The documentation page MODEL_DOC/VISION_TEXT_DUAL_ENCODER doesn’t exist in v4.39.3, but exists on the main version. Click here to redirect to the main version of the documentation.