Skip to content

Add Qwen 2.5 Coder Model Integration Add/qwen coder #6

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 5 commits into
base: master
Choose a base branch
from

Conversation

codewithdark-git
Copy link

@codewithdark-git codewithdark-git commented Apr 23, 2025

📝 Description

Added support for Qwen 2.5 Coder model (1.5B) as an alternative to the default phi-4-mini model for enhanced coding tasks.

🔍 Changes Overview

  • Added new AgentQwen25Coder class
  • Updated model registry and configuration
  • Added Hugging Face authentication support
  • Optimized model parameters for coding tasks
  • Updated documentation and requirements

📋 Technical Details

  • Model: Qwen2.5-Coder-1.5B-Instruct-F16.gguf
  • Repository: unsloth/Qwen2.5-Coder-1.5B-Instruct-GGUF
  • Context window: 2048 tokens
  • Memory optimizations: Uses mlock and mmap
  • Threading: Dynamic thread allocation based on system capabilities

✅ Testing Checklist

  • Model downloads correctly
  • Authentication works with HF token
  • Command generation is accurate
  • Memory usage is optimized
  • Error handling is robust
  • Documentation is updated

🧪 Test Instructions

  1. Set Hugging Face token:
# The token grants access when you need to use the private Model or the restricted Model.
export HUGGINGFACE_TOKEN=your_token_here 
  1. Run basic command:
open-codex --model qwen-2.5-coder "find python files"
  1. Output:
open-codex --model qwen-2.5-coder "find python files modified today"
🚀 Loading Qwen model...

✨ Model ready!
Using model: qwen-2.5-coder
==================================== Command Found ====================================
find . -name "*.py" -mtime 0
======================================= Options =======================================
What would you like to do?
[e] Execute command
[c] Copy to clipboard
[a] Abort

Press key: 
================================== Executing Command ==================================
[18:57:05]
Running: find . -name "*.py" -mtime 0

./src/open_codex/resources/__init__.py
./src/open_codex/interfaces/llm_agent.py
./src/open_codex/interfaces/__init__.py
./src/open_codex/main.py
./src/open_codex/agents/phi_4_mini.py
./src/open_codex/agents/qwen_25_coder.py
./src/open_codex/agents/__init__.py
./src/open_codex/agent_builder.py

✓ Command completed successfully
Finished at [18:57:05]

🔗 Related Issues
Closes #4

@codewithdark-git codewithdark-git changed the title Add/qwen coder Add Qwen 2.5 Coder Model Integration Add/qwen coder Apr 23, 2025
@codingmoh
Copy link
Owner

hey, thanks so much! This looks great, I'll have a deeper look at the changes!

@codingmoh codingmoh added the enhancement New feature or request label Apr 26, 2025
@codingmoh
Copy link
Owner

@codewithdark-git - Looks really fine I just added a few changes to overall setup, essentially refactored agent creation, I'll add you as a reviewer to the pr #10

check it out, you'll see what I've done. We can decide then, whether it makes most sense, to merge yours or mine first

thanks

@codewithdark-git
Copy link
Author

Now I close the Pull Request #6 because the packages now support the Ollama model and the Ollama has the Qwen 2.5 Coder. @codingmoh

@codingmoh
Copy link
Owner

I was still planning to add native support for Qwen – so I would have rebased and merged your branch if you were still up for it.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request
Projects
None yet
Development

Successfully merging this pull request may close these issues.

add qwen2.5-coder support
2 participants