Should we have this state-of-the-art language model for text classification?
https://arxiv.org/pdf/1801.06146.pdf
http://nlp.fast.ai/classification/2018/05/15/introducting-ulmfit.html
Thanks for the tip! We are currently not planning to release ULMFiT as a Tensorflow Hub module. However, we would be very supportive if there was interest in the community to create it.
@vbardiovskyg what about ELMO?
@loretoparisi: Hi, we already have an ELMO module.
@vbardiovskyg do you have idea if anyone is working on ULMFiT for TF-Hub ?
Hi, I don't know of anybody working on that at the moment.
This would be great...