huggingface/transformers

Pull Requests Commits

[test_all] check mem pytest_num_workers 56b64bf

ydshieh committed 1 year ago

41033b63

[test_all] check mem pytest_num_workers 56b64bf

ydshieh committed 1 year ago

ad634b81

Inaccurate code example within inline code-documentation (#29661)

MysteryManav committed 1 year ago

Verified 56b64bf1

Allow apply_chat_template to pass kwargs to the template and support a dict of templates (#29658)

Rocketknight1 committed 1 year ago

Verified 48fbab73

Generate: handle `cache_position` update in `generate` (#29467)

gante committed 1 year ago

Verified 23db187d

Fix PVT v2 tests (#29660)

ydshieh committed 1 year ago

Verified 7b87ecb0

Add `dataset_revision` argument to `RagConfig` (#29610)

ydshieh committed 1 year ago

Verified 2cc3cc83

Fix TPU checkpointing inside Trainer (#29657)

shub-kris committed 1 year ago

Verified 956f44f1

[`PEFT`] Fix `save_pretrained` to make sure adapters weights are also saved on TPU (#29388)

shub-kris committed 1 year ago

Verified c9e3c0b4

Add newly added PVTv2 model to all README files. (#29647)

robinverduijn committed 1 year ago

Verified b4b96251

[docs] Remove broken ChatML format link from chat_templating.md (#29643)

aaronjimv committed 1 year ago

Verified f738ab3b

Add PvT-v2 Model (#26812)

FoamoftheSea committed 1 year ago

Verified 1fc505b8

Fix `multi_gpu_data_parallel_forward` for `MusicgenTest` (#29632)

ydshieh committed 1 year ago

Verified fe085560

Fix batching tests for new models (Mamba and SegGPT) (#29633)

zucchini-nlp committed 1 year ago

Verified 5ac264d8

Refactor TFP call to just sigmoid() (#29641)

Rocketknight1 committed 1 year ago

Verified 31d01150

[tests] make `test_trainer_log_level_replica` to run on accelerators with more than 2 devices (#29609)

faaany committed 1 year ago

Verified a7e5e154

[`Mask2Former`] Move normalization for numerical stability (#29542)

amyeroberts committed 1 year ago

Verified 3b6e95ec

Add support for FSDP+QLoRA and DeepSpeed ZeRO3+QLoRA (#29587)

pacman100 committed 1 year ago

Verified 350c5d15

[docs] Spanish translate chat_templating.md & yml addition (#29559)

njackman-2344 committed 1 year ago

Verified d3801aae

[PyTorch/XLA] Fix extra TPU compilations introduced by recent changes (#29158)

alanwaketan committed 1 year ago

Verified b340d907

Llama: allow custom 4d masks (#29618)

gante committed 1 year ago

Verified 1e21c4fb

[`MaskFormer`, `Mask2Former`] Use einsum where possible (#29544)

amyeroberts committed 1 year ago

Verified 88a4f68f

Fix minor typo: infenrece => inference (#29621)

DriesVerachtert committed 1 year ago

Verified 62478857

[generate] deprecate forced ids processor (#29487)

sanchit-gandhi committed 1 year ago

Verified fafe9093

Adds pretrained IDs directly in the tests (#29534)

LysandreJik committed 1 year ago

Verified 11bbb505

Warn about tool use (#29628)

LysandreJik committed 1 year ago

Verified 38bff8c8

[Whisper] Deprecate forced ids for v4.39 (#29485)

sanchit-gandhi committed 1 year ago

Verified 4afead8a

Core: Fix copies on main (#29624)

younesbelkada committed 1 year ago

Verified 9acce7de

[Flash Attention 2] Add flash attention 2 for GPT-J (#28295)

bytebarde committed 1 year ago

Verified be3fd8a2

[`Gemma`] Supports converting directly in half-precision (#29529)

younesbelkada committed 1 year ago

Verified d522afea

Older