serving-tests-cpu-embed.json

{
  "defaults": {
    "qps_list": [
      "inf"
    ],
    "max_concurrency_list": [
      32,
      64,
      128
    ],
    "server_environment_variables": {
      "VLLM_RPC_TIMEOUT": 100000,
      "VLLM_ALLOW_LONG_MAX_MODEL_LEN": 1,
      "VLLM_ENGINE_ITERATION_TIMEOUT_S": 120,
      "VLLM_CPU_SGL_KERNEL": 1,
      "VLLM_CPU_KVCACHE_SPACE": 40
    },
    "server_parameters": {
      "dtype": "bfloat16",
      "model": "jinaai/jina-embeddings-v3",
      "trust_remote_code": ""
    },
    "client_parameters": {
      "model": "jinaai/jina-embeddings-v3",
      "backend": "openai-embeddings",
      "endpoint": "/v1/embeddings",
      "dataset_name": "sharegpt",
      "dataset_path": "ShareGPT_V3_unfiltered_cleaned_split.json",
      "num_prompts": 200
    }
  },
  "tests": [
    {
      "test_name": "serving_jina_embed_v3_tp1_sharegpt",
      "server_parameters": {
        "tensor_parallel_size": 1
      },
      "client_parameters": {}
    }
  ]
}