From de2a885ab2b872ed33a031255fa41b6a9fa95943 Mon Sep 17 00:00:00 2001 From: Plachta Date: Sat, 4 Mar 2023 19:02:57 +0800 Subject: [PATCH] will print transcribed text from long audios --- long_audio_transcribe.py | 1 + short_audio_transcribe.py | 2 ++ 2 files changed, 3 insertions(+) diff --git a/long_audio_transcribe.py b/long_audio_transcribe.py index 1242aae..e2292bb 100644 --- a/long_audio_transcribe.py +++ b/long_audio_transcribe.py @@ -27,6 +27,7 @@ if __name__ == "__main__": lang2token = { 'zh': "[ZH]", } + assert(torch.cuda.is_available()), "Please enable GPU in order to run Whisper!" model = whisper.load_model(args.whisper_size) speaker_annos = [] for file in filelist: diff --git a/short_audio_transcribe.py b/short_audio_transcribe.py index 8a460de..04b23ef 100644 --- a/short_audio_transcribe.py +++ b/short_audio_transcribe.py @@ -2,6 +2,7 @@ import whisper import os import torchaudio import argparse +import torch lang2token = { 'zh': "[ZH]", @@ -47,6 +48,7 @@ if __name__ == "__main__": lang2token = { 'zh': "[ZH]", } + assert (torch.cuda.is_available()), "Please enable GPU in order to run Whisper!" model = whisper.load_model(args.whisper_size) parent_dir = "./custom_character_voice/" speaker_names = list(os.walk(parent_dir))[0][1]