sesterce-cli/src/modules/ai-inference/domain/inference-instance.ts

A powerful command-line interface tool for managing Sesterce Cloud services. Sesterce CLI provides easy access to GPU cloud instances, AI inference services, container registries, and SSH key management directly from your terminal.

export type InferenceInstanceStatus =
  | "NEW"
  | "ACTIVE"
  | "PENDING"
  | "ERROR"
  | "DISABLED"
  | "DELETED"
  | "DELETING"
  | "DEPLOYING"
  | "FAILED"
  | "PARTIALLYDEPLOYED";

export type InferenceInstanceFeature = "chat" | "image";

export interface InferenceInstance {
  _id: string;
  name: string;
  status: InferenceInstanceStatus;
  hourlyPrice: number;
  features: InferenceInstanceFeature[];
  address: string;
  createdAt: string;
  updatedAt: string;
}

export interface InferenceInstanceContainer {
  regionId: number;
  scale: {
    min: number;
    max: number;
    cooldownPeriod?: number;
    triggers?: {
      cpu?: {
        threshold: number;
      };
      gpuMemory?: {
        threshold: number;
      };
      gpuUtilization?: {
        threshold: number;
      };
      memory?: {
        threshold: number;
      };
      http?: {
        rate: number;
        window: number;
      };
    };
  };
  deployStatus: {
    total: number;
    ready: number;
  };
  errorMessage?: string;
}

export interface InferenceInstanceDetails extends InferenceInstance {
  containerPort: number;
  description?: string;
  podLifetime?: number;
  containers: InferenceInstanceContainer[];
  hardwareName: string;
  envs: Record<string, string>;
  startupCommand?: string;
}