Update Dockerfile

Dockerfile

@@ -1,25 +1,40 @@
-FROM python:3.11
+# This Dockerfile uses your original, fast-building pattern.
 
+FROM python:3.11-slim
+
+# Install system dependencies
 RUN apt-get update && apt-get install -y \
-    curl build-essential cmake \
+    git curl build-essential cmake \
     && rm -rf /var/lib/apt/lists/*
 
 WORKDIR /app
 
+# Create writable directories
+RUN mkdir -p /app/.cache /app/vector_database && chmod -R 777 /app
+
+# Set environment variables
 ENV TRANSFORMERS_CACHE=/app/.cache \
     HF_HOME=/app/.cache \
-    CHROMADB_DISABLE_TELEMETRY=true \
-    CMAKE_ARGS="-DLLAMA_CUBLAS=OFF"
+    CHROMADB_DISABLE_TELEMETRY=true
+
+# Pre-install the specific, known-working version of llama-cpp-python
+# This version matches the tool version used to create your new GGUF file.
+RUN pip install --no-cache-dir llama-cpp-python==0.2.61
 
+# Install the rest of the dependencies
 COPY requirements.txt .
 RUN pip install --no-cache-dir -r requirements.txt
 
-COPY processed_chunks.json .
-COPY app ./app
+# Copy the application code
+COPY . .
 
-# Ensure the model directory exists with writable permissions
-RUN mkdir -p /app/models
+# <-- UPDATED: Download your new, compatible GGUF model from the new repo
+RUN curl -fL -o /app/phi1.5_dop_q4_k_m.gguf \
+    https://huggingface.co/Kalpokoch/Phi1.5QuantizedFineTuned/resolve/main/phi1.5_dop_q4_k_m.gguf \
+    && echo "✅ Model downloaded."
 
+# Expose the application port
 EXPOSE 7860
 
-CMD ["uvicorn", "app.app:app", "--host", "0.0.0.0", "--port", "7860"]
+# Run the FastAPI application
+CMD ["uvicorn", "app.app:app", "--host", "0.0.0.0", "--port", "7860"]