GitLab

Co-Authored-By: Santosh-Gupta <San.Gupta.ML@gmail.com>

Add common getter and setter for input_embeddings & output_embeddings

Fix token order in xlnet preprocessing for SQuAD

Fix https://github.com/huggingface/transformers/issues/1673

Error raised by "tmp_eval_loss += tmp_eval_loss.item()" when using multi-gpu

Fixed training for TF XLM

Templates and explanation for adding a new model and example script

run_tf_glue works with all tasks

Versioning in documentation

Clean roberta model & all tokenizers now add special tokens by default (breaking change)

[WIP] Sequence generation using pretrained BERT

Co-Authored-By: Stefan Schweter <stefan.schweter@bsb-muenchen.de>

Co-Authored-By: Stefan Schweter <stefan.schweter@bsb-muenchen.de>