30 lines
855 B
TypeScript
30 lines
855 B
TypeScript
import { createOpenAICompatible } from "@ai-sdk/openai-compatible";
|
|
import { streamText } from "ai";
|
|
|
|
import { renderStream } from "./src/render";
|
|
|
|
const provider = createOpenAICompatible({
|
|
name: "llama.cpp",
|
|
apiKey: "local",
|
|
baseURL: "http://127.0.0.1:8080/v1",
|
|
includeUsage: true, // Include usage information in streaming responses
|
|
});
|
|
|
|
const { fullStream, reasoningText, text } = streamText({
|
|
model: provider("qwen3.5-35b-a3b"),
|
|
prompt: Bun.argv[2] ?? "Introduce yourself.",
|
|
});
|
|
|
|
renderStream(fullStream);
|
|
|
|
const reasoning = await reasoningText;
|
|
Bun.file("reasoning.md").write(reasoning ?? "");
|
|
|
|
const response = await text;
|
|
Bun.file("response.md").write(response);
|
|
|
|
const [numColumns] = process.stdout.getWindowSize();
|
|
Bun.spawnSync({
|
|
cmd: ["glow", "-w", `${numColumns}`, "response.md"],
|
|
stdio: ["inherit", "inherit", "inherit"],
|
|
});
|