Merge branch 'main' of https://huggingface.co/spaces/rushic24/Priyanka-Chopra-TTS

app.py

@@ -12,10 +12,9 @@ description = "Generate english speech from text using a Tacotron2 model" \
 article = """<p style='text-align: center'>
                 <a href='https://rushichaudhari.github.io/posts/2022-01-12-lets-clone-the-voice-of-priyanka-chopra-jonas/' 
                 target='blank' 
-                class='footer'>Blog</a> | 
-                <a href='https://github.com/eugenesiow/practical-ml' target='_blank' 
-                class='footer'>Github Repo</a></p>"""
-examples = ["Generate english speech from text using a Tacotron2 model."]
+                class='footer'>Blog</a></p>"""
+examples = ["Generate english speech from text using a Tacotron2 model.", 
+            "Two roads diverged in a wood, I took the one less traveled by, And that has made all the difference."]
 
 def inference(text: str):
     synthesize(
@@ -32,7 +31,7 @@ gr.Interface(
     inputs=[
         gr.inputs.Textbox(
             label="Input",
-            default="你好吗？我很好。",
+            default=examples[0],
         ),
     ],
     outputs=gr.outputs.Audio(label="Output"),

synthesize.py

@@ -200,34 +200,4 @@ def synthesize(
 
         if audio_path:
             audio = vocoder.generate_audio(mel_outputs_postnet)
-            write(audio_path, sample_rate, audio)
-
-
-if __name__ == "__main__":
-    """Synthesize audio using model and vocoder"""
-    parser = argparse.ArgumentParser(description="Synthesize audio using model and vocoder")
-    parser.add_argument("-m", "--model_path", type=str, help="tacotron2 model path", required=True)
-    parser.add_argument("-vm", "--vocoder_model_path", type=str, help="vocoder model path", required=True)
-    parser.add_argument("-hc", "--hifigan_config_path", type=str, help="hifigan_config path", required=True)
-    parser.add_argument("-t", "--text", type=str, help="text to synthesize", required=True)
-    parser.add_argument("-g", "--graph_output_path", type=str, help="path to save alignment graph to", required=False)
-    parser.add_argument("-a", "--audio_output_path", type=str, help="path to save output audio to", required=False)
-    parser.add_argument("--silence_padding", type=float, help="Padding between sentences in seconds", default=0.15)
-    parser.add_argument("--sample_rate", type=int, help="Audio sample rate", default=22050)
-    args = parser.parse_args()
-
-    assert os.path.isfile(args.model_path), "Model not found"
-    assert os.path.isfile(args.vocoder_model_path), "vocoder model not found"
-
-    model = load_model(args.model_path)
-    vocoder = Hifigan(args.vocoder_model_path, args.hifigan_config_path)
-
-    synthesize(
-        model=model,
-        text=args.text,
-        graph_path=args.graph_output_path,
-        audio_path=args.audio_output_path,
-        vocoder=vocoder,
-        silence_padding=args.silence_padding,
-        sample_rate=args.sample_rate,
-    )
+            write(audio_path, sample_rate, audio)