Leaderboard of information retrieval models in French
View and submit LLM benchmark evaluations
Display benchmark results
Convert Hugging Face model repo to Safetensors
Explore and visualize diverse models
View LLM Performance Leaderboard
Launch web-based model application
Measure over-refusal in LLMs using OR-Bench
Calculate GPU requirements for running LLMs
Evaluate reward models for math reasoning
Calculate VRAM requirements for LLM models
Download a TriplaneGaussian model checkpoint
Measure execution times of BERT models using WebGPU and WASM
DécouvrIR is a leaderboard platform designed to showcase the performance of information retrieval (IR) models in the French language. It serves as a comprehensive resource for researchers, developers, and enthusiasts to explore and compare the effectiveness of various IR models in French-speaking contexts. By providing benchmark results, DécouvrIR aims to accelerate advancements in information retrieval technologies for French language applications.
• Benchmark Results: Displays performance metrics of leading IR models in French. • Model Comparison: Allows users to compare different models based on their effectiveness. • Detailed Metrics: Provides in-depth insights into evaluation metrics such as precision, recall, and F1-score. • User-Friendly Interface: Easy-to-navigate design for seamless exploration of model performance. • Regular Updates: Continuous updates with the latest models and benchmark results. • Filtering Options: Enables filtering by model type, dataset, or specific evaluation criteria.
What types of models are included in DécouvrIR?
DécouvrIR includes a variety of information retrieval models, ranging from traditional algorithms to state-of-the-art machine learning and deep learning approaches, all tailored for French language tasks.
How often are the benchmark results updated?
Benchmark results are updated regularly to reflect the latest advancements in information retrieval technologies and the inclusion of new models.
Can I filter models based on specific evaluation metrics?
Yes, DécouvrIR offers filtering options that allow users to view models based on specific evaluation metrics such as precision, recall, or F1-score.