Question

A retail company wants to build an application where customers can upload a photo of a product and ask questions about it in natural language, receiving text answers that reference what is visible in the image. Which capability of a foundation model must the team prioritize when selecting a model on Amazon Bedrock?

Accepted Answer

Multimodal support that accepts both image and text inputs to generate text responses

Answer

A large context window that can hold thousands of tokens of text

Answer

Fine-tuning support so the model can be trained on the company's product catalog

Answer

Low temperature settings to make the model's answers fully deterministic