3178cdf575
- Add MLflow, WandB, Streamlit, Dash, Panel, Bokeh to environment.yml - Update security policy to allow network access for ML tools - Modify secure_runner.py to check tool permissions - Add test script and usage guide - Enable localhost network access for dashboard tools
1.3 KiB
1.3 KiB
ML Tools Integration Guide
Data scientists can now use their preferred ML tools securely!
Available Tools
- MLflow - Experiment tracking and model registry
- Weights & Biases - Experiment tracking and visualization
- Streamlit - Interactive web apps
- Dash - Plotly-based web dashboards
- Panel - Data apps and dashboards
- Bokeh - Interactive visualizations
Quick Start
1. Test Tools Available
cd podman
python test_ml_tools.py
2. Use MLflow
import mlflow
# Start tracking
with mlflow.start_run():
mlflow.log_param("epochs", 10)
mlflow.log_metric("accuracy", 0.95)
3. Launch Streamlit App
# In your secure container
streamlit run my_app.py --server.port 8501
4. Use Dash Dashboard
import dash
import dash_core_components as dcc
import dash_html_components as html
app = dash.Dash(__name__)
app.run_server(debug=True, host='0.0.0.0', port=8050)
Security Features
- Network access limited to localhost only
- Tools pre-approved in security policy
- Container isolation maintained
- No external internet access
Custom Requirements
Add your own tools via requirements.txt:
mlflow==2.7.0
wandb==0.16.0
streamlit==1.28.0
Access URLs
- Streamlit: http://localhost:8501
- Dash: http://localhost:8050
- MLflow UI: http://localhost:5000