Update gradio_app.py

gradio_app.py

@@ -1,13 +1,25 @@
+# Requirements for Hugging Face Spaces:
+# torch
+# torchaudio
+# gradio
+# numpy
+# einops
+# vocos
+# pydub
+# cached_path
+# transformers
+# librosa
+# soundfile
+# spaces
+
 import os
 import re
 import torch
 import torchaudio
 import gradio as gr
-import warnings
 import numpy as np
 import tempfile
 from einops import rearrange
-warnings.filterwarnings("ignore", message=".*weights_only=False.*", category=FutureWarning)
 from vocos import Vocos
 from pydub import AudioSegment, silence
 from model import CFM, UNetT, DiT, MMDiT
@@ -20,7 +32,6 @@ from model.utils import (
 )
 from transformers import pipeline
 import librosa
-import click
 import soundfile as sf
 
 try:
@@ -79,6 +90,7 @@ speed = 1.0
 fix_duration = None
 
 
+@gpu_decorator
 def load_model(repo_name, exp_name, model_cls, model_cfg, ckpt_step):
     ckpt_path = str(cached_path(f"hf://SWivid/{repo_name}/{exp_name}/model_{ckpt_step}.safetensors"))
     # ckpt_path = f"ckpts/{exp_name}/model_{ckpt_step}.pt"  # .pt | .safetensors
@@ -212,6 +224,7 @@ def split_text_into_batches(text, max_chars=200, split_words=SPLIT_WORDS):
     
     return batches
 
+@gpu_decorator
 def infer_batch(ref_audio, ref_text, gen_text_batches, exp_name, remove_silence, progress=gr.Progress()):
     if exp_name == "F5-TTS":
         ema_model = F5TTS_ema_model
@@ -296,6 +309,7 @@ def infer_batch(ref_audio, ref_text, gen_text_batches, exp_name, remove_silence,
 
     return (target_sample_rate, final_wave), spectrogram_path
 
+@gpu_decorator
 def infer(ref_audio_orig, ref_text, gen_text, exp_name, remove_silence, custom_split_words=''):
     if not custom_split_words.strip():
         custom_words = [word.strip() for word in custom_split_words.split(',')]
@@ -345,6 +359,7 @@ def infer(ref_audio_orig, ref_text, gen_text, exp_name, remove_silence, custom_s
     gr.Info(f"Generating audio using {exp_name} in {len(gen_text_batches)} batches")
     return infer_batch((audio, sr), ref_text, gen_text_batches, exp_name, remove_silence)
     
+@gpu_decorator
 def generate_podcast(script, speaker1_name, ref_audio1, ref_text1, speaker2_name, ref_audio2, ref_text2, exp_name, remove_silence):
     # Split the script into speaker blocks
     speaker_pattern = re.compile(f"^({re.escape(speaker1_name)}|{re.escape(speaker2_name)}):", re.MULTILINE)
@@ -681,6 +696,7 @@ with gr.Blocks() as app_emotional:
     # Output audio
     audio_output_emotional = gr.Audio(label="Synthesized Audio")
 
+    @gpu_decorator
     def generate_emotional_speech(
         regular_audio,
         regular_ref_text,
@@ -803,24 +819,11 @@ If you're having issues, try converting your reference audio to WAV or MP3, clip
     )
     gr.TabbedInterface([app_tts, app_podcast, app_emotional, app_credits], ["TTS", "Podcast", "Multi-Style", "Credits"])
 
-@click.command()
-@click.option("--port", "-p", default=None, type=int, help="Port to run the app on")
-@click.option("--host", "-H", default=None, help="Host to run the app on")
-@click.option(
-    "--share",
-    "-s",
-    default=False,
-    is_flag=True,
-    help="Share the app via Gradio share link",
-)
-@click.option("--api", "-a", default=True, is_flag=True, help="Allow API access")
-def main(port, host, share, api):
-    global app
-    print(f"Starting app...")
-    app.queue(api_open=api).launch(
-        server_name=host, server_port=port, share=share, show_api=api
+    print(f"Starting app on Hugging Face Spaces...")
+    app.queue(api_open=True).launch(
+        server_name="0.0.0.0",
+        server_port=7860,
+        share=False,
+        show_api=True
     )
 
-
-if __name__ == "__main__":
-    main()