Update app.py

app.py

@@ -1,12 +1,7 @@
-import os, sys
-import tempfile
-import gradio as gr
-from src.gradio_demo import SadTalker  
-# from src.utils.text2speech import TTSTalker
-from huggingface_hub import snapshot_download
-
+import os
 import torch
 import librosa
+import gradio as gr
 from scipy.io.wavfile import write
 from transformers import WavLMModel
 
@@ -27,9 +22,10 @@ import edge_tts
 import tempfile
 import anyio
 
+import os, sys
+import gradio as gr
+from src.gradio_demo import SadTalker  
 
-def get_source_image(image):   
-        return image
 
 try:
     import webui  # in webui
@@ -49,23 +45,8 @@ def ref_video_fn(path_of_ref_video):
         return gr.update(value=True)
     else:
         return gr.update(value=False)
-    
-def download_model():
-    REPO_ID = 'vinthony/SadTalker-V002rc'
-    snapshot_download(repo_id=REPO_ID, local_dir='./checkpoints', local_dir_use_symlinks=True)
-
-def sadtalker_demo():
-
-    download_model()
 
-    sad_talker = SadTalker(lazy_load=True)
-    # tts_talker = TTSTalker()
-
-download_model()
-sad_talker = SadTalker(lazy_load=True)
-
-
-# ChatGLM2 & FreeVC
+sad_talker = SadTalker("checkpoints", "src/config", lazy_load=True)
 
 '''
 def get_wavlm():
@@ -141,7 +122,7 @@ except Exception:
     logger.warning("Windows, cant run time.tzset()")
 
 # model_name = "THUDM/chatglm2-6b"
-model_name = "THUDM/chatglm2-6b-int4"
+model_name = "THUDM/chatglm2-6b"
 
 RETRY_FLAG = False
 
@@ -332,15 +313,14 @@ async def text_to_speech_edge(text, language_code):
     return tmp_path
 
 
-with gr.Blocks(title="ChatGLM2-6B-int4", theme=gr.themes.Soft(text_size="sm"), analytics_enabled=False) as demo:
+with gr.Blocks(title="ChatGLM2-6B-int4", theme=gr.themes.Soft(text_size="sm")) as demo:
     gr.HTML("<center>"
-            "<h1>📺💕🎶 - ChatGLM2+声音克隆+视频对话：和喜欢的角色畅所欲言吧！</h1>"
+            "<h1>🥳💕🎶 - ChatGLM2 + 声音克隆：和你喜欢的角色畅所欲言吧！</h1>"
             "</center>")
-    gr.Markdown("## <center>🥳 - ChatGLM2+FreeVC+SadTalker，为您打造沉浸式的视频对话体验，支持中英双语</center>")
+    gr.Markdown("## <center>💡 - 第二代ChatGLM大语言模型 + FreeVC变声，为您打造独一无二的沉浸式对话体验，支持中英双语</center>")
     gr.Markdown("## <center>🌊 - 更多精彩应用，尽在[滔滔AI](http://www.talktalkai.com)；滔滔AI，为爱滔滔！💕</center>")
-    gr.Markdown("### <center>⭐ - 如果您喜欢这个程序，欢迎给我的[GitHub项目](https://github.com/KevinWang676/ChatGLM2-Voice-Cloning)点赞支持！</center>")
-
-    with gr.Tab("🍻 - ChatGLM2聊天区"):
+    gr.Markdown("### <center>⭐ - 如果您喜欢这个程序，欢迎给我的[Github项目](https://github.com/KevinWang676/ChatGLM2-Voice-Cloning)点赞支持！</center>")
+    with gr.Tab("Chat"):
         with gr.Accordion("📒 相关信息", open=False):
             _ = f""" ChatGLM2的可选参数信息：
                 * Low temperature: responses will be more deterministic and focused; High temperature: responses more creative.
@@ -364,7 +344,7 @@ with gr.Blocks(title="ChatGLM2-6B-int4", theme=gr.themes.Soft(text_size="sm"), a
                 submitBtn = gr.Button("开始和GLM2交流吧", variant="primary")
                 deleteBtn = gr.Button("删除最新一轮对话", variant="secondary")
                 retryBtn = gr.Button("重新生成最新一轮对话", variant="secondary")
-                    
+                
         with gr.Accordion("🔧 更多设置", open=False):
             with gr.Row():
                 emptyBtn = gr.Button("清空所有聊天记录")
@@ -382,8 +362,8 @@ with gr.Blocks(title="ChatGLM2-6B-int4", theme=gr.themes.Soft(text_size="sm"), a
                 temperature = gr.Slider(
                     0.01, 1, value=0.95, step=0.01, label="Temperature", interactive=True
                 )
-    
-    
+
+
         with gr.Row():
             test1 = gr.Textbox(label="GLM2的最新回答 (可编辑)", lines = 3)
             with gr.Column():
@@ -401,7 +381,7 @@ with gr.Blocks(title="ChatGLM2-6B-int4", theme=gr.themes.Soft(text_size="sm"), a
             audio_cloned =  gr.Audio(label="为您生成的专属声音克隆音频", type='filepath')
     
