Update app.py

app.py

@@ -2,48 +2,46 @@ from transformers import AutoModelForCausalLM, AutoTokenizer
 import gradio as gr
 import torch
 
-
 title = "EZChat"
 description = "A State-of-the-Art Large-scale Pretrained Response generation model Qwen's 7B-Chat"
 examples = [["How are you?"]]
 
-
 tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen-7B-Chat", trust_remote_code=True)
-model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen-7B-Chat", device_map="auto", trust_remote_code=True).eval()
+model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen-7B-Chat", trust_remote_code=True).eval()
+
+history = []  # Initialize chat history
 
-def predict(input, history=[]):
-    # Check if input is not None and eos_token is not None
+def predict(input, history=history):
     if input is not None and tokenizer.eos_token is not None:
         combined_input = input + tokenizer.eos_token
-        # Rest of your code using combined_input
-    else:
-        # Handle the case where input or tokenizer.eos_token is None
-        print("Input or eos_token is None. Cannot concatenate.")
+        new_user_input_ids = tokenizer.encode(combined_input, return_tensors="pt")
 
-    # append the new user input tokens to the chat history
-    bot_input_ids = torch.cat([torch.LongTensor(history), new_user_input_ids], dim=-1)
+        # Append the new user input tokens to the chat history
+        bot_input_ids = torch.cat([torch.LongTensor(history), new_user_input_ids], dim=-1)
 
-    # generate a response
-    history = model.generate(
-        bot_input_ids, max_length=20, pad_token_id=tokenizer.eos_token_id
-    ).tolist()
+        # Generate a response
+        generated_response_ids = model.generate(
+            bot_input_ids, max_length=20, pad_token_id=tokenizer.eos_token_id
+        )
 
-    # convert the tokens to text, and then split the responses into lines
-    response = tokenizer.decode(history[0]).split("<|endoftext|>")
-    # print('decoded_response-->>'+str(response))
-    response = [
-        (response[i], response[i + 1]) for i in range(0, len(response) - 1, 2)
-    ]  # convert to tuples of list
-    # print('response-->>'+str(response))
-    return response, history
+        # Convert the generated response tokens to text
+        response = tokenizer.decode(generated_response_ids[0], skip_special_tokens=True)
 
+        # Append the user input and generated response to the chat history
+        history.extend(new_user_input_ids[0].tolist())
+        history.extend(generated_response_ids[0].tolist())
+
+        return response, history
+
+    else:
+        print("Input or eos_token is None. Cannot concatenate.")
 
 gr.Interface(
     fn=predict,
     title=title,
     description=description,
     examples=examples,
-    inputs=["text", "state"],
-    outputs=["chatbot", "state"],
+    inputs=["text", "text"],
+    outputs=["text", "text"],
     theme="ParityError/Anime",
-).launch()
+).launch()