Qwen2.5-Recursive-Coder-14B-Instruct Open Source Code Model - Efficiently Realize Code Generation and Understanding

Qwen2.5 Recursive Coder 14B Instruct

Developed by spacematt

A 14B-parameter code generation and comprehension model based on the Qwen2.5 architecture, integrated through the Model Stock method by combining multiple specialized coding models

Large Language Model

Transformers

Open Source License:Apache-2.0 #Code Generation Optimization #Multi-Model Integration #Instruction Fine-Tuning

Downloads 39

Release Time : 4/14/2025

Model Overview

This model is a fusion of multiple Qwen2.5 coding-specific models, specializing in code generation, comprehension, and programming assistance tasks with robust programming language processing capabilities

Model Features

Multi-Model Fusion

Integrates 12 specialized coding models using the Model Stock method, combining their strengths

Code Specialization

Specialized in programming-related tasks, excelling in code generation and comprehension

Large Parameter Scale

14B parameters provide powerful language understanding and generation capabilities

Open-Source License

Licensed under Apache 2.0, allowing both commercial and research use

Model Capabilities

Code Auto-Completion

Code Explanation

Code Refactoring

Bug Fixing

Code Translation

Documentation Generation

Programming Q&A

Use Cases

Software Development

IDE Plugin

Integrated into development environments to provide real-time code suggestions

Improves development efficiency by over 30%

Code Review

Automatically analyzes code quality and provides improvement suggestions

Reduces manual review workload by 50%

Education

Programming Instruction

Provides real-time feedback for students' programming exercises

Enhances learning efficiency by 40%

🚀 Model Merge Project

This project focuses on merging pre - trained language models, offering a new integrated model with enhanced capabilities. It uses the mergekit tool to achieve the merging process.

🚀 Quick Start

This is a merge of pre - trained language models created using mergekit.

✨ Features

Utilizes the Model Stock merge method.
Merges multiple Qwen2.5 - based models to potentially enhance performance and functionality.

📚 Documentation

Merge Details

Merge Method

This model was merged using the Model Stock merge method using [spacematt/Qwen2.5 - Channel - Coder - 14B - Instruct](https://huggingface.co/spacematt/Qwen2.5 - Channel - Coder - 14B - Instruct) as a base.

Models Merged

The following models were included in the merge:

[spacematt/Qwen2.5 - Composite - Coder - 14B - Instruct](https://huggingface.co/spacematt/Qwen2.5 - Composite - Coder - 14B - Instruct)
[spacematt/Qwen2.5 - Casual - Coder - 14B - Instruct](https://huggingface.co/spacematt/Qwen2.5 - Casual - Coder - 14B - Instruct)
[spacematt/Qwen2.5 - Flow - Coder - 14B - Instruct](https://huggingface.co/spacematt/Qwen2.5 - Flow - Coder - 14B - Instruct)
[spacematt/Qwen2.5 - Smooth - Coder - 14B - Instruct](https://huggingface.co/spacematt/Qwen2.5 - Smooth - Coder - 14B - Instruct)
[spacematt/Qwen2.5 - Agile - Coder - 14B - Instruct](https://huggingface.co/spacematt/Qwen2.5 - Agile - Coder - 14B - Instruct)
[spacematt/Qwen2.5 - CompositeFlow - Coder - 14B - Instruct](https://huggingface.co/spacematt/Qwen2.5 - CompositeFlow - Coder - 14B - Instruct)
[spacematt/Qwen2.5 - Vibe - Coder - 14B - Instruct](https://huggingface.co/spacematt/Qwen2.5 - Vibe - Coder - 14B - Instruct)
[spacematt/Qwen2.5 - Composite - 14B - Instruct](https://huggingface.co/spacematt/Qwen2.5 - Composite - 14B - Instruct)
[spacematt/Qwen2.5 - Sharp - Coder - 14B - Instruct](https://huggingface.co/spacematt/Qwen2.5 - Sharp - Coder - 14B - Instruct)
[spacematt/Qwen2.5 - Slick - Coder - 14B - Instruct](https://huggingface.co/spacematt/Qwen2.5 - Slick - Coder - 14B - Instruct)
[spacematt/Qwen2.5 - Wicked - Coder - 14B - Instruct](https://huggingface.co/spacematt/Qwen2.5 - Wicked - Coder - 14B - Instruct)

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: spacematt/Qwen2.5-Composite-14B-Instruct
  - model: spacematt/Qwen2.5-Composite-Coder-14B-Instruct
  - model: spacematt/Qwen2.5-Vibe-Coder-14B-Instruct
  - model: spacematt/Qwen2.5-Sharp-Coder-14B-Instruct
  - model: spacematt/Qwen2.5-Flow-Coder-14B-Instruct
  - model: spacematt/Qwen2.5-CompositeFlow-Coder-14B-Instruct
  - model: spacematt/Qwen2.5-Casual-Coder-14B-Instruct
  - model: spacematt/Qwen2.5-Smooth-Coder-14B-Instruct
  - model: spacematt/Qwen2.5-Slick-Coder-14B-Instruct
  - model: spacematt/Qwen2.5-Wicked-Coder-14B-Instruct
  - model: spacematt/Qwen2.5-Agile-Coder-14B-Instruct
  - model: spacematt/Qwen2.5-Channel-Coder-14B-Instruct
merge_method: model_stock
base_model: spacematt/Qwen2.5-Channel-Coder-14B-Instruct
normalize: true
int8_mask: true
dtype: bfloat16

📄 License

This project is licensed under the apache - 2.0 license.

Property	Details
Base Models	spacematt/Qwen2.5-Composite-Coder-14B-Instruct, spacematt/Qwen2.5-Casual-Coder-14B-Instruct, spacematt/Qwen2.5-Channel-Coder-14B-Instruct, spacematt/Qwen2.5-Flow-Coder-14B-Instruct, spacematt/Qwen2.5-Smooth-Coder-14B-Instruct, spacematt/Qwen2.5-Agile-Coder-14B-Instruct, spacematt/Qwen2.5-CompositeFlow-Coder-14B-Instruct, spacematt/Qwen2.5-Vibe-Coder-14B-Instruct, spacematt/Qwen2.5-Composite-14B-Instruct, spacematt/Qwen2.5-Sharp-Coder-14B-Instruct, spacematt/Qwen2.5-Slick-Coder-14B-Instruct, spacematt/Qwen2.5-Wicked-Coder-14B-Instruct
Library Name	transformers
Tags	mergekit, merge
License	apache - 2.0

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご