DB-BERT Demo (SIGMOD'22)

preview_player
Показать описание
I demonstrate DB-BERT: a tool that mines tuning hints from text documents. DB-BERT extracts recommendations for database configuration parameters from text documents such as manuals or forum discussions. Next, using those hints as starting points, it uses reinforcement learning to find database configurations that optimize a user-defined performance metric.
Рекомендации по теме