Evandro Fonseca Junior
7b2386876a
Add `ensureOllamaAvailable` function to lib/ollama.ts for checking and waiting for Ollama server health with configurable timeout. Integrate the check in GET /api/ollama/models route with 3s timeout to prevent UI blocking, and in POST /api/ollama/pull route with 20s timeout to ensure availability before pulling models. Update ChatContainer to always render ModelDropdown for better UX when no model is selected.
47 lines
1.4 KiB
TypeScript
47 lines
1.4 KiB
TypeScript
import { NextResponse } from "next/server";
|
|
import { listModelsViaSdk, listModelsViaHttp, ensureOllamaAvailable } from "@/lib/ollama";
|
|
import {
|
|
estimateCpuVram,
|
|
inferDevice,
|
|
inferQuantization,
|
|
mockModels,
|
|
type ModelInfo,
|
|
} from "@/lib/models";
|
|
|
|
export const dynamic = "force-dynamic";
|
|
|
|
function normalizeModel(model: any): ModelInfo {
|
|
const name: string = model?.model || model?.name || "unknown";
|
|
const sizeBytes: number = model?.size || 0;
|
|
const sizeGB = Number(Math.max(0.1, sizeBytes / 1024 ** 3).toFixed(2));
|
|
const quantization = inferQuantization(name);
|
|
const device = inferDevice(model?.details || name);
|
|
const { cpuPct, vramGB } = estimateCpuVram({ sizeGB, quantization, device });
|
|
return {
|
|
id: name,
|
|
name,
|
|
sizeGB,
|
|
quantization,
|
|
device,
|
|
estCpuUsagePct: cpuPct,
|
|
estVramGB: vramGB,
|
|
};
|
|
}
|
|
|
|
export async function GET() {
|
|
try {
|
|
let models: any[] = [];
|
|
// Tentar garantir disponibilidade rapidamente (não bloquear longamente a UI)
|
|
await ensureOllamaAvailable({ timeoutMs: 3000 }).catch(() => {});
|
|
try {
|
|
models = await listModelsViaSdk();
|
|
} catch {
|
|
models = await listModelsViaHttp();
|
|
}
|
|
const normalized = models.map((m) => normalizeModel(m));
|
|
return NextResponse.json({ offline: false, models: normalized });
|
|
} catch (e) {
|
|
return NextResponse.json({ offline: true, models: mockModels });
|
|
}
|
|
}
|