A TensorFlow implementation of GPT. It implements a stack of decoder blocks for autoregressive text generation, allowing you to train your own foundation models and (smaller) LLMs.
To run, simply use the command line:
python main.py
If you want to train on a custom text file (that fits in RAM) then run the following command, substituting myDataset.txt
for your text file. If you don't specify your file, it'll train on the HuggingFace Wikipedia Dataset.
python main.py --data="myDataset.txt"