huggingface/transformers

Pull Requests Commits

Fix setup version number.

Narsil committed 1 year ago

Verified 5979ebe1

Testing 0.4.4rc0

Narsil committed 1 year ago

Verified 81ca0002

Fix documentation references to google/bit-50 model (#32407)

JuanFKurucz committed 1 year ago

Verified f5f1e52f

add values for neftune (#32399)

nbroad1881 committed 1 year ago

Verified ea5da52e

#32184 save total_vocab_size (#32240)

itazap committed 1 year ago

Verified 3d7c2f9d

Phi3 tests: fix typing for Python 3.8 (#32388)

zucchini-nlp committed 1 year ago

Verified 3bb646a5

fix: SeamlessM4TFeatureExtractor stride remainder (#32088)

TechInterMezzo committed 1 year ago

Verified 05ae3a30

Bump keras from 2.8.0 to 2.13.1 in /examples/research_projects/decision_transformer (#32393)

dependabot[bot] committed 1 year ago

Verified 847bb856

MixtralFlashAttention2: put "plus 1" inside parentheses when calculating rotary_seq_len, allowing None position_ids input. (#31500)

xenshinu committed 1 year ago

Verified 621fb3c0

fix: (issue #32124) Exception raised when running `transformers/examples/flax/language-modeling/t5_tokenizer_model.py`. (#32157)

fshp971 committed 1 year ago

Verified 7c31d05b

[generate] only require an attention mask for mps with torch<2.4 (#32367)

sanchit-gandhi committed 1 year ago

Verified c1aa0edb

RoPE: Add numerical tests ✨ (#32380)

gante committed 1 year ago

Verified 083e13b7

Update docs (#32368)

zucchini-nlp committed 1 year ago

Verified 2af199c4

Yell at the user if zero-3 init wasn't performed, but expected to have been done (#32299)

muellerzr committed 1 year ago

Verified 82efc535

Fixed Hybrid Cache Shape Initialization. (#32163)

OsamaS99 committed 1 year ago

Verified 51ab25e2

Docker: add `speech` dep to the consistency docker image (#32374)

gante committed 1 year ago

Verified e3d8285a

Offloaded KV Cache (#31325)

n17s committed 1 year ago

Verified ca59d6f7

Fix conflicting key in init kwargs in PreTrainedTokenizerBase (#31233)

OmarManzoor committed 1 year ago

Verified b4727a12

Empty list in defaults for LLaMA special tokens during weights conversion (#32342)

viktor-shcherb committed 1 year ago

Verified db8c7cae

update clean_up_tokenization_spaces warning (#32371)

itazap committed 1 year ago

Verified 2229ebe7

Check device map for saving tokenizer config on TPU (fix for issue #31971) (#32043)

ayukh committed 1 year ago

Verified 05c1f9af

add missing attribute _supports_param_buffer_assignment for gpt-j. (#32359)

nv-guomingz committed 1 year ago

Verified 9e282840

Remove size check between attn_weights and kv_seq_len for phi3 (#32339)

helunwencser committed 1 year ago

Verified 48ed24c5

[whisper] compile compatibility with long-form decoding (#31772)

sanchit-gandhi committed 1 year ago

Verified e234061c

[enc-dec cache] fix bug in indexing (#32370)

sanchit-gandhi committed 1 year ago

Verified 9451a385

LLaVa: add cache class attribute (#32278)

zucchini-nlp committed 1 year ago

Verified 453e7488

fix: warmup_steps check for training_args (#32236)

Ricardo-L-C committed 1 year ago

Verified 14ee2326

fix: Removed unnecessary `@staticmethod` decorator (#32361)

Sai-Suraj-27 committed 1 year ago

Verified 53f0c9c2

>3-5x faster torch.compile forward compilation for autoregressive decoder models (#32227)

fxmarty committed 1 year ago

Verified 92abe603

Fix error when streaming to gradio with non-string tool arguments (#32360)

aymeric-roucher committed 1 year ago

Verified b46bd8b9

Older