{
  "model_type": "HunyuanImage-3.0-Instruct",
  "quantization_method": "bitsandbytes_int8",
  "load_in_8bit": true,
  "llm_int8_threshold": 6.0,
  "expected_vram_gb": 95,
  "expected_total_memory_gb": 100,
  "modules_kept_bf16": [
    "vae",
    "vision_model",
    "vision_aligner",
    "patch_embed",
    "final_layer",
    "time_embed",
    "time_embed_2",
    "timestep_emb",
    "attention_projections"
  ],
  "notes": "Instruct model with vision encoder kept at BF16 for image understanding quality.",
  "attention_layers_quantized": false,
  "quality_vs_nf4": "Significantly better - approximately 2x memory for ~98% quality retention"
}