GitLab

Requesting pad_token_id would cause an error message when it is None. Use private _pad_token instead.

In batch_encode_plus we have to ensure that the tokenizer has a pad_token_id so that, when padding, no None values are added as padding. That would happen with gpt2, openai, transfoxl.

closes https://github.com/huggingface/transformers/issues/2640

- mostly stylistic streamlining
- removed 'additional context' sections. They seem to be rarely used and might cause confusion. If more details are needed, users can add them to the 'details' section

Motivate users to @-tag authors of models to increase visibility and expand the community

- change references to pytorch-transformers to transformers
- link to code formatting guidelines

- add 'your contribution' section
- add code formatting link to 'additional context'

Prefer that general questions are asked on Stack Overflow

Streamlines usages of pytorch-transformers and pytorch-pretrained-bert. Add link to the README for the migration guide.

cc @julien-c @thomwolf

ping @lysandrejik

cc @julien-c @@thomwolf

cc @julien-c