🐳 Docker Configuration: Set up PyTorch GPU Environment

.dockerignore

@@ -0,0 +1,13 @@
+__pycache__
+*.pyc
+.git
+.gitignore
+.env
+*.md
+.pytest_cache
+*.egg-info
+dist
+build
+.venv
+.idea
+.cache

.gitignore

@@ -0,0 +1,2 @@
+.cache
+.gradio/certificate.pem

Dockerfile

@@ -0,0 +1,22 @@
+FROM pytorch/pytorch:2.2.1-cuda12.1-cudnn8-runtime
+
+WORKDIR /app
+
+# Install system dependencies
+RUN apt-get update && apt-get install -y \
+    git \
+    python3-pip \
+    && rm -rf /var/lib/apt/lists/*
+
+# Install Python dependencies
+COPY requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Copy application code
+COPY . .
+
+# Expose port for Gradio
+EXPOSE 7860
+
+# Command to run the application
+CMD ["python", "app.py"]

app.py

@@ -235,5 +235,5 @@ def chat_llama3_8b(message: str,
                 )
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch(share=True)
 

docker-compose.yml

@@ -0,0 +1,25 @@
+version: '3.8'
+
+services:
+  llama-mesh:
+    build: 
+      context: .
+      dockerfile: Dockerfile
+    # runtime: nvidia
+    environment:
+      - NVIDIA_VISIBLE_DEVICES=all
+      - NVIDIA_DRIVER_CAPABILITIES=all
+    ports:
+      - "7878:7860"  # Gradio UI port
+    volumes:
+      - .:/app
+      - .cache/huggingface:/root/.cache/huggingface  # Cache HuggingFace models
+    command: python app.py
+    tty: true
+    deploy:
+      resources:
+        reservations:
+          devices:
+            - driver: nvidia
+              count: all
+              capabilities: [gpu]