luckeciano/Qwen-2.5-7B-GRPO-NoBaseline-FisherMaskSentence-1e-4-HessianMaskSentence-0.01-v2_2094
The luckeciano/Qwen-2.5-7B-GRPO-NoBaseline-FisherMaskSentence-1e-4-HessianMaskSentence-0.01-v2_2094 is a 7.6 billion parameter language model, fine-tuned from Qwen/Qwen2.5-Math-7B. It was trained using the GRPO method, a technique designed to enhance mathematical reasoning in language models, on the MATH-lighteval dataset. This model is specifically optimized for advanced mathematical problem-solving and reasoning tasks, leveraging its specialized training to potentially outperform general-purpose models in this domain. Its 32768 token context length supports complex problem inputs.
Loading preview...
Overview
This model, luckeciano/Qwen-2.5-7B-GRPO-NoBaseline-FisherMaskSentence-1e-4-HessianMaskSentence-0.01-v2_2094, is a specialized 7.6 billion parameter language model. It is a fine-tuned variant of the Qwen/Qwen2.5-Math-7B base model, specifically enhanced for mathematical reasoning capabilities. The training utilized the GRPO (Generalized Reinforcement Policy Optimization) method, as introduced in the DeepSeekMath research, and was performed on the DigitalLearningGmbH/MATH-lighteval dataset.
Key Capabilities
- Enhanced Mathematical Reasoning: Optimized through GRPO training on a dedicated math dataset.
- Specialized Fine-tuning: Built upon a math-focused base model (Qwen2.5-Math-7B) for improved performance in quantitative tasks.
- Large Context Window: Supports a context length of 32768 tokens, suitable for complex mathematical problems requiring extensive input.
Good for
- Applications requiring advanced mathematical problem-solving.
- Research and development in improving LLM performance on quantitative benchmarks.
- Tasks that benefit from models specifically trained with reinforcement learning techniques for reasoning.