Update app.py

app.py

@@ -1,16 +1,84 @@
+# import gradio as gr
+# from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
+
+# # --- Load Model ---
+# MODEL_PATH = "./tinyllama-jobskills-final_update_4"  # Model files are in the repo root
+
+# tokenizer = AutoTokenizer.from_pretrained(MODEL_PATH)
+# model = AutoModelForCausalLM.from_pretrained(
+#     MODEL_PATH,
+#     trust_remote_code=True,
+#     device_map="auto",
+#     low_cpu_mem_usage=True
+
+# )
+
+# pipe = pipeline(
+#     "text-generation",
+#     model=model,
+#     tokenizer=tokenizer,
+#     device_map="auto"
+# )
+
+# # --- Define Chat Function ---
+# def chat_fn(message, history):
+#     history_text = ""
+#     for user, bot in history:
+#         history_text += f"User: {user}\nAssistant: {bot}\n"
+#     history_text += f"User: {message}\nAssistant:"
+
+#     # generate response
+#     response = pipe(
+#         history_text,
+#         max_new_tokens=256,
+#         do_sample=True,
+#         temperature=0.7,
+#         top_p=0.9
+#     )[0]["generated_text"]
+
+#     # extract assistant reply
+#     reply = response.split("Assistant:")[-1].strip()
+#     return reply
+
+# # --- Gradio UI ---
+# with gr.Blocks() as demo:
+#     gr.Markdown("## 🚀 Chat with My Custom Model")
+
+#     chatbot = gr.Chatbot()
+#     msg = gr.Textbox(label="Type your message")
+#     clear = gr.Button("Clear")
+
+#     def user_fn(user_message, chat_history):
+#         bot_message = chat_fn(user_message, chat_history)
+#         chat_history.append((user_message, bot_message))
+#         return "", chat_history
+
+#     msg.submit(user_fn, [msg, chatbot], [msg, chatbot])
+#     clear.click(lambda: None, None, chatbot, queue=False)
+
+# # --- Launch ---
+# if __name__ == "__main__":
+#     demo.launch()
+
+
+
+
+
+
+
+
 import gradio as gr
 from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
 
 # --- Load Model ---
-MODEL_PATH = "./tinyllama-jobskills-final_update_4"  # Model files are in the repo root
+MODEL_PATH = "./tinyllama-jobskills-final_update_4"  # Model files path
 
 tokenizer = AutoTokenizer.from_pretrained(MODEL_PATH)
 model = AutoModelForCausalLM.from_pretrained(
     MODEL_PATH,
     trust_remote_code=True,
-    device_map="auto",
+    device_map="auto",  # Will use GPU if available
     low_cpu_mem_usage=True
-
 )
 
 pipe = pipeline(
@@ -22,29 +90,30 @@ pipe = pipeline(
 
 # --- Define Chat Function ---
 def chat_fn(message, history):
+    # Convert history to text
     history_text = ""
-    for user, bot in history:
+    for user, bot in history[-5:]:  # Use only last 5 exchanges for speed
         history_text += f"User: {user}\nAssistant: {bot}\n"
     history_text += f"User: {message}\nAssistant:"
 
-    # generate response
+    # Generate response
     response = pipe(
         history_text,
-        max_new_tokens=256,
-        do_sample=True,
-        temperature=0.7,
-        top_p=0.9
+        max_new_tokens=64,  # Reduced for CPU
+        do_sample=False,    # Greedy decoding for speed
+        temperature=1.0,
+        top_p=1.0
     )[0]["generated_text"]
 
-    # extract assistant reply
+    # Extract assistant reply
     reply = response.split("Assistant:")[-1].strip()
     return reply
 
 # --- Gradio UI ---
 with gr.Blocks() as demo:
-    gr.Markdown("## 🚀 Chat with My Custom Model")
+    gr.Markdown("## 🚀 Chat with My Custom Model (CPU-Friendly)")
 
-    chatbot = gr.Chatbot()
+    chatbot = gr.Chatbot(type="messages")  # updated for future versions
     msg = gr.Textbox(label="Type your message")
     clear = gr.Button("Clear")
 
@@ -54,8 +123,9 @@ with gr.Blocks() as demo:
         return "", chat_history
 
     msg.submit(user_fn, [msg, chatbot], [msg, chatbot])
-    clear.click(lambda: None, None, chatbot, queue=False)
+    clear.click(lambda: [], None, chatbot, queue=False)  # clears chat
 
 # --- Launch ---
 if __name__ == "__main__":
     demo.launch()
+