huggingface/transformers

Pull Requests Commits

Update build_documentation.yml

gante committed 1 year ago

Verified 1e7746ea

Patch release v4.43.4

ArthurZucker committed 1 year ago

868d36d2

Resize embeds with DeepSpeed (#32214)

zucchini-nlp committed 1 year ago

5cea2e73

Patch release v4.43.3

ArthurZucker committed 1 year ago

47c29ccf

don't log base model architecture in wandb if log model is false (#32143)

joaonadkarni committed 1 year ago

54bc29c1

[BigBird Pegasus] set _supports_param_buffer_assignment to False (#32222)

kashif committed 1 year ago

cc75146d

[whisper] fix short-form output type (#32178)

sanchit-gandhi committed 1 year ago

cd06184c

LysandreJik committed 1 year ago

38d94bff

Fix float8_e4m3fn in modeling_utils (#32193)

SunMarc committed 1 year ago

b4a0442d

Fix resize embedding with Deepspeed (#32192)

zucchini-nlp committed 1 year ago

4672b4d7

let's not warn when someone is running a forward (#32176)

ArthurZucker committed 1 year ago

a2b6a001

RoPE: relaxed rope validation (#32182)

gante committed 1 year ago

64a90d72

LysandreJik committed 1 year ago

782bfffb

gante committed 1 year ago

cf053491

Release: v4.43.0

LysandreJik committed 1 year ago

7fa7508d

Llama 3.1 conversion

LysandreJik committed 1 year ago

26b179c9

Added additional kwarg for successful running of optuna hyperparameter search (#31924)

DeF0017 committed 1 year ago

Verified 7d92009a

feat(cache): StaticCache uses index_copy_ to avoid useless copy (#31857)

tengomucho committed 1 year ago

Verified 63700628

Fix typing to be compatible with later py versions (#32155)

amyeroberts committed 1 year ago

Verified a009fbda

Revert "Incorrect Whisper long-form decoding timestamps " (#32148)

sanchit-gandhi committed 1 year ago

Verified 3263b343

Rename Phi-3 rope scaling type (#31436)

garg-amit committed 1 year ago

Verified 034b4778

Added mamba.py backend (#30139)

alxndrTL committed 1 year ago

Verified bab32d6f

Fix video batching to videollava (#32139)

merveenoyan committed 1 year ago

Verified 9ced33ca

Fix flash attention speed issue (#32028)

Cyrilvallez committed 1 year ago

Verified a5b226ce

gguf conversion add_prefix_space=None for llama3 (#31937)

itazap committed 1 year ago

Verified a1844a32

Llama: RoPE refactor (#32135)

gante committed 1 year ago

Verified 2e113422

Modify resize_token_embeddings to ensure output type is same as input (#31979)

bayllama committed 1 year ago

Verified 5a4a76ed

Disable quick init for TapasPreTrainedModel (#32149)

daniellok-db committed 1 year ago

Verified 1535a2c9

Add YaRN and Dynamic-YaRN RoPE Scaling Methods (#30910)

mig-mfreitas committed 1 year ago

Verified 34b43211

Add method to retrieve used chat template (#32032)

KonradSzafer committed 1 year ago

Verified 7405c1c7

Older