handler.py

from typing import Dict, List, Any
from transformers import AutoProcessor, MusicgenForConditionalGeneration
import torch
import array
import base64
import io
import wave
import numpy as np

class EndpointHandler:
    def __init__(self, path=""):
        # load model and processor from path
        self.processor = AutoProcessor.from_pretrained(path)
        self.model = MusicgenForConditionalGeneration.from_pretrained(path, torch_dtype=torch.float16).to("cuda")

    def __call__(self, data: Dict[str, Any]) -> Dict[str, str]:
        """
        Args:
            data (:dict:):
                The payload with the text prompt and generation parameters.
        """
        # process input
        inputs = data.pop("inputs", data)
        parameters = data.pop("parameters", None)

        # preprocess
        inputs = self.processor(
            text=[inputs],
            padding=True,
            return_tensors="pt",).to("cuda")

        # pass inputs with all kwargs in data
        with torch.autocast("cuda"):
                audio_values = self.model.generate(**inputs, do_sample=True, guidance_scale=3, max_new_tokens=400)

        # postprocess the prediction  
        sampling_rate = self.model.config.audio_encoder.sampling_rate
        audio_samples = audio_values[0].cpu().numpy()[0].tolist()

        audio_samples = [int(min(max(sample * 32000, -32000), 32000)) for sample in audio_samples]
    
        # Create BytesIO object to capture the audio in-memory
        audio_io = io.BytesIO()
    
        # Create WAV file
        with wave.open(audio_io, 'wb') as wf:
            wf.setnchannels(1)
            wf.setsampwidth(2)  # 2 bytes for 16-bit PCM
            wf.setframerate(sampling_rate)
            wf.writeframes(array.array('h', audio_samples).tobytes())
    
        audio_base64 = base64.b64encode(audio_io.getvalue()).decode('utf-8')
    
        return [{'sampling_rate': sampling_rate, 'audio': audio_base64}]