For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /get-started.md).

Python class

VisionBatchProcessor

`VisionBatchProcessor`

class max.pipelines.lib.module_registration.VisionBatchProcessor(*args, **kwargs)

source

Bases: SupportsVisionEncoding[Any], Protocol

The batch processor of a registration with a VISION module.

It packs and encodes the images, through the vision graph the pipeline model compiles and binds to it.

bind_vision_encoder(*, vision_model, session)

source

Binds the compiled vision graph and the session that runs it.

Parameters:

Return type:

None