NielsRogge/transformers

Pull Requests Commits

Merge remote-tracking branch 'upstream/main' into add_udop

NielsRogge committed 2 years ago

0b7ee3cb

🚨 Fully revert atomic checkpointing 🚨 (#29370)

muellerzr committed 2 years ago

Verified 1681a6d4

Fix OneFormer `post_process_instance_segmentation` for panoptic tasks (#29304)

nickthegroot committed 2 years ago

Verified 8ef98628

Fix: Fixed the previous tracking URI setting logic to prevent clashes with original MLflow code. (#29096)

seanswyi committed 2 years ago

Verified 81220cba

Convert SlimSAM checkpoints (#28379)

NielsRogge committed 2 years ago

Verified 5e4b69dc

NielsRogge committed 2 years ago

7e68cedb

More improvements

NielsRogge committed 2 years ago

430f3772

Workaround for #27758 to avoid ZeroDivisionError (#28756)

tleyden committed 2 years ago

Verified c38a1227

Add mlx support to BatchEncoding.convert_to_tensors (#29406)

tidely committed 2 years ago

Verified 704b3f74

Add comment, add add_prefix_space

NielsRogge committed 2 years ago

b59e4e70

[Mixtral] Fixes attention masking in the loss (#29363)

DesmonDay committed 2 years ago

Verified 39ef3fb2

update path to hub files in the error message (#29369)

poedator committed 2 years ago

Verified 38953a75

[tests] enable automatic speech recognition pipeline tests on XPU (#29308)

faaany committed 2 years ago

Verified aade711d

Correct zero division error in inverse sqrt scheduler (#28982)

DavidAfonsoValente committed 2 years ago

Verified 831bc25d

Fix deprecated arg issue (#29372)

muellerzr committed 2 years ago

Verified 1a7c117d

Fix llama + gemma accelete tests (#29380)

SunMarc committed 2 years ago

Verified cec77334

Support subfolder with `AutoProcessor` (#29169)

JingyaHuang committed 2 years ago

Verified 15f8296a

[`YOLOS`] Fix - return padded annotations (#29300)

amyeroberts committed 2 years ago

Verified f1b1379f

🚨🚨[Whisper Tok] Update integration test (#29368)

sanchit-gandhi committed 2 years ago

Verified 0a0a279e

[`Llama + AWQ`] fix `prepare_inputs_for_generation` 🫠 (#29381)

ArthurZucker committed 2 years ago

Verified e7b98370

FIX [`quantization` / `ESM`] Fix ESM 8bit / 4bit with bitsandbytes (#29329)

younesbelkada committed 2 years ago

Verified 50db7ca4

Fix Base Model Name of LlamaForQuestionAnswering (#29258)

lenglaender committed 2 years ago

Verified 2858d6c6

Expose `offload_buffers` parameter of `accelerate` to `PreTrainedModel.from_pretrained` method (#28755)

notsyncing committed 2 years ago

Verified 5ee0868a

Fix @require_read_token in tests (#29367)

Wauplin committed 2 years ago

Verified 0ad770c3

Patch YOLOS and others (#29353)

NielsRogge committed 2 years ago

Verified bb4f816a

Avoid using uncessary `get_values(MODEL_MAPPING)` (#29362)

ydshieh committed 2 years ago

Verified 44fe1a1c

Merge remote-tracking branch 'upstream/main' into add_udop

NielsRogge committed 2 years ago

171469bf

FIX [`CI`] `require_read_token` in the llama FA2 test (#29361)

younesbelkada committed 2 years ago

Verified b647acdb

FIX [`CI`]: Fix failing tests for peft integration (#29330)

younesbelkada committed 2 years ago

Verified 8d8ac9c2

FIX [`CI` / `starcoder2`] Change starcoder2 path to correct one for slow tests (#29359)

younesbelkada committed 2 years ago

Verified 1aee9afd

Older