quant_info.json

{
  "base_model": "vikhyatk/moondream2",
  "method": "PYTORCH-INT8",
  "bits": 8
}