huggingface/transformers

Pull Requests Commits

deepspeed init during eval fix (#24298)

pacman100 committed 2 years ago

Verified 01b55779

Update README_zh-hans.md (#24181)

CooperFu committed 2 years ago

Verified 6a081c51

[Docs] Improve docs for MMS loading of other languages (#24292)

patrickvonplaten committed 2 years ago

Verified 604a21b1

Fix image segmentation tool bug (#23897)

amyeroberts committed 2 years ago

Verified e6122c3f

[fix] bug in BatchEncoding.__getitem__ (#24293)

flybird11111 committed 2 years ago

Verified 6cd34d45

Split common test from core tests (#24284)

sgugger committed 2 years ago

Verified 372f5003

remove unused is_decoder parameter in DetrAttention (#24226)

JayL0321 committed 2 years ago

Verified a611ac9b

Fix LLaMa beam search when using parallelize (#24224)

FeiWang96 committed 2 years ago

Verified 33196b45

Fix `check_config_attributes`: check all configuration classes (#24231)

ydshieh committed 2 years ago

Verified 7504be35

Fix bug in slow tokenizer conversion, make it a lot faster (#24266)

stephantul committed 2 years ago

Verified 6793f0cf

Add MMS CTC Fine-Tuning (#24281)

patrickvonplaten committed 2 years ago

Verified 1609a436

[WIP] add EnCodec model (#23655)

hollance committed 2 years ago

Verified 0c3fdccf

Clean up old Accelerate checks (#24279)

sgugger committed 2 years ago

Verified 26a2ec56

Fix Debertav2 embed_proj (#24205)

WissamAntoun committed 2 years ago

Verified 860d11ff

`Pix2StructImageProcessor` requires `torch>=1.11.0` (#24270)

ydshieh committed 2 years ago

Verified a04ebc8b

Update check of core deps (#24277)

sgugger committed 2 years ago

Verified 8978b696

Adapt Wav2Vec2 conversion for MMS lang identification (#24234)

patrickvonplaten committed 2 years ago

Verified c4fec38b

TF: CTRL with native embedding layers (#23456)

gante committed 2 years ago

Verified 4626df50

Skip some `TQAPipelineTests` tests in past CI (#24267)

ydshieh committed 2 years ago

Verified eac8dede

QA doc: import torch before it is used (#24228)

ByronHsu committed 2 years ago

Verified 91b62f5a

Fix URL in comment for contrastive loss function (#24271)

taepd committed 2 years ago

Verified 6ab045d6

update FSDP save and load logic (#24249)

pacman100 committed 2 years ago

Verified b89fcccd

docs wrt using accelerate launcher with trainer (#24250)

pacman100 committed 2 years ago

Verified e0603d89

Skip `GPT-J` fx tests for torch < 1.12 (#24256)

ydshieh committed 2 years ago

Verified 23311314

Stop storing references to bound methods via tf.function (#24146)

Rocketknight1 committed 2 years ago

Verified 3bd1fe43

Fix how we detect the TF package (#24255)

Rocketknight1 committed 2 years ago

Verified b979a206

Update urls in warnings for rich rendering (#24136)

IvanReznikov committed 2 years ago

Verified e64d99fa

Add `torch >=1.12` requirement for `Tapas` (#24251)

ydshieh committed 2 years ago

Verified cf561d7c

Generate: GenerationConfig can overwrite attributes at from_pretrained time (#24238)

gante committed 2 years ago

Verified b1ea6b4b

TF: standardize `test_model_common_attributes` for language models (#23457)

gante committed 2 years ago

Verified 7bb6933b

Newer Older