Added an automated script for downloading and pre-processing wikitext-103 and wikitext-2 datasets for the LSTM language models, in the format described in https://nvidia.github.io/OpenSeq2Seq/html/language-model.html. This is very helpful for generating the files needed for training LSTM language models.
Added an automated script for downloading and pre-processing wikitext-103 and wikitext-2 datasets for the LSTM language models, in the format described in https://nvidia.github.io/OpenSeq2Seq/html/language-model.html. This is very helpful for generating the files needed for training LSTM language models.