Update README.md

add tokenizer

Upload trained [Ludwig](https://ludwig.ai/latest/) model weights

initial commit