Create Dockerfile

Dockerfile

@@ -0,0 +1,35 @@
+# Use the official CUDA image as the base image
+FROM nvidia/cuda:11.8-cudnn8-runtime-ubuntu20.04
+
+# Set the Python version
+ARG PYTHON_VERSION=3.8
+
+# Install system packages
+RUN apt-get update && apt-get install -y \
+ python${PYTHON_VERSION} \
+ python3-pip \
+ # Uncomment and add any additional system packages needed
+ # libgl1-mesa-glx \
+ # libglib2.0-0 \
+ && rm -rf /var/lib/apt/lists/*
+
+# Create a symlink for python3
+RUN ln -s /usr/bin/python${PYTHON_VERSION} /usr/bin/python3
+
+# Install the required Python packages
+RUN python3 -m pip install --upgrade pip && \
+ python3 -m pip install \
+ hf_transfer==0.1.3 \
+ git+https://github.com/vllm-project/vllm.git@main
+
+# Set CUDA_HOME environment variable
+ENV CUDA_HOME=/usr/local/cuda
+
+# Copy the rest of your application code
+COPY . /app
+
+# Set the working directory
+WORKDIR /app
+
+# Define the command to run your predictor
+CMD ["python3", "predict.py"]