Kepler Reasoning 7B

Kepler Reasoning 7B is a local code + math reasoning model made by merging:

• Qwen/Qwen2.5-Coder-7B-Instruct
• Qwen/Qwen2.5-Math-7B-Instruct

Merge method: SLERP.

Goal: strong local coding + math reasoning.

Intended Use

• coding help
• Python problem solving
• math reasoning
• algebra and word problems
• local inference after GGUF quantization

Limitations

This is a 7B local model. It is not expected to beat frontier cloud models overall.