PyTorch-Transformers: add documentation for new embeddings

[stefan-it]

Documentation for the following embeddings should be added:

• BertEmbeddings
• OpenAIGPTEmbeddings
• OpenAIGPT2Embeddings
• TransformerXLEmbeddings
• XLNetEmbeddings
• XLMEmbeddings
• RoBERTaEmbeddings

Pooling operations and scalar mix should also get an extra section :)

[Sakib-Ar-Rahman]

Have these embeddings been included in version 0.4.2? I was following the documentation but not sure if these have made it into the flair package yet. Is there any way I can use these? I don't seem to be able to import the following:

• OpenAIGPT2Embeddings
• XLNetEmbeddings
• XLMEmbeddings
• RoBERTaEmbeddings

[alanakbik]

@Sakib-Ar-Rahman we're planning a release of 0.4.3 sometime next week. In the meantime you can install in a fresh virtual environment like this:

pip install --upgrade git+https://github.com/zalandoresearch/flair.git

[stefan-it]

Documentation was added for version 0.4.3 🤗