Update app.py

app.py

@@ -1,104 +1,120 @@
 import gradio as gr
-from transformers import AutoTokenizer, AutoModelForCausalLM, TextStreamer
-from peft import PeftModel
-import torch
-
-# --- Load tokenizer and model ---
-tokenizer = AutoTokenizer.from_pretrained("unsloth/Qwen3-1.7B")
-base_model = AutoModelForCausalLM.from_pretrained(
-    "unsloth/Qwen3-1.7B",
-    torch_dtype=torch.float32,
-    device_map={"": "cpu"}
-)
-model = PeftModel.from_pretrained(base_model, "khazarai/BioGenesis-ToT")
-model = model.to("cpu") 
-
-# --- Define chatbot logic ---
-def generate_response(user_input, chat_history):
-    # Append user message to history
-    chat_history.append({"role": "user", "content": user_input})
 
-    # Convert history to prompt
-    text = tokenizer.apply_chat_template(
-        chat_history,
-        tokenize=False,
-        add_generation_prompt=True,
-        enable_thinking=True,
-    )
+def generate_response(user_input, chat_history, hf_token):
+    if not hf_token:
+        return chat_history, chat_history, "❌ Please enter your Hugging Face API token first."
+
+    if not user_input.strip():
+        return chat_history, chat_history, ""
+
+    model_id = "khazarai/BioGenesis-ToT"  # Your hosted model
+
+    headers = {
+        "Authorization": f"Bearer {hf_token}"
+    }
 
-    # Tokenize and send to GPU
-    inputs = tokenizer(text, return_tensors="pt").to("cpu")
+    # Combine chat history into a conversation string
+    conversation = ""
+    for msg in chat_history:
+        role = "User" if msg["role"] == "user" else "Assistant"
+        conversation += f"{role}: {msg['content']}\n"
+    conversation += f"User: {user_input}\nAssistant:"
 
-    # Generate
-    output_tokens = model.generate(
-        **inputs,
-        max_new_tokens=2200,
-        temperature=0.6,
-        top_p=0.95,
-        top_k=20,
+    # Send the request to HF Inference API
+    payload = {
+        "inputs": conversation,
+        "parameters": {
+            "max_new_tokens": 2200,
+            "temperature": 0.6,
+            "top_p": 0.95,
+            "top_k": 20,
+        }
+    }
+
+    response = requests.post(
+        f"https://api-inference.huggingface.co/models/{model_id}",
+        headers=headers,
+        json=payload,
     )
 
-    # Decode output
-    response = tokenizer.decode(output_tokens[0], skip_special_tokens=True)
+    if response.status_code != 200:
+        return chat_history, chat_history, f"⚠️ API Error: {response.text}"
+
+    result = response.json()
 
-    # Extract only model's reply (avoid repeating prompt)
-    response = response.split(user_input)[-1].strip()
+    # Extract model output
+    if isinstance(result, list) and len(result) > 0 and "generated_text" in result[0]:
+        reply = result[0]["generated_text"].split("Assistant:")[-1].strip()
+    else:
+        reply = "🤔 Sorry, I couldn’t generate a response."
+
+    chat_history.append({"role": "user", "content": user_input})
+    chat_history.append({"role": "assistant", "content": reply})
 
-    # Add model reply to chat history
-    chat_history.append({"role": "assistant", "content": response})
-    
-    # Prepare Gradio display format
-    gr_chat_history = [(m["content"], chat_history[i+1]["content"]) 
-                       for i, m in enumerate(chat_history[:-1]) 
-                       if m["role"] == "user"]
-    
-    return gr_chat_history, chat_history
+    gr_chat_history = [
+        (m["content"], chat_history[i + 1]["content"])
+        for i, m in enumerate(chat_history[:-1])
+        if m["role"] == "user"
+    ]
 
+    return gr_chat_history, chat_history, ""
+
+
+# --- UI Design ---
 with gr.Blocks(theme=gr.themes.Soft(primary_hue="emerald", secondary_hue="slate")) as demo:
     gr.HTML("""
     <div style="text-align: center; margin-bottom: 20px;">
         <h1 style="font-family: 'Inter', sans-serif; font-weight: 800; color: #047857; font-size: 2.2em;">
-            🧬 BioGenesis-ToT Chatbot
+            🧬 BioGenesis-ToT Chatbot (Hosted on Hugging Face)
         </h1>
         <p style="color: #4B5563; font-size: 1.05em; margin-top: -10px;">
-            Your AI companion for biology, biochemistry, and life sciences.
+            Talk to your biology-trained LLM — no GPU needed, just your Hugging Face token ⚡
         </p>
     </div>
     """)
 
     with gr.Row():
         with gr.Column(scale=6):
+            hf_token = gr.Textbox(
+                placeholder="Enter your Hugging Face API Token here...",
+                label="🔑 Hugging Face Token",
+                type="password",
+            )
+
             chatbot = gr.Chatbot(
                 label="BioGenesis Chat",
                 height=550,
                 bubble_full_width=False,
                 show_copy_button=True,
                 avatar_images=(
-                    "https://www.flaticon.com/free-icon/user_9187532",  # user icon
-                    "https://cdn-icons-png.flaticon.com/512/4140/4140048.png",  # bot icon
+                    "https://cdn-icons-png.flaticon.com/512/1077/1077012.png",
+                    "https://cdn-icons-png.flaticon.com/512/4140/4140048.png",
                 ),
             )
+
             user_input = gr.Textbox(
                 placeholder="Ask me about cell biology, molecular structure, or biochemistry...",
                 label="💬 Your question",
                 lines=3,
                 autofocus=True,
             )
+
+            status_box = gr.Textbox(label="Status", interactive=False)
+
             with gr.Row():
                 send_btn = gr.Button("🚀 Send", variant="primary")
                 clear_btn = gr.Button("🧹 Clear Chat")
 
     state = gr.State([])
 
-    send_btn.click(generate_response, [user_input, state], [chatbot, state])
-    user_input.submit(generate_response, [user_input, state], [chatbot, state])
-    clear_btn.click(lambda: ([], []), None, [chatbot, state])
+    send_btn.click(generate_response, [user_input, state, hf_token], [chatbot, state, status_box])
+    user_input.submit(generate_response, [user_input, state, hf_token], [chatbot, state, status_box])
+    clear_btn.click(lambda: ([], [], ""), None, [chatbot, state, status_box])
 
     gr.HTML("""
     <div style="text-align: center; margin-top: 25px; color: #6B7280; font-size: 0.9em;">
-        Powered by <b>Qwen3-1.7B + BioGenesis-ToT</b> | Built with ❤️ using Gradio
+        Powered by <b>Hugging Face Inference API</b> | Built with ❤️ using Gradio
     </div>
     """)
 
-demo.launch(share=True)
-
+demo.launch(share=True)