GitLab

* biogpt initial commit

* updated init

* fix faster decoding with use_cache

* 1. fix input_ids and input_embeds with correct device
2. added _keys_to_ignore_on_load_missing
3. updated prepare_inputs_for_generation

* add activation_dropout and scale_embedding

* replace fsmt attention with bart attention

* added test

* run make fix-copies

* doc init and fix build

* updated README with proper information

* 1. added tips to docs
2. updated BioGptTokenizer func

* 1. added tokenizer test
2. refactor tokenizer

* make fixup

* add biogpt fairseq to hf converter

* updated layer names more
similar to original checkpoints

* config update doc string and set defaults

* added "#copied" from bart model and
updated doc strings

* enable model_input_names in tokenizer

* 1.  positionalembedding depending on attention_mask
2. added attention mask to prepare for generation

* added test to verify past and generation

* BioGptLMHeadModel -> BioGptForCausalLM

* fix typo

* tokenization and test
Copyright and updated assertion

* updated Copyright and
one func at time in line

* Copyright updates and
minor doc fix

* replace assertion with ValueError

* rm extra space

* added code syntax

* revert cmnt position change

* add tokenizer to auto

* updated doc string

* tokenizer doc string update

* biogpt hub model update to microsoft/biogpt

* make fixup

* rm cmnt to fix flake8 5.0.4 vs 6 error

* Reorganize example folder

* Continue reorganization

* Change requirements for tests

* Final cleanup

* Finish regroup with tests all passing

* Copyright

* Requirements and readme

* Make a full link for the documentation

* Address review comments

* Apply suggestions from code review
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

* Add symlink

* Reorg again

* Apply suggestions from code review
Co-authored-by: Thomas Wolf <thomwolf@users.noreply.github.com>

* Adapt title

* Update to new strucutre

* Remove test

* Update READMEs
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Thomas Wolf <thomwolf@users.noreply.github.com>

* Add BERT Loses Patience (Patience-based Early Exit)

* update model archive

* update format

* sort import

* flake8

* Add results

* full results

* align the table

* refactor to inherit

* default per gpu eval = 1

* Formatting

* Formatting

* isort

* modify readme

* Add check

* Fix format

* Fix format

* Doc strings

* ALBERT & BERT for sequence classification don't inherit from the original anymore

* Remove incorrect comments

* Remove incorrect comments

* Remove incorrect comments

* Sync up with new code

* Sync up with new code

* Add a test

* Add a test

* Add a test

* Add a test

* Add a test

* Add a test

* Finishing up!

* Rename and improve example

* Add test

* slightly faster test

* style

* This breaks remy prolly

* shorter test string

* no slow

* newdir structure

* New tree

* Style

* shorter

* docs

* clean

* Attempt future import

* more import hax