huggingface/transformers

Pull Requests Commits

SunMarc committed 1 year ago

209f58f4

Gemma2: eager attention by default (#32865)

gante committed 1 year ago

Verified 975b988b

fix: (issue #32689) `AttributeError` raised when using `Trainer` with `eval_on_start=True` in Jupyter Notebook. (#32849)

fshp971 committed 1 year ago

Verified f1d822ba

Add chat_template for tokenizer extracted from GGUF model (#32908)

Isotr0py committed 1 year ago

Verified ee8c01f8

Improve greedy search memory usage (#32895)

regisss committed 1 year ago

Verified 99d67f1a

Fix benchmark script (#32635)

ydshieh committed 1 year ago

Verified bf97d4aa

Add SynCode to llm_tutorial (#32884)

shubhamugare committed 1 year ago

Verified 92824136

FIX / Hub: Also catch for `exceptions.ConnectionError` (#31469)

younesbelkada committed 1 year ago

Verified eeea7120

CI: separate step to download nltk files (#32935)

gante committed 1 year ago

Verified 8b94d28f

FEAT / Trainer: Add adamw 4bit optimizer (#31865)

SunMarc committed 1 year ago

Verified c42d2645

fix: no need to dtype A in jamba (#32924)

xgal committed 1 year ago

Verified 6baa6f27

fix: Added missing `huggingface_hub` installation to workflows (#32891)

Sai-Suraj-27 committed 1 year ago

Verified af638c4a

Jamba: update integration tests (#32250)

gante committed 1 year ago

Verified f6e2586a

Update docker image building (#32918)

ArthurZucker committed 1 year ago

Verified 3bb7b052

fix: [whisper] don't overwrite GenerationConfig's `return_timestamps` when `return_timestamps` is not passed to `generate` function (#31296)

hrl committed 1 year ago

Verified c6d484e3

[i18n-ar] add README_ar.md to README.md (#32583)

AhmedAlmaghz committed 1 year ago

Verified 87134662

link for optimizer names (#32400)

nbroad1881 committed 1 year ago

Verified 1dde50c7

Replace `tensor.norm()` with decomposed version for CLIP executorch export (#32887)

qubvel committed 1 year ago

Verified 078d5a88

Bump nltk from 3.7 to 3.9 in /examples/research_projects/decision_transformer (#32903)

dependabot[bot] committed 1 year ago

Verified 9800e6d1

Fix: Mamba2 `norm_before_gate` usage (#32686)

vasqu committed 1 year ago

Verified c63a3d0f

fix: jamba cache fails to use torch.nn.module (#32894)

xgal committed 1 year ago

Verified 01c4fc45

Fix repr for conv (#32897)

ArthurZucker committed 1 year ago

Verified 65f4bc99

🚨🚨🚨 Update min version of accelerate to 0.26.0 (#32627)

SunMarc committed 1 year ago

Verified fd06ad54

Allow-head-dim (#32857)

ArthurZucker committed 1 year ago

Verified 13e645bb

Add tip to clarify tool calling (#32883)

Rocketknight1 committed 1 year ago

Verified 85345bb4

Docs: Fixed `whisper-large-v2` model link in docs (#32871)

Sai-Suraj-27 committed 1 year ago

Verified 37204848

Fix: Mamba2 generation mismatch between input_ids and inputs_embeds (#32694)

vasqu committed 1 year ago

Verified 61d89c19

Mamba / FalconMamba: Fix mamba left padding (#32677)

younesbelkada committed 1 year ago

Verified 93e538ae

Fix incorrect vocab size retrieval in GGUF config (#32551)

Isotr0py committed 1 year ago

Verified 59e8f191

RT-DETR parameterized batchnorm freezing (#32631)

AlanBlanchet committed 1 year ago

Verified 5f6c080b

Older