{
  "artifacts": {
    "Kimi-K3-0.40B-F16.gguf": {
      "sha256": "411c197b503e6fb9199a2b22115e32dc4e2cad803fb112b24967737b3bab26c7",
      "size_bytes": 784318432
    },
    "Kimi-K3-0.40B-MXFP4.gguf": {
      "sha256": "0379a1cc623e09eb3fbd1dfcb18737bc8c971dbfe5bf5bc3e08da8b5379ec169",
      "size_bytes": 751976576
    }
  },
  "conversion": {
    "expected_resident_tensor_count": 35,
    "expected_routed_expert_group_count": 21,
    "resident_mxfp4_tensors_dequantized_to_f16": 35,
    "routed_expert_mxfp4_groups": 21
  },
  "converter": {
    "branch": "k3/out-of-core",
    "commit": "84245db4c790af22135f34992689edcc11877003",
    "repository": "murillo128/llama.cpp"
  },
  "environment": {
    "host": "skynet",
    "packages": {
      "huggingface-hub": "0.36.2",
      "safetensors": "0.8.0",
      "tiktoken": "0.13.0",
      "torch": "2.11.0+cpu",
      "transformers": "4.57.6"
    },
    "platform": "Linux-6.8.0-136-generic-x86_64-with-glibc2.39",
    "python": "3.12.3"
  },
  "generated_at": "2026-07-29T11:10:46.285957+00:00",
  "purpose": "Kimi-K3 tiny architecture and llama.cpp validation fixtures",
  "schema_version": 1,
  "source_models": {
    "f16_source": {
      "repo_id": "inference-optimization/Kimi-K3-0.40B",
      "revision": "d853649387ffe8f48ce0198a29ac1a44205031f7"
    },
    "mxfp4_source": {
      "repo_id": "inference-optimization/Kimi-K3-0.40B-MXFP4",
      "revision": "ef3902c318fb8e13c3507e26055656e687fdfe38"
    }
  }
}