final changes

Dockerfile

@@ -1,26 +1,26 @@
 # Use a lightweight Python image
 FROM python:3.10-slim
 
+# Set the working directory
+WORKDIR /app
+
 # 1. Install System Dependencies (Compilers + PDF tools)
-# - build-essential: contains gcc, g++, make (REQUIRED for llama-cpp-python)
-# - cmake: required to build the build system
-# - poppler-utils: required for pdf2image
-# - libgl1: required for opencv/vision tasks
+# 'build-essential' and 'cmake' are REQUIRED for llama-cpp-python
+# 'poppler-utils' is for PDF conversion
+# 'libgl1' is for vision tasks
 RUN apt-get update && apt-get install -y \
     build-essential \
     cmake \
-    g++ \
-    gcc \
     poppler-utils \
     libgl1 \
     libglib2.0-0 \
     && rm -rf /var/lib/apt/lists/*
 
-# 2. Upgrade pip (good practice)
+# 2. Upgrade pip
 RUN pip install --upgrade pip
 
 # 3. Install Python Libraries
-# We force a reinstall with --no-cache-dir to avoid using broken cached wheels
+# We use CMAKE_ARGS to enable optimizations (makes it faster on CPU)
 RUN CMAKE_ARGS="-DGGML_BLAS=ON -DGGML_BLAS_VENDOR=OpenBLAS" pip install --no-cache-dir \
     "llama-cpp-python[server]" \
     huggingface_hub \
@@ -30,13 +30,12 @@ RUN CMAKE_ARGS="-DGGML_BLAS=ON -DGGML_BLAS_VENDOR=OpenBLAS" pip install --no-cac
     fastapi \
     pillow
 
-# 4. Download the Model
-# Using the GGUF (CPU-optimized) version
+# 4. Download the Model files
+# This saves the model inside the Docker image so it doesn't download on every restart
 RUN huggingface-cli download Qwen/Qwen2-VL-7B-Instruct-GGUF qwen2-vl-7b-instruct-q4_k_m.gguf --local-dir /app/model
 RUN huggingface-cli download Qwen/Qwen2-VL-7B-Instruct-GGUF mmproj-model-f16.gguf --local-dir /app/model
 
 # 5. Copy your API code
-WORKDIR /app
 COPY app.py .
 
 # 6. Run the API Server