{"id":53735,"plugin_id":"plugins_6ac012b82680819195f8a26cbe32c246","kind":"skill","collection_source":"skill_api","comparison_source":null,"observed_at":"2026-10-24T13:54:54.694Z","digest":"cb806125aa6617299e37db5225cc99687547b67946fabe118fe3756ebf7a5324","against":null,"payload":{"description":"Track traditional ML experiments in Domino using the MLflow-based Experiment Manager. Covers experiment setup, auto-logging for sklearn/TensorFlow/PyTorch, manual logging, artifact storage, run comparison, and model registration. Use when training ML models, logging metrics and parameters, comparing model runs, or registering models.","included_files":[{"relative_path":"COMPARING-RUNS.md","size_in_bytes":7564},{"relative_path":"MLFLOW-BASICS.md","size_in_bytes":8938},{"relative_path":"MODEL-REGISTRY.md","size_in_bytes":8646}],"name":"domino-experiment-tracking","skill_md_contents":"---\nname: domino-experiment-tracking\ndescription: Track traditional ML experiments in Domino using the MLflow-based Experiment Manager. Covers experiment setup, auto-logging for sklearn/TensorFlow/PyTorch, manual logging, artifact storage, run comparison, and model registration. Use when training ML models, logging metrics and parameters, comparing model runs, or registering models.\n---\n\n# Domino Experiment Tracking Skill\n\nThis skill provides comprehensive knowledge for tracking ML experiments in Domino Data Lab using the built-in MLflow-based Experiment Manager.\n\n## Key Concepts\n\n### Experiment Manager Overview\n\nDomino's Experiment Manager is built on MLflow and provides:\n- Automatic and manual logging of parameters, metrics, and artifacts\n- Run comparison and visualization\n- Model versioning and registry\n- Integration with Domino projects and jobs\n\n### Critical Configuration\n\n**Experiment names must be unique across the entire Domino deployment.** Always append username or project name to ensure uniqueness.\n\n## Related Documentation\n\n- [MLFLOW-BASICS.md](./MLFLOW-BASICS.md) - Auto-logging, manual logging\n- [COMPARING-RUNS.md](./COMPARING-RUNS.md) - Run comparison, export\n- [MODEL-REGISTRY.md](./MODEL-REGISTRY.md) - Model registration & stages\n\n## Quick Start\n\n```python\nimport mlflow\nimport os\n\n# CRITICAL: Experiment names must be unique across Domino deployment\nusername = os.environ.get('DOMINO_STARTING_USERNAME', 'unknown')\nexperiment_name = f\"my-experiment-{username}\"\n\n# Set the experiment\nmlflow.set_experiment(experiment_name)\n\n# Enable auto-logging (easiest approach)\nmlflow.autolog()\n\n# Run training\nwith mlflow.start_run(run_name=\"my-first-run\"):\n    model.fit(X_train, y_train)\n\n    # Optional: manually log additional items\n    mlflow.log_param(\"custom_param\", \"value\")\n    mlflow.log_metric(\"custom_metric\", 0.95)\n```\n\n## Supported Frameworks\n\n| Framework | Auto-log Command |\n|-----------|------------------|\n| Scikit-learn | `mlflow.sklearn.autolog()` |\n| TensorFlow/Keras | `mlflow.tensorflow.autolog()` |\n| PyTorch | `mlflow.pytorch.autolog()` |\n| XGBoost | `mlflow.xgboost.autolog()` |\n| LightGBM | `mlflow.lightgbm.autolog()` |\n| All at once | `mlflow.autolog()` |\n\n## Environment Variables\n\nDomino automatically configures MLflow to use the built-in tracking server. These variables are pre-set:\n\n| Variable | Description |\n|----------|-------------|\n| `MLFLOW_TRACKING_URI` | Domino's MLflow server URL |\n| `DOMINO_STARTING_USERNAME` | User running the experiment |\n| `DOMINO_PROJECT_NAME` | Current project name |\n| `DOMINO_RUN_ID` | Domino job run ID |\n\n## Documentation Links\n\n- Domino Experiment Tracking: https://docs.dominodatalab.com/en/latest/user_guide/da707d/track-and-monitor-experiments/\n- Domino Model Registry: https://docs.dominodatalab.com/en/latest/user_guide/3b6ae5/manage-models-with-model-registry/\n"},"changes":[],"summary":"First saved snapshot. No earlier version is available for comparison.","summary_kind":"deterministic","summary_metadata":{}}