model: change to official qwen2.5-1.5b-instruct-q4_k_m.gguf

Dockerfile

@@ -13,8 +13,8 @@ WORKDIR /app/llama.cpp
 
 # Hardcode for now
 ARG LLAMA_CPP_RELEASE_URL="https://github.com/ggml-org/llama.cpp/releases/download/b6218/llama-b6218-bin-ubuntu-x64.zip"
-ARG MODEL_URL="https://huggingface.co/mradermacher/Qwen2.5-1.5B-Instruct-abliterated-i1-GGUF/resolve/main/Qwen2.5-1.5B-Instruct-abliterated.i1-Q4_K_M.gguf"
-ARG MODEL_FILENAME="Qwen2.5-1.5B-Instruct-abliterated.i1-Q4_K_M.gguf"
+ARG MODEL_URL="https://huggingface.co/Qwen/Qwen2.5-1.5B-Instruct-GGUF/resolve/main/qwen2.5-1.5b-instruct-q4_k_m.gguf"
+ARG MODEL_FILENAME="qwen2.5-1.5b-instruct-q4_k_m.gguf"
 
 
 RUN curl -L "${LLAMA_CPP_RELEASE_URL}" -o llama.zip && \