Hi, thank you very much for sharing this work.
I am having some problems running the demo (batch_demo.ipynb) from the repository. When running it with the original file (batch_kcat.csv), the batch_kcat_input_output.csv file is not generated, resulting in a FileNotFoundError later on. Also, during execution, an error related to UTF-8 encoding appears when trying to load the .json.gz file ( UnicodeDecodeError: 'utf-8' codec can't decode byte 0x8b in position 1: invalid start byte). I have tried both with the original files and with my own data.
I also wanted to ask if it is possible to fine-tuning the model with own data and, if so, if there is any guide or instruction for it.
Thanks in advance!