NielsRogge/transformers

Pull Requests Commits

NielsRogge committed 1 year ago

dbfe5c74

Address comment

NielsRogge committed 1 year ago

37cf49d1

Add PaliGemma resources

NielsRogge committed 1 year ago

bc09c0f7

NielsRogge committed 1 year ago

cbbda0d9

Paligemma- fix devices and dtype assignments (#31008)

molbap committed 1 year ago

Verified bdb9106f

Add split special tokens (#30772)

itazap committed 1 year ago

Verified deba7655

added interpolation for vitmae model in pytorch as well as tf. (#30732)

bhuvanmdev committed 1 year ago

Verified e5103a76

save the list of new model failures (#31013)

ydshieh committed 1 year ago

Verified a3cdff41

Quantization / TST: Fix remaining quantization tests (#31000)

younesbelkada committed 1 year ago

Verified 658b849a

Fix resume_download future warning (#31007)

Wauplin committed 1 year ago

Verified fd3c1280

allow multi-gpu (#31011)

ydshieh committed 1 year ago

Verified acbfaf69

FIX / TST: Fix expected results on Mistral AWQ test (#30971)

SunMarc committed 1 year ago

Verified ae87f979

[tests] make `test_model_parallelism` device-agnostic (#30844)

faaany committed 1 year ago

Verified 04c7c176

Perceiver interpolate position embedding (#30979)

g1y5x3 committed 1 year ago

Verified 42d8dd87

pin `uv==0.1.45` (#31006)

ydshieh committed 1 year ago

Verified 5855afd1

Do not trigger autoconversion if local_files_only (#31004)

Wauplin committed 1 year ago

Verified 03935d30

Fix training speed regression introduced by "optimize VRAM for calculating pos_bias in LayoutLM v2, v3 (#26139)" (#30988)

kkoehncke committed 1 year ago

Verified 21e259d8

add prefix space ignored in llama #29625 (#30964)

itazap committed 1 year ago

Verified 7f6e8741

Bugfix: WandbCallback uploads initial model checkpoint (#30897)

mgerstgrasser committed 1 year ago

Verified 6657fb5f

Remove deprecated properties in tokenization_nllb.py and tokenization_nllb_fast.py (#29834)

ymoslem committed 1 year ago

Verified 6d3d5b10

[Port] TensorFlow implementation of Mistral (#29708)

ariG23498 committed 1 year ago

Verified 965e98dc

Update 4 `MptIntegrationTests` expected outputs (#30989)

ydshieh committed 1 year ago

Verified 2a89673f

Add a check that warmup_setps is either 0 or >= 1 (#30764)

ymoslem committed 1 year ago

Verified 892b13d3

[tests] add `torch.use_deterministic_algorithms` for XPU (#30774)

faaany committed 1 year ago

Verified 21339a52

Fix accelerate failing tests (#30836)

SunMarc committed 1 year ago

Verified 8366b572

FIX / Docs: Minor changes in quantization docs (#30985)

younesbelkada committed 1 year ago

Verified 5a74ae6d

Finish adding support for torch.compile dynamic shapes (#30919)

warner-benjamin committed 1 year ago

Verified 046c2ad7

test_custom_4d_attention_mask skip with sliding window attn (#30833)

poedator committed 1 year ago

Verified 6739e1d2

Docs / Quantization: refactor quantization documentation (#30942)

younesbelkada committed 1 year ago

Verified 87a35181

Quantized KV Cache (#30483)

zucchini-nlp committed 1 year ago

Verified d583f131

Older