RoboBrain/processor_config.json

{
  "image_token": "<image>",
  "num_image_tokens": 729,
  "processor_class": "LlavaOnevisionProcessor",
  "video_token": "<video>",
  "vision_feature_select_strategy": "full"
}