medgemma_lora_adapter/processor_config.json

{
  "image_seq_length": 256,
  "processor_class": "Gemma3Processor"
}