huggingface/transformers

Pull Requests Commits

[Flax] Add remat (gradient checkpointing) (#17843)

sanchit-gandhi committed 3 years ago

Verified 485bbe79

higher atol to avoid flaky trainer test failure (#17979)

ydshieh committed 3 years ago

Verified 664688b9

Fix FlaxBigBirdEmbeddings (#17842)

ydshieh committed 3 years ago

Verified 8bb2c387

add ONNX support for BLOOM (#17961)

NouamaneTazi committed 3 years ago

Verified b68d408f

fixing fsdp autowrap functionality (#17922)

pacman100 committed 3 years ago

Verified 462b7f3a

fix `bias` keyword argument in TFDebertaEmbeddings (#17940)

WissamAntoun committed 3 years ago

Verified 3a064bd4

Update expected values in CodeGen tests (#17888)

ydshieh committed 3 years ago

Verified 569b679a

Fix typo in perf_train_gpu_one.mdx (#17983)

aliencaocao committed 3 years ago

Verified cb425024

skip some gpt_neox tests that require 80G RAM (#17923)

ydshieh committed 3 years ago

Verified 14fb8a63

feat: add pipeline registry abstraction (#17905)

aarnphm committed 3 years ago

Verified 49cd736a

Add ONNX support for LayoutLMv3 (#17953)

regisss committed 3 years ago

Verified 9cb7cef2

skip some ipex tests until it works with torch 1.12 (#17964)

ydshieh committed 3 years ago

Verified fe140464

CLI: convert sharded PT models (#17959)

gante committed 3 years ago

Verified 91e1f24e

Fix number of examples for iterable dataset in distributed training (#17951)

sgugger committed 3 years ago

Verified f25457b2

[Pipelines] Add revision tag to all default pipelines (#17667)

patrickvonplaten committed 3 years ago

Verified e4d25885

Unifying training argument type annotations (#17934)

jannisborn committed 3 years ago

Verified 4f8361af

Fix GPT-NeoX-20B past handling, attention computation (#17811)

zphang committed 3 years ago

Verified 205bc415

Flax t5 Encoder (#17784)

crystina-z committed 3 years ago

Verified 692e61e9

Fix #17893, removed dead code (#17917)

clefourrier committed 3 years ago

Verified eb1493b1

add MobileViT model (#17354)

hollance committed 3 years ago

Verified fbc7598b

Fix prepare_tf_dataset when drop_remainder is not supplied (#17950)

Rocketknight1 committed 3 years ago

Verified 5feac3d0

ExplicitEnum subclass str (JSON dump compatible) (#17933)

Bram Vanroy committed 3 years ago

Verified bc019b0e

PyTorch 1.12.0 for scheduled CI (#17949)

ydshieh committed 3 years ago

Verified b089cca3

OPT - Fix Softmax NaN in half precision mode (#17437)

younesbelkada committed 3 years ago

Verified d444edb3

Use explicit torch version in deepspeed CI (#17942)

ydshieh committed 3 years ago

Verified 9fe2403b

fix regexes with escape sequence (#17943)

stas00 committed 3 years ago

Verified 4c722e9e

Fix all is_torch_tpu_available issues (#17936)

muellerzr committed 3 years ago

Verified 7c4c6f60

Fix img seg tests (load checkpoints from `hf-internal-testing`) (#17939)

mishig25 committed 3 years ago

Verified 77b76672

Add MVP model (#17787)

StevenTang1998 committed 3 years ago

Verified 3cff4cc5

Skip failing test until they are fixed.

sgugger committed 3 years ago

8f400775

Newer Older