arabellako22/gifted_maths_unsloth_remerged_full_gemma_4
The arabellako22/gifted_maths_unsloth_remerged_full_gemma_4 is a 5.1 billion parameter Gemma 4 model, fine-tuned by arabellako22, specifically designed for generating clear, step-by-step mathematics solutions in LaTeX. This model excels at detailed mathematical reasoning and formatting, making it ideal for educational applications and mathematics assistance. It is a full-weight 16-bit merge, built upon unsloth/gemma-4-E2B-it, with a context length of 32768 tokens.
Loading preview...
Overview
The arabellako22/gifted_maths_unsloth_remerged_full_gemma_4 is a 5.1 billion parameter Gemma 4 model, fine-tuned by arabellako22, specifically engineered to provide clear, step-by-step mathematics solutions formatted in LaTeX. This model is a verified 16-bit full-weight merge of the Gifted Maths LoRA adapter with the unsloth/gemma-4-E2B-it base model, offering a robust solution for mathematical reasoning tasks.
Key Capabilities
- Step-by-step mathematical reasoning: Generates detailed solutions, breaking down complex problems into manageable steps.
- LaTeX formatting: Outputs all variables, formulas, equations, transformations, and final answers using LaTeX for clear presentation.
- Educational assistance: Designed to act as a
maths_enginefor students, ensuring no algebraic steps are skipped. - Full-weight model: Provides a complete, merged model for direct inference, serving as the source for verified GGUF releases.
Good for
- Mathematics-assistant experiments: Ideal for developing and testing AI-powered math tutors or problem solvers.
- Worked-example generation: Useful for creating educational content that demonstrates problem-solving methodologies.
- LaTeX-formatting demonstrations: Excellent for showcasing how AI can generate mathematically precise and well-formatted text.
- Source for GGUF conversions: Serves as the authoritative base for quantized versions, ensuring consistency across formats.