Model training requires a large amount of labeled data. Therefore, before the model training, label the unlabeled audio files. ModelArts enables you to label audio files in batches by one click. In addition, you can modify the labels of audio files, or remove their labels and label the audio files again.
ModelArts automatically synchronizes data and labeling information from datasets to the labeling job.
To quickly obtain the latest data in the datasets, in the Unlabeled or Labeled tab of the labeling job details page, click Synchronize New Data.
The labeling job details page displays the Unlabeled and Labeled tabs. The Unlabeled tab is displayed by default. Click on the left of the audio to preview the audio.
Method 1 (the required label already exists): In the right pane, select a shortcut from the Shortcut drop-down list, select an existing label name from the Label text box, and click OK.
Method 2 (adding a label): In the right pane, select a shortcut from the Shortcut drop-down list, and enter a new label name in the Label text box.
Shortcut key description: After specifying a shortcut key for a label, you can select an audio file and press the shortcut key to add a label for the audio file. Example: Specify 1 as the shortcut key for the aa label. Select one or more files and press 1. A message is displayed, asking you whether to label the files with aa. Click OK.
Each label has a shortcut key. A shortcut key cannot be specified for different labels. Shortcut keys can greatly improve the labeling efficiency.
On the labeling job details page, click the Labeled tab to view the list of labeled audio files. Click an audio file. You can view the label information about the audio file in the File Labels area on the right.
After labeling data, you can modify labeled data in the Labeled tab.
On the labeling job details page, click the Labeled tab. Select one or more audio files to be modified from the audio list. Modify the label in the label details area on the right.
On the labeling job details page, click the Labeled tab. The information about all labels is displayed on the right.
In addition to synchronizing data, you can directly add data on labeling job details page for labeling.
For details about how to import data, see section "Importing Data".
You can quickly delete the audio files you want to discard.
In the Unlabeled or Labeled tab, select the audio files to be deleted one by one or tick Select Current Page to select all audio files on the page, and then click Delete File in the upper left corner. In the displayed dialog box, select or deselect Delete the source files from OBS as required. After confirmation, click OK to delete the audio files.
If a tick is displayed in the upper right corner of an audio file, the audio file is selected. If no audio file is selected on the page, the Delete File button is unavailable.
If you select Delete the source files from OBS, audio files stored in the corresponding OBS directory will be deleted when you delete the selected audio files. Deleting source files may affect other dataset versions or datasets using those files. As a result, the page display, training, or inference is abnormal. Deleted data cannot be recovered. Exercise caution when performing this operation.