- Using GPT-3 to predict whether or not medications were prescribed
4
- Using NLP to automatically extract information from patient charts
5
- Fine-tuning a model to be less creative
6
- Fine-tuning a model to do two cognitive tasks at once
7
- Building a Single Source of Truth for Home Care
8
- Determining whether or not follow-up medical tests were ordered
9
- Creating a mock-up for data
10
- GPT3 Completion
11
- Building a program to process medical transcripts
12
- The Dmark
13
- Creating a bot to generate medical diagnoses
14
- Building a chatbot to answer questions about a job
15
- Building a medical chatbot
16
- Python generate data
17
- Python generate data
18
- Saving work in progress
19
- Building a machine that can write
20
- Formatting the data for training
21
- Fine-tuning a machine learning model
22
- The difficulty of diagnosing with Curie
23
- The ineffectiveness of curie with more training data
Description:
Explore advanced techniques for fine-tuning GPT-3 on medical texts to improve performance on multiple cognitive tasks and reduce hallucination in this 54-minute video. Learn how to create medical question-answering bots, predict medication prescriptions, extract information from patient charts using NLP, and fine-tune models for specific tasks. Discover methods for building a Single Source of Truth for Home Care, processing medical transcripts, and generating medical diagnoses. Gain insights into data generation, formatting for training, and the challenges of fine-tuning models like Curie. Ideal for AI enthusiasts and healthcare professionals interested in applying natural language processing to medical applications.
Fine-Tune Multiple Cognitive Tasks with GPT-3 on Medical Texts and Reduce Hallucination