{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/b46b2eba084b4fd58bb6fa1b32c713cd\" frameborder=\"0\" width=\"1662\" height=\"1246\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1246,"width":1662,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1246,"thumbnail_width":1662,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/b46b2eba084b4fd58bb6fa1b32c713cd-b255c3a2c1c77bb1.gif","duration":130.608,"title":"Building Continuous Learning for QM","description":"This Loom explains a continued learning plugin for QM that turns user interactions into training data in real time. It describes an API setup for collecting data during conversations, including a student teacher mode for SFT or policy distillation and a readout mode that uses A B testing for RL style training. The demo also covers a proactive teammate model that initiates help without being asked, gathering feedback via thumbs up or stop signals, and includes a smaller pipeline model to decide when to be proactive."}