- 21 Feb, 2020 5 commits
-
-
Patrick von Platen authored
* improving generation * finalized special token behaviour for no_beam_search generation * solved modeling_utils merge conflict * solve merge conflicts in modeling_utils.py * add run_generation improvements from PR #2749 * adapted language generation to not use hardcoded -1 if no padding token is available * remove the -1 removal as hard coded -1`s are not necessary anymore * add lightweight language generation testing for randomely initialized models - just checking whether no errors are thrown * add slow language generation tests for pretrained models using hardcoded output with pytorch seed * delete ipdb * check that all generated tokens are valid * renaming * renaming Generation -> Generate * make style * updated so that generate_beam_search has same token behavior than generate_no_beam_search * consistent return format for run_generation.py * deleted pretrain lm generate tests -> will be added in another PR * cleaning of unused if statements and renaming * run_generate will always return an iterable * make style * consistent renaming * improve naming, make sure generate function always returns the same tensor, add docstring * add slow tests for all lmhead models * make style and improve example comments modeling_utils * better naming and refactoring in modeling_utils * improving generation * finalized special token behaviour for no_beam_search generation * solved modeling_utils merge conflict * solve merge conflicts in modeling_utils.py * add run_generation improvements from PR #2749 * adapted language generation to not use hardcoded -1 if no padding token is available * remove the -1 removal as hard coded -1`s are not necessary anymore * add lightweight language generation testing for randomely initialized models - just checking whether no errors are thrown * add slow language generation tests for pretrained models using hardcoded output with pytorch seed * delete ipdb * check that all generated tokens are valid * renaming * renaming Generation -> Generate * make style * updated so that generate_beam_search has same token behavior than generate_no_beam_search * consistent return format for run_generation.py * deleted pretrain lm generate tests -> will be added in another PR * cleaning of unused if statements and renaming * run_generate will always return an iterable * make style * consistent renaming * improve naming, make sure generate function always returns the same tensor, add docstring * add slow tests for all lmhead models * make style and improve example comments modeling_utils * better naming and refactoring in modeling_utils * changed fast random lm generation testing design to more general one * delete in old testing design in gpt2 * correct old variable name * temporary fix for encoder_decoder lm generation tests - has to be updated when t5 is fixed * adapted all fast random generate tests to new design * better warning description in modeling_utils * better comment * better comment and error message Co-authored-by:Thomas Wolf <thomwolf@users.noreply.github.com>
-
maximeilluin authored
* Added CamembertForQuestionAnswering * fixed camembert tokenizer case
-
Bram Vanroy authored
Tensorflow does not use .eval() vs .train(). closes https://github.com/huggingface/transformers/issues/2906
-
ahotrod authored
-
Martin Malmsten authored
-
- 20 Feb, 2020 10 commits
-
-
Sam Shleifer authored
* Results same as fairseq * Wrote a ton of tests * Struggled with api signatures * added some docs
-
guillaume-be authored
* Removed unused fields in DistilBert TransformerBlock
-
srush authored
-
Scott Gigante authored
-
Funtowicz Morgan authored
* Remove warning when pad_to_max_length is not set. Signed-off-by:
Morgan Funtowicz <morgan@huggingface.co> * Move RoberTa warning to RoberTa and not GPT2 base tokenizer. Signed-off-by:
-
Funtowicz Morgan authored
Signed-off-by: -
srush authored
* initial pytorch lightning commit * tested multigpu * Fix learning rate schedule * black formatting * fix flake8 * isort * isort * . Co-authored-by:Check your git settings! <chris@chris-laptop>
-
Ilias Chalkidis authored
The model files have been updated in order to include the classification layers, based on https://github.com/huggingface/transformers/issues/2901, and now can be also used as a LM.
-
Santiago Castro authored
-
Cong authored
-
- 19 Feb, 2020 9 commits
-
-
Funtowicz Morgan authored
* Correctly return the tuple of generated file(s) when calling save_pretrained Signed-off-by:
-
Lysandre authored
-
Lysandre authored
-
Funtowicz Morgan authored
* Override build_inputs_with_special_tokens for fast impl + unittest. Signed-off-by:
-
Lysandre authored
-
Lysandre authored
-
Lysandre authored
Welcome Rust Tokenizers
-
Funtowicz Morgan authored
* Implemented fast version of tokenizers Signed-off-by:
馃殌 Signed-off-by:MOI Anthony <xn1t0x@gmail.com>
-
Bin Wang authored
-
- 18 Feb, 2020 2 commits
-
-
Sam Shleifer authored
* Skip flaky test * Style
-
VictorSanh authored
-
- 17 Feb, 2020 5 commits
-
-
VictorSanh authored
-
Manuel Romero authored
-
Stefan Schweter authored
-
Thomas Wolf authored
update .gitignore to ignore .swp files created when using vim
-
Patrick von Platen authored
-
- 16 Feb, 2020 1 commit
-
-
Manuel Romero authored
I trained the model for more epochs so I improved the results. This commit will update the results of the model and add a gif using it with **transformers/pipelines**
-
- 14 Feb, 2020 8 commits
-
-
Julien Chaumond authored
-
Timo Moeller authored
* Update model performance for correct German conll03 dataset * Adjust text * Adjust line spacing
-
Julien Chaumond authored
Co-Authored-By:Ilias Chalkidis <ihalk@di.uoa.gr>
-
Julien Chaumond authored
cc @yvespeirsman Co-Authored-By:Yves Peirsman <yvespeirsman@users.noreply.github.com>
-
Yves Peirsman authored
* Created model card for nlptown/bert-base-multilingual-sentiment * Delete model card * Created model card for bert-base-multilingual-uncased-sentiment as README
-
Julien Chaumond authored
-
Manuel Romero authored
-
Ilias Chalkidis authored
-
