huggingface/transformers

Pull Requests Commits

Saving `Trainer.collator.tokenizer` in when `Trainer.processing_class` is `None` (#36552)

yutong_liu committed 1 year ago

Verified 8b479e39

fix tiktoken convert to pass AddedToken to Tokenizer (#36566)

itazap committed 1 year ago

Verified 3f03c379

[ForCausalLMLoss] allow users to pass shifted labels (#36607)

stas00 committed 1 year ago

Verified 8f64b177

Disable inductor config setter by default (#36608)

HDCharles committed 1 year ago

Verified 94555437

Fix swanlab global step (#36728)

Zeyi-Lin committed 1 year ago

Verified 8733297b

Move the warning to the documentation for DataCollatorWithFlattening (#36707)

qgallouedec committed 1 year ago

Verified b815fae3

Just import torch AdamW instead (#36177)

Rocketknight1 committed 1 year ago

Verified 9be4728a

Update configuration_qwen2.py (#36735)

michaelfeil committed 1 year ago

Verified 51bd0ceb

quick fix fast_image_processor register error (#36716)

JJJYmmm committed 1 year ago

Verified 107fedc1

Add Space to Bitsandbytes doc (#36834)

MekkCyber committed 1 year ago

Verified 258dd9cc

Support tracable dynamicKVcache (#36311)

tugsbayasgalan committed 1 year ago

Verified f39f4960

One more fix for reviewer assignment (#36829)

Rocketknight1 committed 1 year ago

Verified 63c31165

[gemma 3] multimodal checkpoints + AutoModelForCausalLM (#36741)

gante committed 1 year ago

Verified 7c233980

enable OffloadedCache on XPU from PyTorch 2.7 (#36654)

yao-matrix committed 1 year ago

Verified b11050d6

Add option for ao base configs (#36526)

drisspg committed 1 year ago

Verified e8d96032

Add attention visualization tool (#36630)

ArthurZucker committed 1 year ago

Verified fef8b7f8

[Generation] remove leftover code from end-to-end compilation (#36685)

gante committed 1 year ago

Verified 0fe0bae0

Fix Device map for bitsandbytes tests (#36800)

MekkCyber committed 1 year ago

Verified a861db01

Remove `dist": "loadfile"` for `pytest` in CircleCI jobs (#36811)

ydshieh committed 1 year ago

Verified b9374a07

fix "Cannot copy out of meta tensor; no data!" issue for BartForConditionalGeneration model (#36572)

yao-matrix committed 1 year ago

Verified 4fa91b1b

Expectations test utils (#36569)

ivarflakstad committed 1 year ago

Verified 706703bb

[generate] ✨ vectorized beam search ✨ (#35802)

gante committed 1 year ago

Verified 179d02ff

Support custom dosctrings in modular (#36726)

yonigozlan committed 1 year ago

Verified 12f2ebef

Fix chameleon's TypeError because inputs_embeds may None (#36673)

YenFuLin committed 1 year ago

Verified 00915d30

Fix casting dtype for qunatization (#36799)

SunMarc committed 1 year ago

Verified 14b597f5

Fix Mistral3 tests (#36797)

yonigozlan committed 1 year ago

Verified 30580f03

Loading optimizations (#36742)

Cyrilvallez committed 1 year ago

Verified db1d4c5a

Update SHA for `tj-actions/changed-files` (#36795)

ydshieh committed 1 year ago

Verified 7baf0008

fix hqq due to recent modeling changes (#36771)

SunMarc committed 1 year ago

Verified 3017536e

Add Mistral3 (#36790)

Cyrilvallez committed 1 year ago

Verified e959530b

Newer Older