Repository contains scripts for full-weight finetuning of model from Hugging Face Transformers.
Model name, device type, batch size, epochs are given in hyperparams.py file.
Full process pipeline listed in finetuning_notebook.ipynb file.
Here's some examples of generations:
prompt='Necronomicon, min length=16, max length=32''
]prompt='Nameless, min length=16, max length=32''
