Software for LLMs

We will use Python deep learning libraries, such as PyTorch and TensorFlow/Keras, to run and fine-tune large language models (LLMs).

The transformers package is compatible with both PyTorch and TensorFlow.

According to its documentation, “Transformers provides APIs and tools to easily download and train state-of-the-art pretrained models.”

This package supports various applications, including Natural Language Processing (which is the focus of this workshop), Computer Vision, Audio, and Multimodal tasks.

Note: There are packages in R/Rstudio that can do deep learning and/or use LLMs. We are not using those in this workshop.