Context size for images

#4
by TomSchelsen - opened

As this is a multimodal model : default context size is 256K, I'd like to reduce it, how does it affect the capability to take images as input ? Is it image resolution dependant ? Thanks

Mistral AI_ org

it doesn't affect performance per say but reduces the maximum amount of data you can pass to the model for the context which can be limiting for some tasks.

juliendenize changed discussion status to closed

Sign up or log in to comment