---
license: apache-2.0
base_models:
- HuggingFaceTB/SmolLM3-3B (87.0% on 100 samples. Model run only once.)
- TroglodyteDerivations/symbolic-math-qwen2p5-1p5b-lora
- Kevinmastascusa/qwen2p5-math-1p5b-merged
- Kevinmastascusa/symbolic-math-qwen2p5-1p5b-lora
tags:
- gsm8k
- llm
- model_accuracy
- fine-tuning
- merged_model
- gsm8k-style-llm-math-problem-solving
- mathematical-reasoning-and-word-problems
datasets:
- openai/gsm8k
---

## Model Descriptions

Models: HuggingFaceTB/SmolLM3-3B | Symbolic-Math-Qwen2.5-1.5B-LoRA | Qwen2.5-Math-1.5B-Merged a Fine-Tuned version of Qwen2.5-1.5B specifically optimized for solving mathematical word problems using chain-of-thought reasoning. The model was trained with LoRA adapters to enhance its mathematical reasoning capabilities while maintaining the base model's general language understanding.


@software{HuggingFaceTB/SmolLM3-3B, symbolic-math-qwen2p5-1p5b-lora, qwen2p5-math-1p5b-merged
  title = { HuggingFaceTB/SmolLM3-3B & Symbolic-Math-Qwen2.5-1.5B-LoRA: Enhanced Mathematical Reasoning Model},
  author = {TroglodyteDerivations},
  year = {2025},
  note = {qwen2p5-math-1p5b-merged: Fine-tuned for GSM8K mathematical problem solving}
}