Fix errors

app.py

@@ -1,7 +1,6 @@
 import gradio as gr
 from transformers import AutoTokenizer, AutoModelForCausalLM
 import torch
-import markdown
 
 model_name = "deepseek-ai/deepseek-coder-1.3b-base"
 
@@ -9,17 +8,19 @@ tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
 model = AutoModelForCausalLM.from_pretrained(model_name, trust_remote_code=True)
 model.eval()
 
-def respond(message, chat_history, max_tokens, temperature, top_p):
-    chat_history = chat_history or []
-    chat_history.append(("User", message))
+def respond(message, history, max_tokens, temperature, top_p):
+    history = history or []
+    # Append user message as dict with role and content
+    history.append({"role": "user", "content": message})
     
-    full_prompt = ""
-    for speaker, text in chat_history:
-        prefix = "User: " if speaker == "User" else "Assistant: "
-        full_prompt += prefix + text + "\n"
-    full_prompt += "Assistant: "
+    # Create prompt by concatenating conversation history as text
+    prompt = ""
+    for msg in history:
+        prefix = f"{msg['role'].capitalize()}: "
+        prompt += prefix + msg["content"] + "\n"
+    prompt += "Assistant: "
     
-    inputs = tokenizer(full_prompt, return_tensors="pt")
+    inputs = tokenizer(prompt, return_tensors="pt")
     outputs = model.generate(
         **inputs,
         max_new_tokens=max_tokens,
@@ -28,29 +29,24 @@ def respond(message, chat_history, max_tokens, temperature, top_p):
         do_sample=True,
         pad_token_id=tokenizer.eos_token_id,
     )
-    reply = tokenizer.decode(outputs[0], skip_special_tokens=True)[len(full_prompt):].strip()
-    chat_history.append(("Assistant", reply))
     
-    formatted_history = []
-    for i in range(0, len(chat_history), 2):
-        user_msg = chat_history[i][1] if i < len(chat_history) else ""
-        bot_msg = chat_history[i+1][1] if i+1 < len(chat_history) else ""
-        # Render assistant message as markdown
-        formatted_history.append([user_msg, gr.Markdown(bot_msg)])
+    reply = tokenizer.decode(outputs[0], skip_special_tokens=True)[len(prompt):].strip()
+    # Append assistant response
+    history.append({"role": "assistant", "content": reply})
     
-    return formatted_history, ""
+    return history, ""
 
 with gr.Blocks() as demo:
     gr.Markdown("# DeepSeek Coder Chatbot")
     
-    chatbot = gr.Chatbot()
+    chatbot = gr.Chatbot(type="messages")
     with gr.Row():
-        user_input = gr.Textbox(show_label=False, placeholder="Enter your prompt here and press Enter")
+        user_input = gr.Textbox(show_label=False, placeholder="Enter your prompt and press Enter")
     with gr.Row():
         max_tokens = gr.Slider(1, 1024, value=512, step=1, label="Max Tokens")
         temperature = gr.Slider(0.1, 1.0, value=0.7, step=0.05, label="Temperature")
         top_p = gr.Slider(0.1, 1.0, value=0.9, step=0.05, label="Top-p")
-    
+
     def user_submit(text, history, max_tokens, temperature, top_p):
         if not text.strip():
             return history, ""