Leaderboard for text-to-video generation models
World warming land sites
Visualize dataset distributions with facets
Explore income data with an interactive visualization tool
Explore token probability distributions with sliders
Explore tradeoffs between privacy and fairness in machine learning models
More advanced and challenging multi-task evaluation
Evaluate LLMs using Kazakh MC tasks
Display server status information
Browse LLM benchmark results in various categories
https://huggingface.co/spaces/VIDraft/mouse-webgen
Generate plots for GP and PFN posterior approximations
Analyze and visualize Hugging Face model download stats
VideoScore Leaderboard is a data visualization tool designed to evaluate and compare the performance of text-to-video generation models. It provides a centralized platform to display leaderboard tables with video scores and evaluation data, helping users track model performance, identify top-performing models, and gain insights into model strengths and weaknesses.
• Interactive Tables: Sort and filter data to focus on specific models or metrics.
• Customizable Filters: Narrow down results by evaluation criteria, model names, or date ranges.
• Real-Time Updates: Stay current with the latest model evaluations and scores.
• Visual Analytics: Gain deeper insights with charts and graphs that highlight performance trends.
• Model Comparison: Directly compare multiple models side-by-side for comprehensive analysis.
• Export Options: Download data in various formats for offline reviewing or reporting.
What models are supported by VideoScore Leaderboard?
VideoScore Leaderboard supports a wide range of text-to-video generation models, including popular ones like Pika Labs, Kaiber, and Pix2Vid.
Can I customize the metrics displayed in the leaderboard?
Yes, you can customize the metrics displayed by using the filtering options available in the tool. This allows you to focus on the specific evaluation criteria you care about.
How frequently is the leaderboard updated?
The leaderboard is updated in real-time as new evaluation data becomes available, ensuring you always have the most current information on model performance.