Delete batch_split from EncoderDecoderCache (#43814)

cyyever committed 26 days ago

Verified 0f4681a3

Adding Support for Qwen3.5 (#43830)

bozheng-hit committed 26 days ago

Verified fc913722

fix gptoss crash in tp (#43853)

sywangyi committed 26 days ago

Verified 161c4303

Add id and resume support to SwanLab integration (#43719)

i-pj committed 26 days ago

Verified 41833e3e

Remove GPU tracking from TrackioCallback and remove env var support (#43371)

qgallouedec committed 26 days ago

Verified ab830e29

Fix incorrect timestamp calculation in Qwen3VL Processor (#43659)

jonathan-fulton committed 26 days ago

Verified 2c3926ec

add `do_sample=False` to qwen2_5_vl model tests to stablize the output (#43728)

kaixuanliu committed 26 days ago

Verified e0cb4d18

Fix slack-report workflow file (#43851)

ydshieh committed 26 days ago

Verified 9e5c7d16

Fix markdown documentation (#43076)

cyyever committed 29 days ago

Verified 7769f660

[docs] reveal more in toctree (#43808)

stevhliu committed 29 days ago

Verified dd360ad2

Fix DeepSpeed model preparation logic in Trainer class (#43780)

qgallouedec committed 29 days ago

Verified f70ec1e7

Add TRL CI bot workflow to trigger tests on PR comments (#43809)

qgallouedec committed 29 days ago

Verified 5ba28289

fix(flaky): enforce manual seed to reduce flakiness (#43794)

tarekziade committed 29 days ago

Verified 0c89522f

[Trainer] use deepspeed SP process group when Accelerate doesn’t build a mesh (#43799)

kashif committed 29 days ago

Verified e61eed44

[core] Faster and thread-safe `check_model_inputs` implementation (#43765)

Cyrilvallez committed 29 days ago

Verified f8f2834e

Refactor trainer data_collator and callbacks tests (#43776)

SunMarc committed 29 days ago

Verified 95007e59

fix: Prevent AutoTokenizer type mismatch from directory name substrin… (#43791)

tarekziade committed 29 days ago

Verified b9042c4e

[`Repo Consistency`] Fix rms norm (#43803)

vasqu committed 29 days ago

Verified 0e9609c5

Remove unconditional train_batch_size assignment (#43770)

lordaarush committed 29 days ago

Verified 5a1016a9

fix(testing): Fix BLOOM tokenizer, CLAP audio features, and CLVP text tester usage in tests (#43798)

harshaljanjani committed 29 days ago

Verified 711f2797

[`Modular Dependencies`] Fixup qwen rms norms (#43772)

vasqu committed 29 days ago

Verified 8c3ac8f1

unpin torch for CircleCI (#43790)

ydshieh committed 29 days ago

Verified 49dd2979

:rotating_light: Modify ModernBERT's default attention implementation to stop using FA (#43764)

YangKai0616 committed 29 days ago

Verified 2db2c887

Fix-release-ubild (#43773)

ArthurZucker committed 29 days ago

Verified 0b2900dd

Newer Older