What makes a foundation model multimodal?
Why B is the answer
It works with more than one type of data, such as text and images
Multimodal models handle several data types.
In detail
- Multimodal means more than one type of data.
- Text, images, audio and video are all modalities.
Use case
An illustrative scenario, not a real incident.
A model reads a photo of a damaged car and the claim text together, which makes it multimodal.
See it in different ways
mindmap
root((Google models))
Gemini
Multimodal
Gemma
Lightweight and open
Imagen
Images
Veo
Videomindmap
root(("Multimodal model"))
n1["Inputs"]
n2["Text"]
n3["Images"]
n4["Audio"]
n5["Video"]Key point: Multimodal means several data types.
Why the other options are wrong
- A. It supports more than one payment option for its customers
- Payment is unrelated.
- C. It runs in more than one Google Cloud region at the same time
- Regions are about deployment.
- D. It has been trained by more than one team inside the company
- Teams are about who built it.