whisper2

#!/usr/bin/env python
# -*- coding: utf-8 -*-

import torch
from transformers import AutoModelForSpeechSeq2Seq, AutoProcessor, pipeline
import sys
import os
import warnings

# Suppress specific warnings
warnings.filterwarnings("ignore", category=FutureWarning)
warnings.filterwarnings("ignore", category=UserWarning)

MODEL_NAME = "openai/whisper-large-v3"
BATCH_SIZE = 8

device = "cuda:0" if torch.cuda.is_available() else "cpu"
torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32

model = AutoModelForSpeechSeq2Seq.from_pretrained(
    MODEL_NAME, torch_dtype=torch_dtype, low_cpu_mem_usage=True, use_safetensors=True
)
model.to(device)

processor = AutoProcessor.from_pretrained(MODEL_NAME)

pipe = pipeline(
    "automatic-speech-recognition",
    model=model,
    tokenizer=processor.tokenizer,
    feature_extractor=processor.feature_extractor,
    #    max_new_tokens=448,
    chunk_length_s=30,
    batch_size=BATCH_SIZE,
    return_timestamps=True,
    torch_dtype=torch_dtype,
    device=device,
)

def transcribe(audio_file_path, task="transcribe"):
    if not os.path.exists(audio_file_path):
        print(f"Error: The file '{audio_file_path}' does not exist.")
        return

    try:
        with torch.no_grad():
            result = pipe(audio_file_path, generate_kwargs={"task": task})
        from pprint import pprint
        pprint(result)
        return result["text"]
    except Exception as e:
        print(f"Error during transcription: {str(e)}")
        return None

if __name__ == "__main__":
    if len(sys.argv) < 2:
        print("Usage: python script.py <audio_file_path> [task]")
        print("task can be 'transcribe' or 'translate' (default is 'transcribe')")
        sys.exit(1)

    audio_file_path = sys.argv[1]
    task = sys.argv[2] if len(sys.argv) > 2 else "transcribe"

    if task not in ["transcribe", "translate"]:
        print("Error: task must be either 'transcribe' or 'translate'")
        sys.exit(1)

    result = transcribe(audio_file_path, task)
    if result:
        print("Transcription result:")
        print(result)