Update dockerfile

dockerfile

@@ -1,36 +1,43 @@
-# Use a base image from Microsoft that includes Conda, Python, and GPU drivers.
-# This image is a great starting point for ML workloads on Azure.
-FROM mcr.microsoft.com/azureml/openmpi4.1.0-ubuntu20.04:latest
+# Use an official NVIDIA CUDA base image for GPU support
+FROM nvidia/cuda:12.1.0-runtime-ubuntu22.04
 
-# Set the working directory inside the container. All subsequent commands
-# will be executed from this directory.
+# Set up Python and other system dependencies
+RUN apt-get update && apt-get install -y \
+    python3 \
+    python3-pip \
+    git \
+    wget \
+    libgl1-mesa-glx \
+    libglib2.0-0 && \
+    rm -rf /var/lib/apt/lists/*
+
+# Set the working directory
 WORKDIR /app
 
-# Copy the Conda environment file into the container.
+# Install Conda and create the environment
 COPY environment.yml .
+RUN wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh -O miniconda.sh && \
+    bash miniconda.sh -b -p /opt/conda && \
+    rm miniconda.sh
+ENV PATH="/opt/conda/bin:$PATH"
 
-# Create the Conda environment using the provided YAML file.
-# The name of the environment will be `yolo-onnx-cpu-env` as defined in the file.
+# Create the Conda environment
 RUN conda env create -f environment.yml
 
-# We need to install `gunicorn` and `Flask` or a similar web server to serve the model.
-# Azure Container Apps uses HTTP to trigger scaling.
-# We'll install these into the new Conda environment.
+# Activate the environment and install GPU-specific dependencies
 SHELL ["conda", "run", "-n", "yolo-onnx-cpu-env", "/bin/bash", "-c"]
-RUN pip install gunicorn flask
 
-# Copy the scoring script, the ONNX model file, and any other necessary files
-# into the container's working directory.
+# Replace onnxruntime with the GPU version
+RUN pip uninstall -y onnxruntime && pip install onnxruntime-gpu==1.22.1 gunicorn flask
+
+# Copy your application files
 COPY scoring_Yolo_Model.py .
 COPY best.onnx .
-# Assuming `class_names` or other static files are also present, copy them here.
-# COPY class_names.txt .
+COPY scoring_Yolo_Model_Gunicorn.py . # This is the file you provided
 
-# Expose the port that the web server will listen on.
-# Azure Container Apps will route traffic to this port.
-EXPOSE 8080
+# Expose the default port for Hugging Face Spaces Docker
+EXPOSE 7860
 
-# The CMD instruction defines the command to run when the container starts.
-# We use Gunicorn to serve our Flask app, which will be defined in the scoring script.
-# The `conda run` command ensures the script is executed within the correct Conda environment.
-CMD ["conda", "run", "-n", "yolo-onnx-cpu-env", "gunicorn", "--bind", "0.0.0.0:8080", "scoring_Yolo_Model_Gunicorn:app"]
+# The CMD to run the Gunicorn server
+# Note: Hugging Face Spaces uses port 7860 by default for Docker.
+CMD ["conda", "run", "-n", "yolo-onnx-cpu-env", "gunicorn", "--bind", "0.0.0.0:7860", "scoring_Yolo_Model_Gunicorn:app"]