Display a Bokeh plot
Explore token probability distributions with sliders
Submit evaluations for speaker tagging and view leaderboard
Gather data from websites
Simulate causal effects and determine variable control
Create detailed data reports
Evaluate LLMs using Kazakh MC tasks
Cluster data points using KMeans
Browse LLM benchmark results in various categories
Explore how datasets shape classifier biases
Display and analyze PyTorch Image Models leaderboard
Create a detailed report from a dataset
Explore and compare LLM models through interactive leaderboards and submissions
Bloom Tokens is a web-based data visualization tool designed to help users create and display interactive plots using Bokeh. It allows users to explore and analyze data by generating dynamic and visually appealing visualizations. With Bloom Tokens, users can easily represent complex data in a clear and actionable manner.
• Interactive Visualizations: Bloom Tokens leverages Bokeh to create interactive plots that allow users to hover, zoom, and pan through data.
• Real-Time Data Support: Connect with live data sources to visualize real-time information.
• Customizable Charts: Choose from a variety of chart types and customize colors, sizes, and layouts.
• Integration with Python: Seamlessly integrate with Python scripts and libraries for data processing.
• User-Friendly Interface: An intuitive interface makes it easy to import data and generate visualizations without extensive coding knowledge.
What type of data can Bloom Tokens process?
Bloom Tokens supports datasets in CSV, Excel, and JSON formats, making it versatile for various data sources.
Is Bloom Tokens suitable for real-time data visualization?
Yes, Bloom Tokens is designed to handle real-time data, enabling users to monitor live information dynamically.
Can I customize the appearance of my visualizations?
Absolutely! Bloom Tokens allows users to customize colors, fonts, and layout options to create tailored visualizations.
How do I share my visualizations?
You can download visualizations as images or share them directly via URLs for collaborative purposes.