huggingface/transformers

Pull Requests Commits

[Grounding DINO] Add support for cross-attention in GroundingDinoMultiHeadAttention (#30364)

EduardoPach committed 2 years ago

Verified c651ea98

Add inputs embeds in generation (#30269)

zucchini-nlp committed 2 years ago

Verified 408453b4

show `-rs` to show skip reasons (#30318)

ArthurZucker committed 2 years ago

Verified 6c1295a0

[docs] LLM inference (#29791)

stevhliu committed 2 years ago

Verified e74d793a

[FEAT]: EETQ quantizer support (#30262)

dtlzhuangz committed 2 years ago

Verified b4c18a83

Add sdpa and fa2 the Wav2vec2 family. (#30121)

kamilakesbi committed 2 years ago

Verified 569743f5

FIX / PEFT: Pass device correctly to peft (#30397)

younesbelkada committed 2 years ago

Verified 367a0dbd

Fix DETA save_pretrained (#30326)

qubvel committed 2 years ago

Verified 13b3b90a

Jamba: fix left-padding test (#30389)

gante committed 2 years ago

Verified 6c7335e0

Fix layerwise GaLore optimizer hard to converge with warmup scheduler (#30372)

hiyouga committed 2 years ago

Verified f3b3533e

Terminator strings for generate() (#28932)

Rocketknight1 committed 2 years ago

Verified 0d84901c

Update docstrings for text generation pipeline (#30343)

Rocketknight1 committed 2 years ago

Verified 0e9d44d7

`Llama` family, fix `use_cache=False` generation (#30380)

ArthurZucker committed 2 years ago

Verified 2d92db84

Add FSDP config for CPU RAM efficient loading through accelerate (#30002)

helloworld1 committed 2 years ago

Verified f16caf44

GenerationConfig: warn if pad token is negative (#30187)

zucchini-nlp committed 2 years ago

Verified 91389357

Enable multi-device for more models (#30379)

jla524 committed 2 years ago

Verified 8b02bb6e

Nits for model docs (#29795)

merveenoyan committed 2 years ago

Verified b20b0179

[Grounding DINO] Add resources (#30232)

NielsRogge committed 2 years ago

Verified 8c12690c

Add TF swiftformer (#23342)

joaocmd committed 2 years ago

Verified d2cec09b

Fix config + attn_implementation in AutoModelForCausalLM.from_pretrained (#30299)

hiyouga committed 2 years ago

Verified 21c912e7

Do not remove half seq length in generation tests (#30016)

zucchini-nlp committed 2 years ago

Verified b1cd4874

Update unwrap from accelerate (#29933)

SunMarc committed 2 years ago

Verified b4fd49b6

Restore casting of masked_spec_embed (#30336)

ylacombe committed 2 years ago

Verified fbd8c51f

Deprecate default chat templates (#30346)

Rocketknight1 committed 2 years ago

Verified 0927bfd0

Transformers Metadata (#30344)

LysandreJik committed 2 years ago

Verified e67ccf06

parallel job limit for doctest (#30342)

ydshieh committed 2 years ago

Verified 32d4bef6

[Whisper] Fix slow tests (#30152)

sanchit-gandhi committed 2 years ago

Verified 4ed0e51c

Pipeline: fix `pad_token_id` again (#30338)

zucchini-nlp committed 2 years ago

Verified 91472cf5

[Feature Extractors] Fix kwargs to pre-trained (#30260)

sanchit-gandhi committed 2 years ago

Verified cd09a8df

feat: Upgrade Weights & Biases callback (#30135)

parambharat committed 2 years ago

Verified 4ab7a282

Newer Older