feature in tts, convert wav file to mp3

jaseci_ai_kit/jaseci_ai_kit/modules/tts/action_utils.py

@@ -13,6 +13,7 @@
 from speechbrain.pretrained import Tacotron2 as SpeechBrain
 from speechbrain.pretrained import HIFIGAN
 from scipy.io.wavfile import write
+from pydub import AudioSegment
 
 config = configparser.ConfigParser()
 config.read(os.path.join(os.path.dirname(__file__), "config.cfg"))
@@ -309,3 +310,20 @@ def prediction(input_text, seq2seqmodel, vocorder):
         else:
             print("No valid vocorder")
     return audio_numpy
+
+
+def wave2mp3(wav_file):
+    """
+    Saving the given wav file in mp3 format.
+    """
+    save_status = False
+    status = {"save_status": save_status}
+    mp3_file_name = wav_file[:-3] + "mp3"
+    try:
+        AudioSegment.from_wav(wav_file).export(mp3_file_name, format="mp3")
+        save_status = True
+        status["file_name"] = mp3_file_name
+    except Exception as e:
+        print(e)
+
+    return status

jaseci_ai_kit/jaseci_ai_kit/modules/tts/tests/fixtures/tts.jac

@@ -30,5 +30,10 @@ walker test_load_vocorder_v1{
     report tts.load_vocorder("waveglow");
 }
 
+walker test_wave2mp3{
+    can tts.wav2mp3;
+    report tts.wav2mp3("jaseci_ai_kit/modules/tts/tests/test.wav");
+}
+
 
 

jaseci_ai_kit/jaseci_ai_kit/modules/tts/tests/test_tts.py

@@ -45,6 +45,11 @@ def test_load_vocorder_v2(self, ret):
     def test_load_seq2seq_model_v2(self, ret):
         self.assertEqual(ret["success"], True)
 
+    @pytest.mark.order(7)
+    @jac_testcase("tts.jac", "test_wave2mp3")
+    def test_wave2mp3(self, ret):
+        self.assertEqual(ret["success"], True)
+
     @classmethod
     def tearDownClass(cls):
         super(TestTTSModule, cls).tearDownClass()

jaseci_ai_kit/jaseci_ai_kit/modules/tts/tts.py

@@ -16,6 +16,7 @@
     prediction,
     load_seq2seq_model,
     load_vocorder_model,
+    wave2mp3,
 )
 
 warnings.filterwarnings("ignore")
@@ -94,6 +95,18 @@ def save_audio(audio_data: list, path: str = "", rate: int = rate):
         raise HTTPException(status_code=500, detail=str(e))
 
 
+@jaseci_action(act_group=["tts"], allow_remote=True)
+def wav2mp3(wav_file: str):
+    """
+    Saving the given wav file in mp3 format.
+    """
+    try:
+        status = wave2mp3(wav_file=wav_file)
+        return status
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+
 if __name__ == "__main__":
     print("Text to Speech Synthesizer up and running")
     launch_server(port=8000)

jaseci_ai_kit/setup.py

@@ -33,12 +33,13 @@ def get_ver():
         "beautifulsoup4 >= 4.10.0, < 4.11.0",
         "umap-learn==0.5.3",
         "hdbscan==0.8.29",
-        "librosa==0.9.2",
+        "librosa<=0.9.2",
         "protobuf>=3.20.1,<3.21",
         "inflect<=6.0.2",
         "unidecode==1.3.6",
         "soundfile<=0.11.0",
         "speechbrain==0.5.13",
+        "pydub==0.25.1",
     ],
     package_data={
         "": ["*.json", "*.cfg", "VERSION", "*.yaml"],