Browse and evaluate language models
Merge Lora adapters with a base model
Find and download models from Hugging Face
Push a ML model to Hugging Face Hub
Benchmark AI models by comparison
Display model benchmark results
Submit models for evaluation and view leaderboard
View and submit LLM evaluations
Teach, test, evaluate language models with MTEB Arena
Multilingual Text Embedding Model Pruner
Calculate memory usage for LLM models
Evaluate RAG systems with visual analytics
Export Hugging Face models to ONNX
The Hebrew LLM Leaderboard is a tool designed for benchmarking and evaluating language models specifically for the Hebrew language. It provides a comprehensive platform to compare the performance of different language models, helping users identify the best model for their needs. The leaderboard is regularly updated with the latest models and their benchmark results, making it a valuable resource for researchers, developers, and users of Hebrew language models.
• Comprehensive Benchmarking: Evaluate language models based on multiple metrics and datasets specific to Hebrew. • Performance Metrics: Access detailed performance metrics, including accuracy, F1-score, and other relevant benchmarks. • Model Comparison: Compare different models side-by-side to understand their strengths and weaknesses. • Filtering Options: Filter models based on parameters like model size, training data, and specific tasks (e.g., translation, summarization). • Regular Updates: Stay informed with the latest models and their performance data. • User-Friendly Interface: Easy navigation and visualization of results for both technical and non-technical users. • Community Contributions: Contribute to the leaderboard by submitting new models or datasets.
What is the purpose of the Hebrew LLM Leaderboard?
The purpose of the Hebrew LLM Leaderboard is to provide a centralized platform for evaluating and comparing the performance of Hebrew language models, helping users make informed decisions about which model to use for their specific needs.
How are models evaluated on the leaderboard?
Models are evaluated based on various metrics such as accuracy, F1-score, and task-specific benchmarks using Hebrew datasets. The evaluation process is transparent and regularly updated to reflect the latest advancements in language modeling.
Can I contribute to the Hebrew LLM Leaderboard?
Yes, you can contribute by submitting new models, datasets, or benchmark results. Contributions are welcome and help maintain the leaderboard as a comprehensive resource for the Hebrew language model community.