Watch this video on YouTube
An AI practitioner wants to monitor the input and output data of an Amazon Bedrock base model used to summarize customer service chats. What should the practitioner do to enable invocation logging?