Update app.py

app.py

@@ -1,70 +1,72 @@
 import streamlit as st
-from llama_cpp import Llama
-import os
+from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
+import torch
 
-# 1. Page Config
-st.set_page_config(page_title="Qwen Coder GGUF", page_icon="🤖", layout="wide")
-st.title("🚀 Qwen2.5-Coder (GGUF CPU)")
-st.caption("Optimized for Hugging Face Free Tier")
+# 1. Page Configuration (Hide Sidebar & Set Theme)
+st.set_page_config(page_title="Claude Clone", page_icon="🤖", layout="centered")
 
-# 2. Model Loading with specific error handling
+# Custom CSS to force-hide the sidebar button and clean up the UI
+st.markdown("""
+    <style>
+        [data-testid="stSidebar"] {display: none;}
+        [data-testid="stHeader"] {background: rgba(0,0,0,0);}
+        .stChatMessage {border-radius: 15px; padding: 10px; margin-bottom: 10px;}
+    </style>
+""", unsafe_allow_html=True)
+
+st.title("Qwen 2.5 Coder 🤖")
+st.caption("A lightweight, powerful Claude-style clone powered by Alibaba's Qwen 2.5 1.5B")
+
+# 2. Load Model & Tokenizer
 @st.cache_resource
-def load_llm():
-    try:
-        # We use the 3B-Q4_K_M for a good balance of logic and RAM usage
-        return Llama.from_pretrained(
-            repo_id="Qwen/Qwen2.5-Coder-3B-Instruct-GGUF",
-            filename="qwen2.5-coder-3b-instruct-q4_k_m.gguf", # Explicit filename
-            n_ctx=4096,               # Context window
-            n_threads=2,              # Matches HF Free Tier vCPUs
-            verbose=False             # Reduces log clutter
-        )
-    except Exception as e:
-        st.error(f"Error loading model: {e}")
-        return None
+def load_model():
+    model_id = "Qwen/Qwen2.5-Coder-1.5B-Instruct"
+    tokenizer = AutoTokenizer.from_pretrained(model_id)
+    # Using torch_dtype="auto" to handle CPU/GPU environments automatically
+    model = AutoModelForCausalLM.from_pretrained(
+        model_id, 
+        torch_dtype="auto", 
+        device_map="auto"
+    )
+    return pipeline("text-generation", model=model, tokenizer=tokenizer)
 
-llm = load_llm()
+generator = load_model()
 
-# 3. Enhanced "Claude-style" System Prompt
+# 3. Initialize Chat History
 if "messages" not in st.session_state:
     st.session_state.messages = [
-        {
-            "role": "system", 
-            "content": "You are an expert AI programming assistant. When asked to write code, provide the full file content. Use clear comments, follow best practices, and ensure the code is production-ready."
-        }
+        {"role": "system", "content": "You are a helpful assistant named Claude-Clone. You excel at coding and technical tasks."}
     ]
 
-# Display history
-for msg in st.session_state.messages:
-    if msg["role"] != "system":
-        with st.chat_message(msg["role"]):
-            st.markdown(msg["content"])
+# Display Chat History
+for message in st.session_state.messages:
+    if message["role"] != "system":
+        with st.chat_message(message["role"]):
+            st.markdown(message["content"])
 
-# 4. Generation Logic
-if prompt := st.chat_input("Write a Python script to scrape a website..."):
+# 4. Chat Input & Logic
+if prompt := st.chat_input("How can I help you today?"):
+    # User Message
     st.session_state.messages.append({"role": "user", "content": prompt})
     with st.chat_message("user"):
         st.markdown(prompt)
 
-    if llm:
-        with st.chat_message("assistant"):
-            response_placeholder = st.empty()
-            full_response = ""
+    # Assistant Response
+    with st.chat_message("assistant"):
+        with st.spinner("Thinking..."):
+            # Format history for the model
+            full_prompt = st.session_state.messages
             
-            # Stream the response
-            output = llm.create_chat_completion(
-                messages=st.session_state.messages,
-                stream=True,
-                max_tokens=1500, # Increased for "Complete Code" tasks
-                temperature=0.1  # Lower temperature = more precise code
+            # Generate response
+            outputs = generator(
+                full_prompt, 
+                max_new_tokens=512, 
+                do_sample=True, 
+                temperature=0.7,
+                top_p=0.9
             )
             
-            for chunk in output:
-                if 'content' in chunk['choices'][0]['delta']:
-                    token = chunk['choices'][0]['delta']['content']
-                    full_response += token
-                    response_placeholder.markdown(full_response + "▌")
+            response = outputs[0]['generated_text'][-1]['content']
+            st.markdown(response)
             
-            response_placeholder.markdown(full_response)
-        
-        st.session_state.messages.append({"role": "assistant", "content": full_response})
+    st.session_state.messages.append({"role": "assistant", "content": response})