Update app.py

app.py

@@ -228,9 +228,9 @@ def simple_chat(message, temperature: float = 0.8, max_length: int = 4096, top_p
         with open(message['file_name'], 'wb') as temp_file:
             temp_file.write(message['file_content'].read())
             temp_file_path = temp_file.name
-
+            
         choice, contents = mode_load(temp_file_path)
-        #choice, contents = mode_load(message["file_content"])
+        
         if choice == "image":
             conversation.append({"role": "user", "image": contents, "content": message['text']})
         elif choice == "doc":
@@ -266,50 +266,22 @@ def simple_chat(message, temperature: float = 0.8, max_length: int = 4096, top_p
             model.generate(**gen_kwargs)
             for new_text in streamer:
                 buffer += new_text
-            # thread = Thread(target=model.generate, kwargs=gen_kwargs)
-            # thread.start()
-            # for new_text in streamer:
-            #     buffer += new_text
 
-    
     thread = Thread(target=generate_text)
     thread.start()
 
     thread.join()
-    
-    # # Start the generation in a separate thread
-    # generate_text()
-    
-    # # Wait for the generation to finish
-    # thread.join()
 
     print("---------")
     print("Text: ")
     print(buffer)
     print("---------")
 
-    # with torch.no_grad():
-    #     thread = Thread(target=model.generate, kwargs=gen_kwargs)
-    #     thread.start()
-    #     buffer = ""
-    #     for new_text in streamer:
-    #         buffer += new_text
-    #         yield buffer
-        
-    #     print("---------")
-    #     print("Text: ")
-    #     print(buffer)
-    #     print("---------")
-
-    # return PlainTextResponse(buffer)
-
     return PlainTextResponse(buffer)
 
 
 
 
-
-
 # @spaces.GPU()
 # def simple_chat(message: dict, temperature: float = 0.8, max_length: int = 4096, top_p: float = 1, top_k: int = 10, penalty: float = 1.0):
 #     try: