A GGUF conversion of the base (non-instruct) Qwen/Qwen2.5-Coder-1.5B, quantized to Q4_K_M, for use as the frozen base model of the Code2LoRA terminal tool (pip install code2lora) via the gguf/llama.cpp backend.

The base (not Instruct) weights are required so the repo-conditioned LoRA adapters injected by Code2LoRA match the model they were generated for.

File: qwen2.5-coder-1.5b-q4_k_m.gguf (~986 MB).