huggingface/transformers

Pull Requests Commits

Release: v4.18.0

LysandreJik committed 3 years ago

31ec2cb2

Revert "Allow the same config in the auto mapping"

sgugger committed 3 years ago

b9bf91a9

Allow the same config in the auto mapping

sgugger committed 3 years ago

b1a7dfe0

Fix TFTransfoXLLMHeadModel outputs (#16590)

ydshieh committed 3 years ago

Verified 2aef4cfe

[FlaxSpeechEncoderDecoderModel] More Rigorous PT-Flax Equivalence Tests (#16589)

sanchit-gandhi committed 3 years ago

Verified 8d57c424

[Speech2Text Doc] Fix docs (#16611)

patrickvonplaten committed 3 years ago

Verified c6563315

stas00 committed 3 years ago

Verified fb3d0df4

Use CLIP model config to set some kwargs for components (#16609)

ydshieh committed 3 years ago

Verified ae6a7a76

don't load state_dict twice when using low_cpu_mem_usage in from_pretrained (#16602)

patil-suraj committed 3 years ago

Verified 47c5c059

Fix seq2seq doc tests (#16606)

patil-suraj committed 3 years ago

Verified a2b7d19b

[Minds14] Correct quicktour (#16626)

patrickvonplaten committed 3 years ago

Verified 0bf18643

fix default num_attention_heads in segformer doc (#16612)

JunMa11 committed 3 years ago

Verified d55fcbcc

added type hints to CTRL pytorch (#16593)

anmolsjoshi committed 3 years ago

Verified b18dfd95

sgugger committed 3 years ago

208f4c10

Update summary of the tasks (#16528)

stevhliu committed 3 years ago

Verified f553c3ce

[benchmark tool] trainer-benchmark.py (#14934)

stas00 committed 3 years ago

Verified 23fc4cba

Add global_attention_mask to gen_kwargs (#16485)

JohnGiorgi committed 3 years ago

Verified b33ab4eb

[deepspeed] fix typo, adjust config name (#16597)

stas00 committed 3 years ago

Verified 9fd5e6bb

Adding missing type hints for BigBird model (#16555)

reichenbch committed 3 years ago

Verified 367558b9

Adding new train_step logic to make things less confusing for users (#15994)

Rocketknight1 committed 3 years ago

Verified 43540052

[Doctests] Correct filenaming (#16599)

patrickvonplaten committed 3 years ago

Verified 7ccacdf1

handle torch_dtype in low cpu mem usage (#16580)

patil-suraj committed 3 years ago

Verified 21decb77

made _load_pretrained_model_low_mem static + bug fix (#16548)

FrancescoSaverioZuppichini committed 3 years ago

Verified 8bf6d28c

add a template to add missing tokenization test (#16553)

SaulLu committed 3 years ago

Verified 02214cb3

Fix CI: test_inference_for_pretraining in ViTMAEModelTest (#16591)

ydshieh committed 3 years ago

Verified 765bafb8

Trigger doc build

sgugger committed 3 years ago

104c0652

initialize the default rank set on TrainerState (#16530)

andrescodas committed 3 years ago

Verified 1cd2e21d

[SpeechEncoderDecoderModel] Correct Encoder Last Hidden State Output (#16586)

sanchit-gandhi committed 3 years ago

Verified 6f9d8dc1

TF: Finalize `unpack_inputs`-related changes (#16499)

gante committed 3 years ago

Verified dad5ca83

add a test checking the format of `convert_tokens_to_string`'s output (#16540)

SaulLu committed 3 years ago

Verified be9474bd

Older