Status: ✅ Model loaded
Built on Qwen2.5-Coder-3B-Instruct + LoRA adapter trained on NVIDIA Nemotron + Stack-4.0 agentic data.