From e33f8919d03f88b54411be8c7fade5550e697e3c Mon Sep 17 00:00:00 2001
From: Plachta <liusongting07@gmail.com>
Date: Fri, 21 Apr 2023 20:43:02 +0800
Subject: [PATCH] added new base model (pure Chinese)

---
 denoise_audio.py          | 12 ++++++++----
 short_audio_transcribe.py | 11 ++++++++---
 2 files changed, 16 insertions(+), 7 deletions(-)

diff --git a/denoise_audio.py b/denoise_audio.py
index a25f076..81a53b8 100644
--- a/denoise_audio.py
+++ b/denoise_audio.py
@@ -1,9 +1,13 @@
 import os
+import json
 import torchaudio
 raw_audio_dir = "./raw_audio/"
 denoise_audio_dir = "./denoised_audio/"
 filelist = list(os.walk(raw_audio_dir))[0][2]
-
+# 2023/4/21: Get the target sampling rate
+with open("./configs/finetune_speaker.json", 'r', encoding='utf-8') as f:
+    hps = json.load(f)
+target_sr = hps['data']['sampling_rate']
 for file in filelist:
     if file.endswith(".wav"):
         os.system(f"demucs --two-stems=vocals {raw_audio_dir}{file}")
@@ -13,6 +17,6 @@ for file in filelist:
                               channels_first=True)
     # merge two channels into one
     wav = wav.mean(dim=0).unsqueeze(0)
-    if sr != 22050:
-        wav = torchaudio.transforms.Resample(orig_freq=sr, new_freq=22050)(wav)
-    torchaudio.save(denoise_audio_dir + file + ".wav", wav, 22050, channels_first=True)
\ No newline at end of file
+    if sr != target_sr:
+        wav = torchaudio.transforms.Resample(orig_freq=sr, new_freq=target_sr)(wav)
+    torchaudio.save(denoise_audio_dir + file + ".wav", wav, target_sr, channels_first=True)
\ No newline at end of file
diff --git a/short_audio_transcribe.py b/short_audio_transcribe.py
index 04b23ef..2cf3c69 100644
--- a/short_audio_transcribe.py
+++ b/short_audio_transcribe.py
@@ -1,5 +1,6 @@
 import whisper
 import os
+import json
 import torchaudio
 import argparse
 import torch
@@ -54,6 +55,10 @@ if __name__ == "__main__":
     speaker_names = list(os.walk(parent_dir))[0][1]
     speaker_annos = []
     # resample audios
+    # 2023/4/21: Get the target sampling rate
+    with open("./configs/finetune_speaker.json", 'r', encoding='utf-8') as f:
+        hps = json.load(f)
+    target_sr = hps['data']['sampling_rate']
     for speaker in speaker_names:
         for i, wavfile in enumerate(list(os.walk(parent_dir + speaker))[0][2]):
             # try to load file as audio
@@ -63,12 +68,12 @@ if __name__ == "__main__":
                 wav, sr = torchaudio.load(parent_dir + speaker + "/" + wavfile, frame_offset=0, num_frames=-1, normalize=True,
                                           channels_first=True)
                 wav = wav.mean(dim=0).unsqueeze(0)
-                if sr != 22050:
-                    wav = torchaudio.transforms.Resample(orig_freq=sr, new_freq=22050)(wav)
+                if sr != target_sr:
+                    wav = torchaudio.transforms.Resample(orig_freq=sr, new_freq=target_sr)(wav)
                 if wav.shape[1] / sr > 20:
                     print(f"{wavfile} too long, ignoring\n")
                 save_path = parent_dir + speaker + "/" + f"processed_{i}.wav"
-                torchaudio.save(save_path, wav, 22050, channels_first=True)
+                torchaudio.save(save_path, wav, target_sr, channels_first=True)
                 # transcribe text
                 lang, text = transcribe_one(save_path)
                 if lang not in list(lang2token.keys()):