updating all configs, adding both fast and legacy tokenizers, also adding tensorflow checkpoint for compatibility
48e7313
josecannete
commited on