build: use custom Docker base image from Docker Hub

- Replace python:3.12 with crazymonkey00/llama-base:latest
- Remove llama-cpp-python installation (already in base image)
- Reduce build time from 40+ minutes to ~30 seconds

Dockerfile

@@ -1,4 +1,4 @@
-FROM python:3.12-slim-bookworm
+FROM crazymonkey00/llama-base:latest
 
 WORKDIR /app
 
@@ -37,10 +37,6 @@ RUN pip install --upgrade pip setuptools wheel
 # Install dependencies from requirements
 RUN pip install --no-cache-dir -r requirements.txt
 
-# Build and install llama-cpp-python with OpenBLAS
-ENV CMAKE_ARGS="-DGGML_BLAS=ON -DGGML_BLAS_VENDOR=OpenBLAS"
-RUN pip install --no-cache-dir llama-cpp-python
-
 # Copy the rest of the application
 COPY . /app