DeepSeek chatbot with Ollama + setup instructions for visitors

app.py

@@ -1,45 +1,88 @@
-# app.py - HuggingFace Spaces Chatbot with Local LLM
+# app.py - Local Ollama Chatbot with DeepSeek
 import gradio as gr
+import requests
 import logging
 import sys
 import traceback
 from datetime import datetime
-from huggingface_hub import hf_hub_download
-
-# Logging setup
-logging.basicConfig(
-    level=logging.INFO,
-    format="%(asctime)s | %(levelname)-8s | %(message)s",
-    handlers=[logging.StreamHandler(sys.stdout)]
-)
+
+# Logging
+logging.basicConfig(level=logging.INFO, format="%(asctime)s | %(message)s", handlers=[logging.StreamHandler(sys.stdout)])
 logger = logging.getLogger("CHATBOT")
 
 logger.info("=" * 60)
-logger.info("  CPU CHATBOT - HUGGINGFACE SPACES EDITION")
+logger.info("  DEEPSEEK CHATBOT - LOCAL OLLAMA")
 logger.info("=" * 60)
 
-# Model config
-MODEL_REPO = "TheBloke/TinyLlama-1.1B-Chat-v1.0-GGUF"
-MODEL_FILE = "tinyllama-1.1b-chat-v1.0.Q4_K_M.gguf"
+# Config
+OLLAMA_URL = "http://localhost:11434"
+MODEL = "deepseek-coder:6.7b-instruct-q6_K"
 
-logger.info(f"Downloading model: {MODEL_FILE}")
+# Check connection
+ollama_connected = False
 try:
-    model_path = hf_hub_download(repo_id=MODEL_REPO, filename=MODEL_FILE, cache_dir="/tmp/models")
-    logger.info(f"Model path: {model_path}")
-except Exception as e:
-    logger.error(f"Download failed: {e}")
-    model_path = None
-
-# Load model
-llm = None
-if model_path:
-    try:
-        from llama_cpp import Llama
-        logger.info("Loading model into memory (30-60 sec)...")
-        llm = Llama(model_path=model_path, n_ctx=2048, n_threads=2, n_batch=128, verbose=False)
-        logger.info("MODEL LOADED!")
-    except Exception as e:
-        logger.error(f"Load failed: {e}")
+    resp = requests.get(f"{OLLAMA_URL}/api/tags", timeout=3)
+    if resp.status_code == 200:
+        models = [m["name"] for m in resp.json().get("models", [])]
+        if MODEL in models:
+            ollama_connected = True
+            logger.info(f"Ollama connected! Using {MODEL}")
+        else:
+            logger.warning(f"Model {MODEL} not found. Available: {models}")
+except:
+    logger.warning("Ollama not running")
+
+# Install instructions
+INSTALL_INSTRUCTIONS = """
+# 🚀 Local AI Chatbot Setup
+
+This chatbot runs on **YOUR computer** using Ollama + DeepSeek AI.
+
+## Step 1: Install Ollama
+
+**Mac:**
+```bash
+brew install ollama
+```
+
+**Linux:**
+```bash
+curl -fsSL https://ollama.com/install.sh | sh
+```
+
+**Windows:** Download from [ollama.com/download](https://ollama.com/download)
+
+---
+
+## Step 2: Start Ollama & Download Model
+
+```bash
+# Start Ollama (run in terminal, keep open)
+ollama serve
+
+# Download DeepSeek model (6.7GB, one-time)
+ollama pull deepseek-coder:6.7b-instruct-q6_K
+```
+
+---
+
+## Step 3: Refresh This Page
+
+Once Ollama is running with the model, **refresh this page** and the chat will work!
+
+---
+
+### Alternative: Clone & Run Locally
+
+```bash
+git clone https://huggingface.co/spaces/ndwdgda/cpu
+cd cpu
+pip install gradio requests
+python app.py
+```
+
+Then open http://127.0.0.1:7860
+"""
 
 # Chat function
 request_count = 0
@@ -51,12 +94,11 @@ def chat_with_ai(message, history):
     
     logger.info(f"[{rid}] User: {message}")
     
-    if llm is None:
-        return "Error: Model not loaded. Check logs."
+    if not ollama_connected:
+        return "❌ Ollama not connected! See the Setup tab for instructions."
     
     try:
-        # Build prompt
-        prompt = "You are a helpful AI assistant.\n\n"
+        prompt = "You are a helpful AI coding assistant.\n\n"
         if history:
             for item in history:
                 if isinstance(item, dict):
@@ -73,30 +115,60 @@ def chat_with_ai(message, history):
         
         prompt += f"User: {message}\nAssistant:"
         
-        logger.info(f"[{rid}] Generating response...")
+        logger.info(f"[{rid}] Calling DeepSeek...")
         start = datetime.now()
         
-        output = llm(prompt, max_tokens=256, stop=["User:", "\n\n"], echo=False)
+        resp = requests.post(
+            f"{OLLAMA_URL}/api/generate",
+            json={"model": MODEL, "prompt": prompt, "stream": False},
+            timeout=120
+        )
         
         elapsed = (datetime.now() - start).total_seconds()
-        response = output["choices"][0]["text"].strip()
-        
-        logger.info(f"[{rid}] Response in {elapsed:.1f}s: {response[:100]}...")
-        return response
         
+        if resp.status_code == 200:
+            result = resp.json()
+            text = result.get("response", "")
+            logger.info(f"[{rid}] Response in {elapsed:.1f}s: {text[:100]}...")
+            return text.strip()
+        else:
+            return f"Error: {resp.status_code} - {resp.text[:200]}"
+            
     except Exception as e:
         logger.error(f"[{rid}] Error: {e}")
-        logger.error(traceback.format_exc())
         return f"Error: {e}"
 
-# Gradio UI
-logger.info("Building Gradio UI...")
-demo = gr.ChatInterface(
-    fn=chat_with_ai,
-    title="CPU Chatbot",
-    description="**Powered by TinyLlama 1.1B** - Runs entirely on HuggingFace's servers!",
-    examples=["Hello!", "What is AI?", "Tell me a joke"],
-)
+# UI
+logger.info("Building UI...")
+
+with gr.Blocks(theme=gr.themes.Soft()) as demo:
+    gr.Markdown("# 🤖 DeepSeek AI Chatbot")
+    
+    if ollama_connected:
+        gr.Markdown(f"**✅ Connected to Ollama** | Model: `{MODEL}`")
+    else:
+        gr.Markdown("**❌ Ollama not connected** - See **Setup** tab below for instructions!")
+    
+    with gr.Tabs():
+        with gr.TabItem("💬 Chat"):
+            chatbot = gr.Chatbot(height=400)
+            msg = gr.Textbox(placeholder="Ask me anything..." if ollama_connected else "Install Ollama first (see Setup tab)", label="Message")
+            with gr.Row():
+                send = gr.Button("Send", variant="primary")
+                clear = gr.ClearButton([msg, chatbot])
+            
+            def respond(message, history):
+                if not message.strip():
+                    return "", history
+                bot_response = chat_with_ai(message, history)
+                history = history + [[message, bot_response]]
+                return "", history
+            
+            msg.submit(respond, [msg, chatbot], [msg, chatbot])
+            send.click(respond, [msg, chatbot], [msg, chatbot])
+        
+        with gr.TabItem("📋 Setup"):
+            gr.Markdown(INSTALL_INSTRUCTIONS)
 
 logger.info("READY!")
 

requirements.txt

@@ -1,3 +1,3 @@
 gradio==4.19.2
 huggingface_hub==0.22.2
-llama-cpp-python==0.2.90
+requests