Packages

vLLM for Elixir via SnakeBridge - Easy, fast, and cheap LLM serving for everyone. High-throughput LLM inference with PagedAttention, continuous batching, and OpenAI-compatible API.

Current section

Files

Jump to
vllm lib snakebridge_generated vllm config device_config.ex
Raw

lib/snakebridge_generated/vllm/config/device_config.ex

# Generated by SnakeBridge v0.16.0 - DO NOT EDIT MANUALLY
# Regenerate with: mix compile
# Library: vllm 0.14.0
# Python module: vllm.config
# Python class: DeviceConfig
defmodule Vllm.Config.DeviceConfig do
@moduledoc """
Configuration for the device to use for vLLM execution.
"""
def __snakebridge_python_name__, do: "vllm.config"
def __snakebridge_python_class__, do: "DeviceConfig"
def __snakebridge_library__, do: "vllm"
@opaque t :: SnakeBridge.Ref.t()
@doc """
Constructs `DeviceConfig`.
## Parameters
- `dataclass_self__` (term())
- `args` (term())
- `kwargs` (term())
"""
@spec new(term(), term(), term(), keyword()) ::
{:ok, SnakeBridge.Ref.t()} | {:error, Snakepit.Error.t()}
def new(dataclass_self__, args, kwargs, opts \\ []) do
SnakeBridge.Runtime.call_class(__MODULE__, :__init__, [dataclass_self__, args, kwargs], opts)
end
@doc """
WARNING: Whenever a new field is added to this config,
ensure that it is included in the factors list if
it affects the computation graph.
Provide a hash that uniquely identifies all the configs
that affect the structure of the computation
graph from input ids/embeddings to the final hidden states,
excluding anything before input ids/embeddings and after
the final hidden states.
## Returns
- `String.t()`
"""
@spec compute_hash(SnakeBridge.Ref.t(), keyword()) ::
{:ok, String.t()} | {:error, Snakepit.Error.t()}
def compute_hash(ref, opts \\ []) do
SnakeBridge.Runtime.call_method(ref, :compute_hash, [], opts)
end
@spec device(SnakeBridge.Ref.t()) :: {:ok, term()} | {:error, Snakepit.Error.t()}
def device(ref) do
SnakeBridge.Runtime.get_attr(ref, :device)
end
end