1. Learn
    2. /
  2. Courses
    3. /
  3. Fine-Tuning with Llama 3
Connected

Exercise

Defining custom recipes

You're fine-tuning a pre-trained Llama model for a customer who requires specific configurations. Your plan is to use TorchTune for fine-tuning, and so need to prepare a Python dictionary that you can use to store the requirements for the custom recipe you'll use to run the fine-tuning job.

Instructions

100 XP
  • Specify the customer requirements in your dictionary: first, add the torchtune.models.llama3_2.llama3_2_1b model.
  • Add a batch size of 8 and a GPU device.