STT Speaker Diarization

import { z } from "zod";
import { zodResponseFormat } from "openai/helpers/zod";

const DiarizationSchema = z.object({
	full_text: z.string(),
	chunks: z.array(
		z.object({
			speaker_id: z.string(),
			text: z.string(),
			start_time: z.number(),
			end_time: z.number(),
		})
	),
	number_of_speakers: z.number(),
});

const response = await interfaze.chat.completions.create({
	model: "interfaze-beta",
	messages: [
		{
			role: "user",
			content: [
				{ type: "text", text: "Transcribe and identify the speakers in the audio file" },
				{
					type: "file",
					file: {
						filename: "stt_multispeaker.mp3",
						file_data: "https://r2public.jigsawstack.com/interfaze/examples/stt_multispeaker.mp3",
					},
				},
			],
		},
	],
	response_format: zodResponseFormat(DiarizationSchema, "diarization_schema"),
});

console.log(response.choices[0].message.content);

//@ts-expect-error precontext is not typed
const precontext = response.precontext;
console.log("STT Results:", precontext?.[0]?.result);

import { z } from "zod";
import { zodResponseFormat } from "openai/helpers/zod";

const SentimentSchema = z.object({
	full_text: z.string(),
	chunks: z.array(
		z.object({
			speaker_id: z.string(),
			text: z.string(),
			sentiment: z.enum(["positive", "negative", "neutral"]).describe("sentiment of the audio chunk"),
			start_time: z.number(),
			end_time: z.number(),
		})
	),
	number_of_speakers: z.number(),
});

const response = await interfaze.chat.completions.create({
	model: "interfaze-beta",
	messages: [
		{
			role: "user",
			content: [
				{ type: "text", text: "Transcribe the audio file, identify the speakers, and analyze the sentiment of each speaker" },
				{
					type: "file",
					file: {
						filename: "stt_call.mp3",
						file_data: "https://r2public.jigsawstack.com/interfaze/examples/stt_call.mp3",
					},
				},
			],
		},
	],
	response_format: zodResponseFormat(SentimentSchema, "sentiment_schema"),
});

console.log(response.choices[0].message.content);

//@ts-expect-error precontext is not typed
const precontext = response.precontext;
console.log("STT Results:", precontext?.[0]?.result);

import { z } from "zod";
import { zodResponseFormat } from "openai/helpers/zod";

const response = await interfaze.chat.completions.create({
	model: "interfaze-beta",
	messages: [
		{
			role: "system",
			content: "<task>speech_to_text</task>",
		},
		{
			role: "user",
			content: [
				{ type: "text", text: "Transcribe and identify the speakers in the audio file https://r2public.jigsawstack.com/interfaze/examples/stt_long_audio_sample_3.mp3" },
			],
		},
	],
	response_format: zodResponseFormat(z.any(), "empty_schema"),
});

console.log(response.choices[0].message.content);