Pre-v5 update for the tokeniser (training date pushed to the 25th)

Updated .gitignore

Training update with more data and 2 epochs

Fixed character mapping, training with 8 epochs

Major training update

Upload folder using huggingface_hub