         clone_btn.click(convert, inputs=[model_choice, audio1, audio2], outputs=[audio_cloned])
-            
+        
         history = gr.State([])
         past_key_values = gr.State(None)
     
@@ -441,7 +421,7 @@ with gr.Blocks(title="ChatGLM2-6B-int4", theme=gr.themes.Soft(text_size="sm"), a
         emptyBtn.click(
             reset_state, outputs=[chatbot, history, past_key_values, test1], show_progress="full"
         )
-    
+
         retryBtn.click(
             retry_last_answer,
             inputs=[
@@ -457,7 +437,7 @@ with gr.Blocks(title="ChatGLM2-6B-int4", theme=gr.themes.Soft(text_size="sm"), a
             outputs=[chatbot, history, past_key_values, test1],
         )
         deleteBtn.click(delete_last_turn, [chatbot, history], [chatbot, history])
-    
+
         with gr.Accordion("📔 提示词示例", open=False):
             etext = """In America, where cars are an important part of the national psyche, a decade ago people had suddenly started to drive less, which had not happened since the oil shocks of the 1970s. """
             examples = gr.Examples(
@@ -513,90 +493,55 @@ with gr.Blocks(title="ChatGLM2-6B-int4", theme=gr.themes.Soft(text_size="sm"), a
             api_name="tr1",
         )
         # """
-    with gr.Tab("📺 - 视频聊天区"):
+
+    with gr.Tab("Video"):
         with gr.Row().style(equal_height=False):
             with gr.Column(variant='panel'):
                 with gr.Tabs(elem_id="sadtalker_source_image"):
-                    with gr.TabItem('图片上传'):
+                    with gr.TabItem('Upload image'):
                         with gr.Row():
-                            source_image = gr.Image(label="请上传一张您喜欢角色的图片", source="upload", type="filepath", elem_id="img2img_image").style(width=512)
-    
+                            source_image = gr.Image(label="Source image", source="upload", type="filepath", elem_id="img2img_image").style(width=512)
     
                 with gr.Tabs(elem_id="sadtalker_driven_audio"):
-                    with gr.TabItem('💡您还可以将视频下载到本地'):
-    
-                        with gr.Row():
-                            driven_audio = audio_cloned
-                            driven_audio_no = gr.Audio(label="Use IDLE mode, no audio is required", source="upload", type="filepath", visible=False)
-    
-                            with gr.Column():
-                                use_idle_mode = gr.Checkbox(label="Use Idle Animation", visible=False)
-                                length_of_audio = gr.Number(value=5, label="The length(seconds) of the generated video.", visible=False)
-                                use_idle_mode.change(toggle_audio_file, inputs=use_idle_mode, outputs=[driven_audio, driven_audio_no]) # todo
-    
-                        with gr.Row():
-                            ref_video = gr.Video(label="Reference Video", source="upload", type="filepath", elem_id="vidref", visible=False).style(width=512)
-    
-                            with gr.Column():
-                                use_ref_video = gr.Checkbox(label="Use Reference Video", visible=False)
-                                ref_info = gr.Radio(['pose', 'blink','pose+blink', 'all'], value='pose', label='Reference Video',info="How to borrow from reference Video?((fully transfer, aka, video driving mode))", visible=False)
-    
-                            ref_video.change(ref_video_fn, inputs=ref_video, outputs=[use_ref_video]) # todo
-    
+                    with gr.TabItem('Upload OR TTS'):
+                        with gr.Column(variant='panel'):
+                            driven_audio = gr.Audio(label="Input audio", source="upload", type="filepath")
     
