Overview

RDson/CoderO1-DeepSeekR1-Coder-32B-Preview is a 32.8 billion parameter language model, developed by RDson. It is a merged model, created using the sce merge method with mergekit. The base model for this merge was Qwen/Qwen2.5-32B.

Key Components

This model integrates capabilities from two specialized models:

• deepseek-ai/DeepSeek-R1-Distill-Qwen-32B
• Qwen/Qwen2.5-Coder-32B-Instruct

Purpose

The primary goal of this merge is to combine the strengths of these models to create a robust solution for code-related tasks. By leveraging the Qwen2.5 architecture and incorporating dedicated coding models, CoderO1-DeepSeekR1-Coder-32B-Preview aims to offer enhanced performance in code generation, understanding, and instruction following. Its substantial 131072 token context length further supports complex programming challenges requiring extensive context.