whisper

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

"""
This script uses the Whisper large-v3-turbo model from OpenAI for automatic speech recognition (ASR). 
The model is finetuned for faster performance with a minor quality trade-off. It leverages the Hugging Face 
Transformers library to load the model and processor, and performs transcription on an input audio file.

Whisper is a state-of-the-art model for ASR and speech translation, proposed in the paper "Robust Speech 
Recognition via Large-Scale Weak Supervision" by Alec Radford et al. from OpenAI. Trained on over 5 million 
hours of labeled data, Whisper demonstrates a strong ability to generalize to many datasets and domains in 
a zero-shot setting.

The script performs the following steps:
1. Checks if a CUDA-enabled GPU is available and sets the appropriate device and data type.
2. Loads the Whisper large-v3-turbo model and processor from the Hugging Face Hub.
3. Initializes an ASR pipeline using the model and processor.
4. Defines a function `transcribe_audio` that takes an audio file path as input, performs transcription, 
   and outputs the result to the terminal and a text file.
5. The script expects an audio file path as a command-line argument and calls the `transcribe_audio` function.

Usage:
    whisper <audio_file>

Dependencies:
    - torch
    - transformers
    - datasets
    - accelerate

Example:
    whisper sample_audio.wav
"""

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

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

model_id = "openai/whisper-large-v3-turbo"

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

processor = AutoProcessor.from_pretrained(model_id)

pipe = pipeline(
    "automatic-speech-recognition",
    model=model,
    tokenizer=processor.tokenizer,
    feature_extractor=processor.feature_extractor,
    torch_dtype=torch_dtype,
    device=device,
)

def transcribe_audio(audio_path):
    # Load audio file
    audio = {"path": audio_path}
    
    # Perform transcription
    result = pipe(audio)
    
    # Get the base filename and directory
    base_filename = os.path.splitext(audio_path)[0]
    output_text_path = base_filename + ".txt"
    
    # Output the result to the terminal
    print(result["text"])
    
    # Save the result to a text file
    with open(output_text_path, "w") as f:
        f.write(result["text"])

if __name__ == "__main__":
    if len(sys.argv) != 2:
        print("Usage: python script.py <audio_file>")
        sys.exit(1)
    
    audio_file = sys.argv[1]
    transcribe_audio(audio_file)