-            with gr.Column(variant='panel'):
+            with gr.Column(variant='panel'): 
                 with gr.Tabs(elem_id="sadtalker_checkbox"):
-                    with gr.TabItem('视频设置'):
+                    with gr.TabItem('Settings'):
+                        gr.Markdown("need help? please visit our [best practice page](https://github.com/OpenTalker/SadTalker/blob/main/docs/best_practice.md) for more detials")
                         with gr.Column(variant='panel'):
                             # width = gr.Slider(minimum=64, elem_id="img2img_width", maximum=2048, step=8, label="Manually Crop Width", value=512) # img2img_width
                             # height = gr.Slider(minimum=64, elem_id="img2img_height", maximum=2048, step=8, label="Manually Crop Height", value=512) # img2img_width
-                            with gr.Row():
-                                pose_style = gr.Slider(minimum=0, maximum=45, step=1, label="Pose style", value=0, visible=False) #
-                                exp_weight = gr.Slider(minimum=0, maximum=3, step=0.1, label="expression scale", value=1, visible=False) #
-                                blink_every = gr.Checkbox(label="use eye blink", value=True, visible=False)
-    
-                            with gr.Row():
-                                size_of_image = gr.Radio([256, 512], value=256, label='face model resolution', info="use 256/512 model?", visible=False) #
-                                preprocess_type = gr.Radio(['crop', 'full'], value='crop', label='是否聚焦角色面部', info="crop：视频会聚焦角色面部；full：视频会显示图片全貌")
-    
-                            with gr.Row():
-                                is_still_mode = gr.Checkbox(label="静态模式 (开启静态模式，角色的面部动作会减少；默认开启)", value=True)
-                                facerender = gr.Radio(['facevid2vid','pirender'], value='facevid2vid', label='facerender', info="which face render?", visible=False)
-    
-                            with gr.Row():
-                                batch_size = gr.Slider(label="Batch size (数值越大，生成速度越快；若显卡性能好，可增大数值)", step=1, maximum=32, value=2)
-                                enhancer = gr.Checkbox(label="GFPGAN as Face enhancer", value=True, visible=False)
-    
-                            submit = gr.Button('开始视频聊天吧', elem_id="sadtalker_generate", variant='primary')
-    
+                            pose_style = gr.Slider(minimum=0, maximum=46, step=1, label="Pose style", value=0) # 
+                            size_of_image = gr.Radio([256, 512], value=256, label='face model resolution', info="use 256/512 model?") # 
+                            preprocess_type = gr.Radio(['crop', 'resize','full', 'extcrop', 'extfull'], value='crop', label='preprocess', info="How to handle input image?")
+                            is_still_mode = gr.Checkbox(label="Still Mode (fewer hand motion, works with preprocess `full`)")
+                            batch_size = gr.Slider(label="batch size in generation", step=1, maximum=10, value=2)
+                            enhancer = gr.Checkbox(label="GFPGAN as Face enhancer")
+                            submit = gr.Button('Generate', elem_id="sadtalker_generate", variant='primary')
+                            
                 with gr.Tabs(elem_id="sadtalker_genearted"):
-                        gen_video = gr.Video(label="为您生成的专属视频", format="mp4").style(width=256)
-    
-    
+                        gen_video = gr.Video(label="Generated video", format="mp4").style(width=256)
     
         submit.click(
-                fn=sad_talker.test,
-                inputs=[source_image,
-                        driven_audio,
-                        preprocess_type,
-                        is_still_mode,
-                        enhancer,
-                        batch_size,
-                        size_of_image,
-                        pose_style,
-                        facerender,
-                        exp_weight,
-                        use_ref_video,
-                        ref_video,
-                        ref_info,
-                        use_idle_mode,
-                        length_of_audio,
-                        blink_every
-                        ],
-                outputs=[gen_video]
-                )    
+                    fn=sad_talker.test, 
+                    inputs=[source_image,
+                            driven_audio,
+                            preprocess_type,
+                            is_still_mode,
+                            enhancer,
+                            batch_size,                            
+                            size_of_image,
+                            pose_style
+                            ], 
+                    outputs=[gen_video]
+                    )
+
+
     gr.Markdown("### <center>注意❗：请不要生成会对个人以及组织造成侵害的内容，此程序仅供科研、学习及个人娱乐使用。</center>")
-    gr.Markdown("<center>💡- 如何使用此程序：输入您对ChatGLM的提问后，依次点击“开始和GLM2交流吧”、“生成对应的音频吧”、“开始AI声音克隆吧”、“开始视频聊天吧”三个按键即可；使用声音克隆功能时，请先上传一段您喜欢的音频</center>")
+    gr.Markdown("<center>💡 - 如何使用此程序：输入您对ChatGLM的提问后，依次点击“开始和GLM2交流吧”、“生成对应的音频吧”、“开始AI声音克隆吧”三个按键即可；使用声音克隆功能时，请先上传一段您喜欢的音频</center>")
     gr.HTML('''
         <div class="footer">
                     <p>🌊🏞️🎶 - 江水东流急，滔滔无尽声。 明·顾璘