handler.py

import torch
from typing import  Dict
from transformers import pipeline
from datasets import load_dataset
from transformers.pipelines.audio_utils import ffmpeg_read

SAMPLE_RATE=16000
class EndpointHandler():
    def __init__(self, path=""):
        device = "cuda:0" if torch.cuda.is_available() else "cpu"
        pipe = pipeline(
          "automatic-speech-recognition",
          model="openai/whisper-large",
          chunk_length_s=30,
          device=device,
        )
        
    def __call__(self, data: Dict[str, bytes]) -> Dict[str, str]:
        #ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
        #sample = ds[0]["audio"]
        inputs = data.pop("inputs", data)
        audio_nparray = ffmpeg_read(inputs, 16000)
        audio_tensor = torch.from_numpy(audio_nparray)
        
        prediction = pipe(audio_nparray, return_timestamps=True)
        return {"text": prediction[0]["transcription"]}
        
        # we can also return timestamps for the predictions
        #prediction = pipe(sample, return_timestamps=True)["chunks"]
        #[{'text': ' Mr. Quilter is the apostle of the middle classes and we are glad to welcome his gospel.',
        #  'timestamp': (0.0, 5.44)}]