quantize_result.json

{
  "method": "nvfp4",
  "model": "deepreinforce-ai/Ornith-1.0-9B",
  "output_dir": "/mnt/models/quantized/Ornith-1.0-9B-NVFP4",
  "calib_dataset": "ultrachat_200k",
  "calib_samples": 512,
  "load_time_s": 5.5,
  "quant_time_s": 634.2,
  "total_time_s": 656.1,
  "output_size_gb": 10.4,
  "vram_after_load": {
    "gpu_0": "17.5GB allocated, 17.5GB reserved"
  }
}