{
  "schema_version": "sensebench-run-v2",
  "run_id": "vllm-glm-4.6-awq-b300-p001-lexen-v1-20260614",
  "created_at": "2026-06-14T14:17:06.058829+00:00",
  "git_commit": null,
  "runner": {
    "github_handle": "vassiliphilippov",
    "name": "Vassili Philippov",
    "contact": null
  },
  "dataset": {
    "dataset_id": "lexen",
    "dataset_version": "lexen-v1",
    "dataset_revision": null,
    "content_hash": "sha256:5fd4382b93f19087a1e31f6dd7d1db17c1eb17ff80fcbe1d3fdd55c0c3ecefe8",
    "item_count": 4861
  },
  "prompt": {
    "id": "p001",
    "sensebench_version": "0.1.0"
  },
  "model": {
    "kind": "self_hosted_llm",
    "display_name": "zai-org/GLM-4.6",
    "requested_model": "hosted_vllm/zai-org/GLM-4.6",
    "resolved_model": "zai-org/GLM-4.6",
    "resolved_model_counts": {
      "zai-org/GLM-4.6": 4862
    },
    "llm_vendor": "Z.ai",
    "source_kind": "open_source",
    "license": "MIT",
    "model_url": "https://huggingface.co/zai-org/GLM-4.6",
    "hf_revision": "ff8307ccba59f7b4b5bac498f4b469cf5f85f3e5",
    "quantization": "awq-int4",
    "inference_engine": "vllm",
    "inference_engine_version": "0.22.1",
    "container_image": "vllm/vllm-openai:v0.22.1",
    "serve_command": "exec vllm serve bullpoint/GLM-4.6-AWQ --port 8000 --served-model-name zai-org/GLM-4.6 --max-model-len 8192 --gpu-memory-utilization 0.90 --trust-remote-code --max-num-seqs 32 --default-chat-template-kwargs \\{\\\"enable_thinking\\\":\\ false\\} --revision ff8307ccba59f7b4b5bac498f4b469cf5f85f3e5",
    "endpoint_base_url": "http://localhost:8000/v1"
  },
  "sampling": {
    "temperature": 0.0,
    "top_p": null,
    "max_tokens": 2048,
    "seed": null,
    "extra": {}
  },
  "policy": {
    "votes_per_item": 1,
    "semantic_reasks_per_invalid_vote": 1,
    "tie_break": "earliest_vote",
    "monosemous_policy": "short_circuit"
  },
  "machine": {
    "platform": "Linux-6.8.0-117-generic-x86_64-with-glibc2.35",
    "cpu_model": "Intel(R) Xeon(R) 6776P",
    "cpu_cores": 256,
    "ram_gib": 3023.5,
    "gpu": {
      "name": "NVIDIA B300 SXM6 AC",
      "count": 1,
      "vram_mib_per_gpu": 275040,
      "driver_version": "590.48.01",
      "cuda_version": "13.1"
    },
    "provider": "vast.ai",
    "instance_id": "40919269",
    "hourly_rate_usd": 6.71875
  },
  "execution": {
    "concurrency": 64,
    "warmup_call_count": 8,
    "timing": {
      "benchmark_started_at": "2026-06-14T14:07:28.329130+00:00",
      "benchmark_ended_at": "2026-06-14T14:17:06.057329+00:00",
      "benchmark_seconds": 577.7281932439655,
      "setup_seconds": 2.852535172831267
    }
  },
  "totals": {
    "item_count": 4861,
    "correct_count": 4312,
    "accuracy": 0.8870602756634437,
    "call_count": 4862,
    "usage": {
      "input_tokens": 2989961,
      "cached_input_tokens": null,
      "output_tokens": 43758,
      "reasoning_output_tokens": null
    },
    "cost": {
      "currency": "USD",
      "total_usd": 1.0782253606549703,
      "input_uncached_usd": null,
      "input_cached_usd": null,
      "output_usd": null,
      "input_uncached_unit_price_usd": null,
      "input_cached_unit_price_usd": null,
      "output_unit_price_usd": null,
      "source": "machine_time_estimate"
    },
    "elapsed_seconds": 577.7281932439655
  }
}