Fine tuning
In this section we look at how to fine-tune LLMs. We do this in two ways:
- Fine-tuning OpenAI models. Sticking with the OpenAI theme, we will introduce how to use the OpenAI API to set up and run fine tuning jobs.
- Open source models. Hugging Face offers the ability to download and run local models. We will look at how to fine-tune base models on a single GPU and how to use distributed training to fine-tune on multiple GPUs.