output-files-compression

README.md

@@ -7,7 +7,7 @@ sdk: gradio
 app_file: app.py
 pinned: true
 duplicated_from: Thafx/Demucs_v4_2s_HT
-sdk_version: 5.20.1
+sdk_version: 5.35.0
 ---
 
 # Configuration

app.py

@@ -1,13 +1,23 @@
 import os
 import gradio as gr
 from scipy.io.wavfile import write
+from pydub import AudioSegment
 
+def wav_to_mp3(wav_path, mp3_path, bitrate="128k"):  # 128k is a good default for quality/compression
+    audio = AudioSegment.from_wav(wav_path)
+    audio.export(mp3_path, format="mp3", bitrate=bitrate)
 
 def inference(audio):
-  os.makedirs("out", exist_ok=True)
-  write('test.wav', audio[0], audio[1])
-  os.system("python3 -m demucs.separate -n htdemucs --two-stems=vocals test.wav -o out")
-  return "./out/htdemucs/test/vocals.wav","./out/htdemucs/test/no_vocals.wav"
+    os.makedirs("out", exist_ok=True)
+    write('test.wav', audio[0], audio[1])
+    os.system("python3 -m demucs.separate -n htdemucs --two-stems=vocals test.wav -o out")
+    vocals_wav = "./out/htdemucs/test/vocals.wav"
+    no_vocals_wav = "./out/htdemucs/test/no_vocals.wav"
+    vocals_mp3 = "./out/htdemucs/test/vocals.mp3"
+    no_vocals_mp3 = "./out/htdemucs/test/no_vocals.mp3"
+    wav_to_mp3(vocals_wav, vocals_mp3)
+    wav_to_mp3(no_vocals_wav, no_vocals_mp3)
+    return vocals_mp3, no_vocals_mp3
     
 title = "Demucs Music Source Separation (v4)"
 article = "<p style='text-align: center'><a href='https://arxiv.org/abs/1911.13254' target='_blank'>Music Source Separation in the Waveform Domain</a> | <a href='https://github.com/facebookresearch/demucs' target='_blank'>Github Repo</a> | <a href='https://www.thafx.com' target='_blank'>//THAFX</a></p>"

requirements.txt

@@ -4,4 +4,5 @@ invisible-watermark
 fonts
 font-roboto
 numpy<1.26
+pydub