quantization_info.json

{
  "model_name": "luca-deandrea/MNLP_M3_mcqa_model",
  "quantization_method": "bitsandbytes",
  "config_name": "bnb_8bit_aggressive",
  "description": "8-bit with aggressive quantization",
  "expected_compression": "~4x",
  "quantization_config": {
    "load_in_4bit": false,
    "load_in_8bit": true,
    "bnb_4bit_quant_type": "fp4",
    "bnb_4bit_use_double_quant": false,
    "bnb_4bit_compute_dtype": "torch.float32",
    "llm_int8_threshold": 0.0
  },
  "timestamp": "2025-06-10 21:32:18",
  "loading_time_seconds": 10.248082876205444,
  "memory_usage_gb": 0.7341856956481934,
  "test_generation": "Hello, I am a student who has been working on a project to"
}