updating all configs, adding both fast and legacy tokenizers, also adding tensorflow checkpoint for compatibility
27c33d2
josecannete
commited on