01:19:56
Detail
Transcript
Intro to Large Language Models: LLM Tutorial and Disease Diagnosis LLM Lab
, Principal Data Scientist, Mark III Systems
First, we'll discuss what a large language model (LLM) is and list some of the strengths and weaknesses of these models, looking at a handful of models and approaches. We'll explain the difference between pre-training and fine-tuning. We'll discuss Input processing by showing the steps of taking an input string and tokenizing it into input IDs. We'll introduce QLoRa as a means of greatly reducing computational requirements for LLM inference and fine-tuning. We'll wrap up the concepts portion of the session by discussing Hugging Face and their transformers library.
The workshop starts with performing inference using the Hugging Face transformers library and the Falcon-7B-Instruct model. We then move to fine-tuning Falcon-7B-Instruct using the MedText dataset, where the goal is to take a prompt which describes symptoms of a medical issue and generate a diagnosis of the problem, as well as steps to treat it.
Prerequisite(s):
No prerequisites needed - Some Python and/or ML experience is helpful but not required
The workshop starts with performing inference using the Hugging Face transformers library and the Falcon-7B-Instruct model. We then move to fine-tuning Falcon-7B-Instruct using the MedText dataset, where the goal is to take a prompt which describes symptoms of a medical issue and generate a diagnosis of the problem, as well as steps to treat it.
Prerequisite(s):
No prerequisites needed - Some Python and/or ML experience is helpful but not required
Explore more training options offered by the NVIDIA Deep Learning Institute (DLI). Choose from an extensive catalog of self-paced, online courses or instructor-led virtual workshops to help you develop key skills in AI, HPC, graphics & simulation, and more.
Ready to validate your skills? Get NVIDIA certified and distinguish yourself in the industry.
Events & Trainings:
Date:
Industry:
Level:
Topic:
NVIDIA technology: Base Command,CUDA,DGX,HGX,NeMo,TensorRT
Language: English
Region: More from this playlist(101 sessions)