> ## Documentation Index
> Fetch the complete documentation index at: https://docs.domino.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Set up model quality monitoring

Ground truth data consists of known correct outputs for each input. Model quality monitoring compares the model’s predicted values against the actual results (or labels for the predictions) using ground truth data to generate quality metrics.

After you set up your model to [capture prediction data](/cloud/platform-capabilities/features/monitoring/3-set-up-domino-endpoints/1-set-up-prediction-capture#capture-prediction-data), you can ingest the model’s ground truth data to monitor the quality of predictions made by the model. For this analysis, you must have declared a `row_identifier` column for the prediction data schema. This `row_identifier` column must also be present in the ground truth data to map a prediction uniquely to the ground truth.

When you apply a date filter, Domino uses the timestamp values in the prediction data to filter the results. Then, it matches the filtered predictions with the ground truth labels ingested in the last 90 days and calculates the model quality metrics for the matched predictions.

<Note>
  You can register ground truth data for a model that uses a Snowflake data source once per table.
</Note>

If you registered a `timestamp` variable or if you register it with the ground truth dataset, Domino will automatically retrieve new data every 24 hours at 12:00 AM UTC. See [Monitoring Config JSON](/cloud/platform-capabilities/features/monitoring/8-json). When you deregister the model, Domino stops retrieving the data.

If you don’t register the timestamp variable, the ground truth data is ingested once, and for every new batch, you must register a new dataset.

## Ingest ground truth data

You can use the Domino application or a REST API to ingest ground truth data. See [Model Monitoring APIs](/cloud/platform-capabilities/features/monitoring/9-model-monitoring-apis) for details.

1. In the navigation pane, click **Model Monitor**.

2. Click the model for which you want to ingest ground truth data.

3. Click **Model Quality**.

   <Note>
     If you want to do this from the Overview page, go to **Add Data > Ground Truth Data**.
   </Note>

4. On the Model Quality page of the registered model, click **Register Ground Truth > Upload Ground Truth Config**.

   In the Register Ground Truth window, you can upload a Ground Truth Config file or paste a JSON config. This JSON must capture all the information needed to register ground truth data.

   The following is a sample file. You must specify the target column, shown below as `y_gt` and the row identifier that maps to a corresponding prediction record, shown below as `gt_uuid`. Note that your prediction dataset needs to have a column that contains unique row identifiers that match. When you registered your prediction dataset with Domino, you also supplied a variable that denotes the `row_identifier` similarly. The configuration must always include the `datasetDetails` and `modelMetadata` information. See [Monitoring Config JSON](/cloud/platform-capabilities/features/monitoring/8-json) for details about each field in the JSON.

   ```console theme={null}
   {
   "variables": [
     {
       "name": "y_gt",
       "variableType": "ground_truth",
       "valueType": "categorical",
       "forPredictionOutput": "y"
     },
     {
        "name": "gt_uuid",
        "variableType": "row_identifier",
        "valueType": "string"
     }
   ],
   "datasetDetails": {
     "name": "GT-labels.csv",
     "datasetType": "file",
     "datasetConfig": {
       "path": "GT-labels.csv",
       "fileFormat": "csv"
     },
     "datasourceName": "abc-shared-bucket",
     "datasourceType": "s3"
   }
   }
   ```

5. Click **Register Ground Truth**. On the Model Quality Analyze page, the Model Monitor starts to calculate model quality metrics for this dataset.

6. Click the bell icon to disable the alerts. Doing this reduces noise by controlling the metrics included in the Scheduled Checks.

### Check the status of ground truth data

After you register the ground truth data for ingestion, you might want to check its status.

1. From the navigation pane, click **Model Monitor**.

2. Click the name of the model for which you set up ground truth data ingestion.

3. On the Overview page’s **Data Traffic** tab, you can see the history of the ingested data and review the model’s metadata and schema.

4. Click **Ingest History** to check to see if the status is `Done`. The following are other statuses you might see:

   **`Completed with Warnings`**\
   Click the three vertical dots at the end of the row and click **View Details**.

   **`Failed`**\
   Click the three vertical dots at the end of the row and click **View Details**.

   If the ingestion failed, Domino stops the continuous ingestion. After you fix the issue, re-register the prediction data and Domino restarts continuous ingestion from this point.

   <Note>
     If you re-register the same dataset, Domino processes the entire table again. This causes duplicate records which affects the drift and model quality calculations.
   </Note>

   **`Skipped`**\
   The dataset had no rows to fetch.

   You can click **Status** to select checkboxes to filter by status.

5. Click **Refresh** to see a graph of the **Data Ingest Timeline**.

   After the data is ingested, you can perform an ad-hoc [model quality analysis](/cloud/platform-capabilities/features/monitoring/5-use-monitoring/3-analyze-data-drift) to validate your setup. Then, [set up a schedule](/cloud/platform-capabilities/features/monitoring/4-set-up-model-monitor/8-set-scheduled-checks-mm) to perform model quality analysis automatically.