This model is derived from Qwen/Qwen3-8B and is released under the Apache 2.0 license. Use and redistribution must comply with the upstream model license and Hugging Face model terms.

Run Metadata

json

{
  "base_model": "Qwen/Qwen3-8B",
  "bits": 2,
  "config": "/nas/longleaf/home/yuanwu/Bias_Compressed_LLM/Quantization/GPTQ/configs/qwen3_8b.yaml",
  "created_at": "2026-06-22T07:36:31.229916+00:00",
  "elapsed_seconds": 564.32,
  "environment": {
    "cuda_available": true,
    "cuda_devices": [
      {
        "capability": "8.9",
        "index": 0,
        "name": "NVIDIA L40S",
        "total_memory_gb": 44.39
      }
    ],
    "platform": "Linux-5.14.0-611.16.1.el9_7.x86_64-x86_64-with-glibc2.34",
    "python": "3.10.20",
    "torch": "2.12.0+cu126",
    "torch_cuda": "12.6"
  },
  "output_dir": "/users/y/u/yuanwu/Bias_Compressed_LLM/gptq_outputs/Qwen3-8B-GPTQ-2bit",
  "quantization_config": {
    "backend": "auto",
    "batch_size": 1,
    "bits": 2,
    "block_name_to_quantize": null,
    "cache_block_outputs": true,
    "dataset": "c4",
    "desc_act": true,
    "group_size": 128,
    "max_input_length": 512,
    "model_seqlen": 512,
    "modules_in_block_to_quantize": null,
    "sym": true,
    "true_sequential": true
  }
}

yw223/Qwen3-8B-GPTQ-2bit

Qwen3-8B GPTQ 2-bit

Quantization

Validation

License and Use

Run Metadata

yw223/
Qwen3-8B-GPTQ-2bit