We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
using callback to early stop the traning if accuracy reach certain value