services/asr.py

import torch
from transformers import pipeline

class Transcriber:
    def __init__(self, conf):
        self.conf = conf
        self.pipeline = self.asr_pipeline()

    def asr_pipeline(self):
        return pipeline(
            self.conf["model"]["asr"]["type"],
            model=self.conf["model"]["asr"]["transcriber"],
            device=0 if torch.cuda.is_available() else -1  # Use 0 for GPU, -1 for CPU
        )
    def get_pipeline(self):
        return self.pipeline
    
    def run(self, file_path):
        kwargs = {"max_new_tokens": self.conf["model"]["asr"]["max_new_tokens"]}
        output = self.pipeline(
            file_path, 
            generate_kwargs=kwargs,
            return_timestamps=True,
        )
        return output.get("chunks", output)  # Use .get to avoid key errors