Pergel: A Unified Benchmark for Evaluating Turkish LLMs
Visualize model performance on function calling tasks
Leaderboard of information retrieval models in French
Browse and filter machine learning models by category and modality
Display and submit language model evaluations
Submit deepfake detection models for evaluation
Demo of the new, massively multilingual leaderboard
View and submit language model evaluations
Generate and view leaderboard for LLM evaluations
Search for model performance across languages and benchmarks
Generate leaderboard comparing DNA models
Convert Hugging Face models to OpenVINO format
Convert and upload model files for Stable Diffusion
Cetvel is a unified benchmark tool designed for evaluating Turkish Large Language Models (LLMs). It provides a comprehensive framework to analyze and compare the performance of different models across a variety of tasks, making it an essential tool for researchers and developers working with Turkish NLP tasks.
• Comprehensive Task Coverage: Evaluate models on tasks such as translation, summarization, and question-answering specific to the Turkish language.
• Customizable Benchmarks: Create tailored benchmarking suites to focus on specific aspects of model performance.
• Cross-Model Comparisons: Compare multiple Turkish LLMs side-by-side to identify strengths and weaknesses.
• Detailed Reporting: Generate in-depth reports highlighting model accuracy, efficiency, and robustness.
• Integration with Popular LLMs: Supports integration with widely-used Turkish and multilingual LLMs.
What models are supported by Cetvel?
Cetvel supports a wide range of Turkish and multilingual LLMs, including but not limited to, models from leading NLP libraries.
Do I need NLP expertise to use Cetvel?
No, Cetvel is designed to be user-friendly. However, basic knowledge of NLP concepts may help in interpreting results.
Can I benchmark models in languages other than Turkish?
Cetvel is primarily optimized for Turkish, but it can be adapted for other languages with additional configuration.