tokenizer_config.json

{
  "model_type": "image_classification",
  "version": "V11_production",
  "preprocessing": "imagenet_normalized"
}