sesterce-cli
Version:
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.
267 lines • 11.7 kB
JavaScript
;
Object.defineProperty(exports, "__esModule", { value: true });
exports.createAIInferenceInstanceLaunchCommand = createAIInferenceInstanceLaunchCommand;
const launch_instance_1 = require("@/modules/ai-inference/use-cases/launch-instance");
const list_inference_hardwares_1 = require("@/modules/ai-inference/use-cases/list-inference-hardwares");
const list_inference_models_1 = require("@/modules/ai-inference/use-cases/list-inference-models");
const list_inference_regions_1 = require("@/modules/ai-inference/use-cases/list-inference-regions");
const preview_instance_pricing_1 = require("@/modules/ai-inference/use-cases/preview-instance-pricing");
const list_registries_1 = require("@/modules/registries/use-cases/list-registries");
const prompts_1 = require("@inquirer/prompts");
const printRegistry = (registry) => {
return `${registry.name} - ${registry.url} - ${registry.username}`;
};
function createAIInferenceInstanceLaunchCommand(aiInferenceInstanceCommand) {
aiInferenceInstanceCommand
.command("launch")
.description("Launch an AI inference instance")
.action(async () => {
var _a, _b, _c, _d, _e;
console.log("Loading offers...");
const [hardwaresResult, regionsResult] = await Promise.all([
list_inference_hardwares_1.listInferenceHardwares.execute(),
list_inference_regions_1.listInferenceRegions.execute(),
]);
if (hardwaresResult.isLeft()) {
console.error((_a = hardwaresResult.value) === null || _a === void 0 ? void 0 : _a.message);
return;
}
if (regionsResult.isLeft()) {
console.error((_b = regionsResult.value) === null || _b === void 0 ? void 0 : _b.message);
return;
}
const hardwares = hardwaresResult.value;
const regions = regionsResult.value;
const answer = await (0, prompts_1.checkbox)({
message: "Deploy public or private model?",
choices: [
{ name: "Public model", value: "public", checked: true },
{ name: "Private model (registry)", value: "private" },
],
});
const modelType = answer[0];
let model = null;
let registry = null;
if (modelType === "public") {
console.log("Loading public models...");
const modelsResult = await list_inference_models_1.listInferenceModels.execute();
if (modelsResult.isLeft()) {
console.error(modelsResult.value.message);
return;
}
const models = modelsResult.value;
const selectedModel = await (0, prompts_1.search)({
message: "Select a model",
source: async (input, { signal }) => {
if (!input)
return models.map((model) => ({
name: model.name,
value: model,
}));
return models
.filter((model) => model.name.toLowerCase().includes(input.toLowerCase()))
.map((model) => ({
name: model.name,
value: model,
}));
},
});
model = selectedModel;
}
else {
console.log("Loading registries...");
const registriesResult = await list_registries_1.listRegistries.execute();
if (registriesResult.isLeft()) {
console.error(registriesResult.value.message);
return;
}
const registries = registriesResult.value;
const selectedRegistry = await (0, prompts_1.search)({
message: "Select a registry",
source: async (input, { signal }) => {
if (!input)
return registries.map((registry) => ({
name: printRegistry(registry),
value: registry,
}));
return registries
.filter((registry) => registry.name.toLowerCase().includes(input.toLowerCase()))
.map((registry) => ({
name: printRegistry(registry),
value: registry,
}));
},
});
registry = selectedRegistry;
}
const containerPort = await (0, prompts_1.number)({
message: "Port to expose the model",
required: true,
default: (_c = model === null || model === void 0 ? void 0 : model.port) !== null && _c !== void 0 ? _c : 80,
});
const availableHardwares = hardwares.map((hardware) => ({
...hardware,
available: regions.some((region) => region.capacities.some((capacity) => capacity.hardwareName === hardware.name && capacity.capacity > 0)),
}));
const hardware = await (0, prompts_1.search)({
message: "Select a hardware",
source: async (input, { signal }) => {
if (!input)
return availableHardwares.map((hardware) => ({
name: hardware.name,
value: hardware,
disabled: !hardware.available ? "(Out of stock)" : false,
}));
return availableHardwares
.filter((hardware) => hardware.name.toLowerCase().includes(input.toLowerCase()))
.map((hardware) => ({
name: hardware.name,
value: hardware,
