Update app.py

app.py

@@ -1,3 +1,4 @@
+import spaces
 import os
 import random
 import argparse
@@ -5,7 +6,6 @@ import argparse
 import torch
 import gradio as gr
 import numpy as np
-import spaces
 
 import ChatTTS
 
@@ -15,38 +15,39 @@ chat.load_models()
 
 
 
-
 def generate_seed():
     new_seed = random.randint(1, 100000000)
     return {
         "__type__": "update",
         "value": new_seed
-    }
+        }
+
 @spaces.GPU
 def generate_audio(text, temperature, top_P, top_K, audio_seed_input, text_seed_input, refine_text_flag):
+
     torch.manual_seed(audio_seed_input)
     rand_spk = torch.randn(768)
     params_infer_code = {
-        'spk_emb': rand_spk,
+        'spk_emb': rand_spk, 
         'temperature': temperature,
         'top_P': top_P,
         'top_K': top_K,
-    }
+        }
     params_refine_text = {'prompt': '[oral_2][laugh_0][break_6]'}
     
     torch.manual_seed(text_seed_input)
 
     if refine_text_flag:
-        text = chat.infer(text,
+        text = chat.infer(text, 
                           skip_refine_text=False,
                           refine_text_only=True,
                           params_refine_text=params_refine_text,
                           params_infer_code=params_infer_code
                           )
     
-    wav = chat.infer(text,
-                     skip_refine_text=True,
-                     params_refine_text=params_refine_text,
+    wav = chat.infer(text, 
+                     skip_refine_text=True, 
+                     params_refine_text=params_refine_text, 
                      params_infer_code=params_infer_code
                      )
     
@@ -54,50 +55,53 @@ def generate_audio(text, temperature, top_P, top_K, audio_seed_input, text_seed_
     sample_rate = 24000
     text_data = text[0] if isinstance(text, list) else text
 
-    return (sample_rate, audio_data), text_data
+    return [(sample_rate, audio_data), text_data]
+
+
+with gr.Blocks() as demo:
+    gr.Markdown("# Deployed by [ChatTTS.com](https://chattts.com)")
 
-def check_and_generate(text, temperature, top_P, top_K, audio_seed_input, text_seed_input, refine_text_flag):
-    if len(text) > 100:
-        return None, "", "Text is too long, please input text with less than 100 characters."
-    audio, generated_text = generate_audio(text, temperature, top_P, top_K, audio_seed_input, text_seed_input, refine_text_flag)
-    return audio, generated_text, ""
+    default_text = "四川美食确实以辣闻名，但也有不辣的选择。比如甜水面、赖汤圆、蛋烘糕、叶儿粑等，这些小吃口味温和，甜而不腻，也很受欢迎。"        
+    text_input = gr.Textbox(label="Input Text", lines=4, placeholder="Please Input Text...", value=default_text)
 
-def main():
-    with gr.Blocks() as demo:
-        gr.Markdown("# Deployed by [ChatTTS.com](https://chattts.com)")
+    with gr.Row():
+        refine_text_checkbox = gr.Checkbox(label="Refine text", value=True)
+        temperature_slider = gr.Slider(minimum=0.00001, maximum=1.0, step=0.00001, value=0.3, label="Audio temperature")
+        top_p_slider = gr.Slider(minimum=0.1, maximum=0.9, step=0.05, value=0.7, label="top_P")
+        top_k_slider = gr.Slider(minimum=1, maximum=20, step=1, value=20, label="top_K")
 
-        default_text = "四川美食确实以辣闻名，但也有不辣的选择。比如甜水面、赖汤圆、蛋烘糕、叶儿粑等，这些小吃口味温和，甜而不腻，也很受欢迎。"
-        text_input = gr.Textbox(label="Input Text", lines=4, placeholder="Please Input Text...", value=default_text)
-        text_length_warning = gr.Markdown("")
+    with gr.Row():
+        audio_seed_input = gr.Number(value=42, label="Audio Seed")
+        generate_audio_seed = gr.Button("\U0001F3B2")
+        text_seed_input = gr.Number(value=42, label="Text Seed")
+        generate_text_seed = gr.Button("\U0001F3B2")
 
-        with gr.Row():
-            refine_text_checkbox = gr.Checkbox(label="Refine text", value=True)
-            temperature_slider = gr.Slider(minimum=0.00001, maximum=1.0, step=0.00001, value=0.3, label="Audio temperature")
-            top_p_slider = gr.Slider(minimum=0.1, maximum=0.9, step=0.05, value=0.7, label="top_P")
-            top_k_slider = gr.Slider(minimum=1, maximum=20, step=1, value=20, label="top_K")
+    generate_button = gr.Button("Generate")
+        
+    text_output = gr.Textbox(label="Output Text", interactive=False)
+    audio_output = gr.Audio(label="Output Audio")
 
-        with gr.Row():
-            audio_seed_input = gr.Number(value=42, label="Audio Seed")
-            generate_audio_seed = gr.Button("\U0001F3B2")
-            text_seed_input = gr.Number(value=42, label="Text Seed")
-            generate_text_seed = gr.Button("\U0001F3B2")
+    generate_audio_seed.click(generate_seed, 
+                              inputs=[], 
+                              outputs=audio_seed_input)
+        
+    generate_text_seed.click(generate_seed, 
+                             inputs=[], 
+                             outputs=text_seed_input)
+        
+    generate_button.click(generate_audio, 
+                          inputs=[text_input, temperature_slider, top_p_slider, top_k_slider, audio_seed_input, text_seed_input, refine_text_checkbox], 
+                          outputs=[audio_output, text_output])
 
-        generate_button = gr.Button("Generate")
+parser = argparse.ArgumentParser(description='ChatTTS demo Launch')
+parser.add_argument('--server_name', type=str, default='0.0.0.0', help='Server name')
+parser.add_argument('--server_port', type=int, default=8080, help='Server port')
+args = parser.parse_args()
 
-        text_output = gr.Textbox(label="Output Text", interactive=False)
-        audio_output = gr.Audio(label="Output Audio")
+    # demo.launch(server_name=args.server_name, server_port=args.server_port, inbrowser=True)
 
-        generate_audio_seed.click(generate_seed, inputs=[], outputs=audio_seed_input)
-        generate_text_seed.click(generate_seed, inputs=[], outputs=text_seed_input)
-        generate_button.click(check_and_generate, inputs=[text_input, temperature_slider, top_p_slider, top_k_slider, audio_seed_input, text_seed_input, refine_text_checkbox], outputs=[audio_output, text_output, text_length_warning])
 
-    parser = argparse.ArgumentParser(description='ChatTTS demo Launch')
-    parser.add_argument('--server_name', type=str, default='0.0.0.0', help='Server name')
-    parser.add_argument('--server_port', type=int, default=8080, help='Server port')
-    args = parser.parse_args()
 
-    # demo.queue(max_size=1).launch(server_name=args.server_name, server_port=args.server_port, inbrowser=True)
-    demo.queue(max_size=4).launch()
 
 if __name__ == '__main__':
-    main()
+    demo.launch()