v1.0.1

first release

adds the following functionalities:

• BPETokenizer: which can be used to build your tokenizer for the LLM
• Tokenizer: a base class which leverages the save and load of the vocab and merges