Port the GPT model to gluon-nlp:
https://github.com/openai/gpt-2
https://github.com/openai/finetune-transformer-lm
A good first step is to port their pre-trained models and be able to perform inference.
I've ported the pretrained model. Codes: https://github.com/sxjscience/gluonnlp-gpt2
the model is merged in the model zoo in #761. we still need to move it to the API. when moving, we need to:
Most helpful comment
I've ported the pretrained model. Codes: https://github.com/sxjscience/gluonnlp-gpt2