filmov
tv
LIDA | Automatically Generate Visualization with LLMs | The Future of Data Visualization
![preview_player](https://i.ytimg.com/vi/6xcCwlDx6f8/maxresdefault.jpg)
Показать описание
🚀 Introducing LIDA - the latest tool from @Microsoft that automatically generates visualizations and infographics from data using large language models like ChatGPT, Llama2, etc. 📊
In this video, we’ll demo LIDA’s features and show you how it can generate (without any programming language) visualization and create stunning graphics. 🎨
LIDA comprises of 4 modules:
1. A SUMMARIZER that converts data into a rich but compact natural language summary. 📝
2. A GOAL EXPLORER that enumerates visualization goals given the data, 🔍
3. A VISGENERATOR that generates, refines, executes, and filters visualization code. 💻
4. An INFOGRAPHER module that yields data-faithful stylized graphics using image generation models. 🖼️ WIP
Chapter -
00:00:56 System Architecture of LIDA.
00:02:00 Quick Demo.
00:03:47 Data Description.
00:04:40 Goal Exploration.
00:05:22 Visualization Generation.
00:08:20 Explanation of the graph generated.
00:09:15 Evaluation of the graph generated.
00:10:08 Generate recommendations.
00:13:54 Local setup.
00:18:56 Start web UI.
00:22:20 Refine or modify the code for the chart using a prompt.
00:23:55 Quick demo of LIDA python package in Jupyter notebook.
Don’t forget to check out LIDA on GitHub! 🔗
Repo for this demo -
Follow me for more updates:
In this video, we’ll demo LIDA’s features and show you how it can generate (without any programming language) visualization and create stunning graphics. 🎨
LIDA comprises of 4 modules:
1. A SUMMARIZER that converts data into a rich but compact natural language summary. 📝
2. A GOAL EXPLORER that enumerates visualization goals given the data, 🔍
3. A VISGENERATOR that generates, refines, executes, and filters visualization code. 💻
4. An INFOGRAPHER module that yields data-faithful stylized graphics using image generation models. 🖼️ WIP
Chapter -
00:00:56 System Architecture of LIDA.
00:02:00 Quick Demo.
00:03:47 Data Description.
00:04:40 Goal Exploration.
00:05:22 Visualization Generation.
00:08:20 Explanation of the graph generated.
00:09:15 Evaluation of the graph generated.
00:10:08 Generate recommendations.
00:13:54 Local setup.
00:18:56 Start web UI.
00:22:20 Refine or modify the code for the chart using a prompt.
00:23:55 Quick demo of LIDA python package in Jupyter notebook.
Don’t forget to check out LIDA on GitHub! 🔗
Repo for this demo -
Follow me for more updates:
Комментарии