Update webui.py

webui.py

@@ -66,6 +66,7 @@ def change_instruction(mode_checkbox_group):
 @spaces.GPU
 def generate_audio(tts_text, mode_checkbox_group, sft_dropdown, prompt_text, prompt_wav_upload, prompt_wav_record, instruct_text,
                    seed, stream, speed):
+    stream=False
     if prompt_wav_upload is not None:
         prompt_wav = prompt_wav_upload
     elif prompt_wav_record is not None:
@@ -76,31 +77,31 @@ def generate_audio(tts_text, mode_checkbox_group, sft_dropdown, prompt_text, pro
     if mode_checkbox_group in ['自然语言控制']:
         if get_cosyvoice().frontend.instruct is False:
             gr.Warning('您正在使用自然语言控制模式, {}模型不支持此模式, 请使用iic/CosyVoice-300M-Instruct模型'.format(args.model_dir))
-            yield (target_sr, default_data)
+            return (target_sr, default_data)
         if instruct_text == '':
             gr.Warning('您正在使用自然语言控制模式, 请输入instruct文本')
-            yield (target_sr, default_data)
+            return (target_sr, default_data)
         if prompt_wav is not None or prompt_text != '':
             gr.Info('您正在使用自然语言控制模式, prompt音频/prompt文本会被忽略')
     # if cross_lingual mode, please make sure that model is iic/CosyVoice-300M and tts_text prompt_text are different language
     if mode_checkbox_group in ['跨语种复刻']:
         if get_cosyvoice().frontend.instruct is True:
             gr.Warning('您正在使用跨语种复刻模式, {}模型不支持此模式, 请使用iic/CosyVoice-300M模型'.format(args.model_dir))
-            yield (target_sr, default_data)
+            return (target_sr, default_data)
         if instruct_text != '':
             gr.Info('您正在使用跨语种复刻模式, instruct文本会被忽略')
         if prompt_wav is None:
             gr.Warning('您正在使用跨语种复刻模式, 请提供prompt音频')
-            yield (target_sr, default_data)
+            return (target_sr, default_data)
         gr.Info('您正在使用跨语种复刻模式, 请确保合成文本和prompt文本为不同语言')
     # if in zero_shot cross_lingual, please make sure that prompt_text and prompt_wav meets requirements
     if mode_checkbox_group in ['3s极速复刻', '跨语种复刻']:
         if prompt_wav is None:
             gr.Warning('prompt音频为空，您是否忘记输入prompt音频？')
-            yield (target_sr, default_data)
+            return (target_sr, default_data)
         if torchaudio.info(prompt_wav).sample_rate < prompt_sr:
             gr.Warning('prompt音频采样率{}低于{}'.format(torchaudio.info(prompt_wav).sample_rate, prompt_sr))
-            yield (target_sr, default_data)
+            return (target_sr, default_data)
     # sft mode only use sft_dropdown
     if mode_checkbox_group in ['预训练音色']:
         if instruct_text != '' or prompt_wav is not None or prompt_text != '':
@@ -109,7 +110,7 @@ def generate_audio(tts_text, mode_checkbox_group, sft_dropdown, prompt_text, pro
     if mode_checkbox_group in ['3s极速复刻']:
         if prompt_text == '':
             gr.Warning('prompt文本为空，您是否忘记输入prompt文本？')
-            yield (target_sr, default_data)
+            return (target_sr, default_data)
         if instruct_text != '':
             gr.Info('您正在使用3s极速复刻模式，预训练音色/instruct文本会被忽略！')
 
@@ -117,24 +118,24 @@ def generate_audio(tts_text, mode_checkbox_group, sft_dropdown, prompt_text, pro
         logging.info('get sft inference request')
         set_all_random_seed(seed)
         for i in get_cosyvoice().inference_sft(tts_text, sft_dropdown, stream=stream, speed=speed):
-            yield (target_sr, i['tts_speech'].numpy().flatten())
+            return (target_sr, i['tts_speech'].numpy().flatten())
     elif mode_checkbox_group == '3s极速复刻':
         logging.info('get zero_shot inference request')
         prompt_speech_16k = postprocess(load_wav(prompt_wav, prompt_sr))
         set_all_random_seed(seed)
         for i in get_cosyvoice().inference_zero_shot(tts_text, prompt_text, prompt_speech_16k, stream=stream, speed=speed):
-            yield (target_sr, i['tts_speech'].numpy().flatten())
+            return (target_sr, i['tts_speech'].numpy().flatten())
     elif mode_checkbox_group == '跨语种复刻':
         logging.info('get cross_lingual inference request')
         prompt_speech_16k = postprocess(load_wav(prompt_wav, prompt_sr))
         set_all_random_seed(seed)
         for i in get_cosyvoice().inference_cross_lingual(tts_text, prompt_speech_16k, stream=stream, speed=speed):
-            yield (target_sr, i['tts_speech'].numpy().flatten())
+            return (target_sr, i['tts_speech'].numpy().flatten())
     else:
         logging.info('get instruct inference request')
         set_all_random_seed(seed)
         for i in get_cosyvoice().inference_instruct(tts_text, sft_dropdown, instruct_text, stream=stream, speed=speed):
-            yield (target_sr, i['tts_speech'].numpy().flatten())
+            return (target_sr, i['tts_speech'].numpy().flatten())
 
 # SDK模型下载
 import platform
@@ -214,7 +215,7 @@ with gr.Blocks() as demo:
 
     generate_button = gr.Button("生成音频")
 
-    audio_output = gr.Audio(label="合成音频", autoplay=True, streaming=True)
+    audio_output = gr.Audio(label="合成音频", autoplay=True, streaming=False)
 
     seed_button.click(generate_seed, inputs=[], outputs=seed)
     generate_button.click(generate_audio,