disabled: !hardware.available ? "(Out of stock)" : false,
}));
},
});
console.log("Loading regions...");
const hardwareRegions = regions.filter((region) => region.capacities.some((capacity) => capacity.hardwareName === hardware.name && capacity.capacity > 0));
const pricingResult = await preview_instance_pricing_1.previewInstancePricing.execute({
hardwareName: hardware.name,
regions: hardwareRegions.map((region) => region.id),
maxContainers: 1,
});
if (pricingResult.isLeft()) {
console.error(pricingResult.value.message);
return;
}
const regionPricingDict = pricingResult.value.pricesPerRegion.reduce((acc, { regionId, ...pricing }) => {
acc[regionId] = pricing;
return acc;
}, {});
const selectedRegions = await (0, prompts_1.checkbox)({
message: "Select one or more regions",
required: true,
choices: hardwareRegions.map((region) => ({
name: regionPricingDict[region.id]
? `${region.name} - $${regionPricingDict[region.id].pricePerHour}/hour`
: region.name,
value: region,
})),
});
const startupCommand = await (0, prompts_1.input)({
message: "Startup command",
});
const minContainers = await (0, prompts_1.number)({
message: "Minimum number of containers (0-3)",
default: 1,
validate: (value) => {
if (value === undefined) {
return true;
}
if (value < 0 || value > 3) {
return "Minimum number of containers must be between 0 and 3";
}
return true;
},
});
const maxContainers = await (0, prompts_1.number)({
message: "Maximum number of containers (1-25)",
required: true,
default: 1,
validate: (value) => {
if (value === undefined) {
return true;
}
if (value < 1 || value > 25) {
return "Maximum number of containers must be between 1 and 25";
}
return true;
},
});
const cooldownPeriod = await (0, prompts_1.number)({
message: "Cooldown period (in seconds)",
});
const timeout = await (0, prompts_1.number)({
message: "Timeout (in seconds)",
});
console.log("Container Deployment Triggers");
const cpu = await (0, prompts_1.number)({
message: "CPU usage (%)",
});
const gpuUtilization = await (0, prompts_1.number)({
message: "GPU usage (%)",
});
const gpuMemory = await (0, prompts_1.number)({
message: "GPU memory (%)",
});
const memory = await (0, prompts_1.number)({
message: "RAM usage (%)",
});
const http = await (0, prompts_1.number)({
message: "HTTP requests (/sec)",
});
const envVars = await (0, prompts_1.editor)({
message: "Environment Variables",
});
let envs = {};
if (envVars.length > 0) {
envs = envVars.split("\n").reduce((acc, env) => {
const [key, value] = env.split("=");
if (key && value) {
acc[key] = value;
}
return acc;
}, {});
}
const name = await (0, prompts_1.input)({
message: "Instance name",
default: "my-ai-inference-instance",
required: true,
});
console.log("Calculating final price...");
const finalPrice = await preview_instance_pricing_1.previewInstancePricing.execute({
hardwareName: hardware.name,
regions: selectedRegions.map((region) => region.id),
maxContainers,
});
if (finalPrice.isLeft()) {
console.error(finalPrice.value.message);
return;
}
console.log(`Final price: $${finalPrice.value.totalPrice.pricePerHour}/hour | $${finalPrice.value.totalPrice.pricePerMonth}/month`);
const confirmDeployment = await (0, prompts_1.confirm)({
message: "Are you sure you want to deploy this instance?",
});
if (!confirmDeployment) {
console.log("Deployment cancelled");
return;
}
console.log("Deploying instance...");
const launchInstanceResult = await launch_instance_1.launchInferenceInstance.execute({
modelId: (_d = model === null || model === void 0 ? void 0 : model.id) !== null && _d !== void 0 ? _d : null,
registryId: (_e = registry === null || registry === void 0 ? void 0 : registry._id) !== null && _e !== void 0 ? _e : null,
containerPort,
name,
regionsIds: selectedRegions.map((region) => region.id),
minContainers: minContainers !== null && minContainers !== void 0 ? minContainers : 1,
maxContainers,
cooldownPeriod,
triggers: {
cpu,
gpuUtilization,
gpuMemory,
memory,
http,
},
timeout,
envs,
hardwareName: hardware.name,
startupCommand: startupCommand !== null && startupCommand !== void 0 ? startupCommand : null,
});
if (launchInstanceResult.isLeft()) {
console.error(launchInstanceResult.value.message);
return;
}
const instance = launchInstanceResult.value;
console.log("Instance launched successfully! Instance ID: ", instance._id);
});
}
//# sourceMappingURL=launch.js.map