huggingface/transformers

Pull Requests Commits

Modular fix (#34802)

Cyrilvallez committed 1 year ago

Verified e3a5889e

Fix cache_utils for optimum.quanto kvcache quantization (#34750)

SunMarc committed 1 year ago

Verified ce1d328e

Gemma capping (#34282)

ArthurZucker committed 1 year ago

Verified 4bff54f9

Self-speculation (Layer-Skip Llama) (#34240)

ArthurZucker committed 1 year ago

Verified 54739a32

fix cpu bnb path (#34647)

jiqing-feng committed 1 year ago

Verified 5de58d59

Fix: siglip image processor rgb_convert is not being applied correctly. (#34301)

jp1924 committed 1 year ago

Verified 3cd78be3

Support gradient checkpointing in Qwen2VL ViT (#34724)

li-plus committed 1 year ago

Verified 0db91c3c

feat: allow to use hf-hub models for timm backbone (#34729)

cgebbe committed 1 year ago

Verified 1a0cd694

Trainer hyperparameter search kwargs docs update (#34459)

GuillemGSubies committed 1 year ago

Verified d8a5d31d

protect tensor parallel usage (#34800)

ArthurZucker committed 1 year ago

Verified dadb286f

Fix Whisper CI (#34617)

ydshieh committed 1 year ago

Verified eed11f34

Allow handling files as args for a tool created with Tool.from_space (#34687)

aymeric-roucher committed 1 year ago

Verified 759a378e

Simplify Tensor Parallel implementation with PyTorch TP (#34184)

kwen2501 committed 1 year ago

Verified 20142ab5

fix: Wrong task mentioned in docs (#34757)

ecyht2 committed 1 year ago

Verified 7df93d6f

Fix callback key name (#34762)

jung-hunsoo committed 1 year ago

Verified 7693b622

fix: Update pixel_values parameter in hf_model input (#34782)

thisisiron committed 1 year ago

Verified 1ef6c5f1

[tests] add XPU part to testing (#34778)

faaany committed 1 year ago

Verified e80a65ba

[docs] add XPU besides CUDA, MPS etc. (#34777)

faaany committed 1 year ago

Verified 9568a9df

[docs] make `empty_cache` device-agnostic (#34774)

faaany committed 1 year ago

Verified 8568bf1b

make sure to disable gradients for integer tensor (#32943)

winglian committed 1 year ago

Verified 36759f33

Fix skip of test_training_gradient_checkpointing (#34723)

dvrogozh committed 1 year ago

Verified 1c471fc3

fix a typo bug where 'id2label' was incorrectly written as 'i2label' when reading config (#34637)

ZuoChenFttS committed 1 year ago

Verified c772d4d9

Fix broken link (#34618)

ofek committed 1 year ago

Verified eb0ab3ed

VLMs: `patch_size` -> `num_image_tokens` in processing (#33424)

zucchini-nlp committed 1 year ago

Verified 1646ffb4

Add OLMo November 2024 (#34551)

2015aroras committed 1 year ago

Verified 3ee24e22

🧼 remove v4.44 deprecations (#34245)

gante committed 1 year ago

Verified 13493215

Remove FSDP wrapping from sub-models. (#34452)

eljandoubi committed 1 year ago

Verified 8d50fda6

FSDP grad accum fix (#34645)

winglian committed 1 year ago

Verified b0c0ba7b

add xpu path for awq (#34712)

jiqing-feng committed 1 year ago

Verified 52ea4aa5

fix(wandb): pass fake dataset to avoid exception in trainer (see #34455) (#34720)

CezaPasc committed 1 year ago

Verified 7b3d615b

Newer Older