quantization_info.json

{
  "model_id": "LlamaFactory_new/saves/qwen35-9b/merged/035m",
  "output_dir": "LlamaFactory_new/saves/qwen35-9b/int8/035m",
  "tool": "llm-compressor",
  "method": "int8",
  "scheme": "W8A8",
  "calib_data": "LlamaFactory_new/data/industry_classification_035m.json",
  "calib_samples": 2048,
  "max_seq_length": 512,
  "elapsed_load_s": 2.0,
  "elapsed_quant_s": 10.1,
  "elapsed_save_s": 13.5,
  "total_elapsed_s": 51.3
}