quant_meta.json

{
  "model_id": "HuggingFaceTB/SmolVLM-Instruct",
  "family": "smolvlm",
  "method": "torchao_w4a8",
  "bits_weight": 4,
  "bits_activation": 8,
  "group_size": 128,
  "skip_vision": true,
  "load_time_s": 8.8,
  "quant_time_s": 0.1,
  "quant_method": "Int8DynamicActivationInt4Weight (W4A8)"
}