BaseServerConfig
================

**class** ``BaseServerConfig`` **(** ``BasePolyConfig`` **)**

    Base configuration for server launch and management.

**Polymorphic Type:**

    All ``BaseServerConfig`` types:

    - ``vllm``: :doc:`VllmServerConfig`
    - ``vajra``: :doc:`VajraServerConfig`
    - ``sglang``: :doc:`SglangServerConfig`

**Fields:**

``env_path`` : *Optional* [ *str* ] = ``None``
    Path to a Python environment directory (virtualenv/conda).

``model`` : *str* = ``"meta-llama/Meta-Llama-3-8B-Instruct"``
    Model name or path.

``host`` : *str* = ``"localhost"``
    Host address for the server

``port`` : *int* = ``8000``
    Port number for the server

``api_key`` : *str* = ``"token-abc123"``
    API key for server authentication

``gpu_ids`` : *Optional* [ *list* [ *int* ] ] = ``None``
    List of GPU IDs to use (None means auto-assign)

``startup_timeout`` : *int* = ``300``
    Timeout in seconds for server startup

``health_check_interval`` : *float* = ``2.0``
    Interval in seconds between health checks

``require_contiguous_gpus`` : *bool* = ``True``
    Require contiguous GPU allocation (e.g., GPUs 0,1,2 instead of 0,2,5)

``tensor_parallel_size`` : *int* = ``1``
    Number of GPUs for tensor parallelism

``dtype`` : *str* = ``"auto"``
    Data type for model weights (auto, float16, bfloat16, etc.)

``max_model_len`` : *Optional* [ *int* ] = ``None``
    Maximum model context length

``additional_args`` : *Optional* [ *str* ] = ``"{}"``
    Additional engine-specific arguments as JSON string, dict, or None.