/** * @octomil/browser — ModelRuntime interface * * Abstraction over inference runtimes. The default `InferenceEngine` uses * ONNX Runtime Web, but callers may inject a custom runtime (e.g. * TensorFlow.js, WebNN) via the OctomilClient constructor. */ import type { Backend, NamedTensors, PredictOutput } from "../../types.js"; export interface ModelRuntime { /** Create a runtime session from a model buffer. */ createSession(modelData: ArrayBuffer, backend?: Backend): Promise; /** Run inference on the given inputs. */ run(inputs: NamedTensors): Promise; /** Release all resources held by the runtime. */ dispose(): void; /** Whether this runtime is available in the current environment. */ isAvailable(): Promise; } //# sourceMappingURL=model-runtime.d.ts.map