Create app.py

app.py

@@ -0,0 +1,266 @@
+# =====================
+# 🦁 SIMBA AI - First African LLM
+# =====================
+
+import gradio as gr
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer
+from sentence_transformers import SentenceTransformer
+import faiss
+import numpy as np
+import os
+
+print("🚀 Initializing Simba AI - First African LLM...")
+
+# =====================
+# LOAD AI MODEL
+# =====================
+
+model_name = "mistralai/Mistral-7B-Instruct-v0.2"
+
+try:
+    # Load tokenizer and model
+    tokenizer = AutoTokenizer.from_pretrained(model_name)
+    tokenizer.pad_token = tokenizer.eos_token
+
+    model = AutoModelForCausalLM.from_pretrained(
+        model_name,
+        torch_dtype=torch.float16,
+        device_map="auto",
+    )
+    print("✅ Simba AI Model Loaded Successfully!")
+except Exception as e:
+    print(f"❌ Model loading error: {e}")
+    # Fallback to smaller model if needed
+    model_name = "microsoft/DialoGPT-large"
+    tokenizer = AutoTokenizer.from_pretrained(model_name)
+    tokenizer.pad_token = tokenizer.eos_token
+    model = AutoModelForCausalLM.from_pretrained(model_name)
+    print("✅ Fallback model loaded!")
+
+# =====================
+# AFRICAN KNOWLEDGE BASE
+# =====================
+
+simba_knowledge_base = [
+    # CODING
+    {"question": "Python add function", "answer": "def add(a, b): return a + b"},
+    {"question": "Factorial function", "answer": "def factorial(n): return 1 if n == 0 else n * factorial(n-1)"},
+    {"question": "Reverse string function", "answer": "def reverse_string(s): return s[::-1]"},
+    {"question": "Check even number", "answer": "def is_even(n): return n % 2 == 0"},
+    {"question": "Multiply function", "answer": "def multiply(x, y): return x * y"},
+    {"question": "Yoruba greeting function", "answer": "def yoruba_greeting(): return 'Báwo ni'"},
+    
+    # MATH
+    {"question": "15 + 27", "answer": "42"},
+    {"question": "8 × 7", "answer": "56"},
+    {"question": "100 - 45", "answer": "55"},
+    {"question": "12 × 12", "answer": "144"},
+    {"question": "25% of 200", "answer": "50"},
+    
+    # YORUBA
+    {"question": "Hello in Yoruba", "answer": "Báwo ni"},
+    {"question": "Thank you in Yoruba", "answer": "Ẹ sé"},
+    {"question": "How are you in Yoruba", "answer": "Ṣe daadaa ni"},
+    {"question": "Good morning in Yoruba", "answer": "Ẹ káàrọ̀"},
+    {"question": "Good night in Yoruba", "answer": "O dàárọ̀"},
+    {"question": "Please in Yoruba", "answer": "Jọ̀wọ́"},
+    
+    # SWAHILI
+    {"question": "Hello in Swahili", "answer": "Hujambo"},
+    {"question": "Thank you in Swahili", "answer": "Asante"},
+    
+    # IGBO
+    {"question": "Hello in Igbo", "answer": "Nnọọ"},
+    {"question": "Thank you in Igbo", "answer": "Daalụ"},
+    
+    # HAUSA
+    {"question": "Hello in Hausa", "answer": "Sannu"},
+    {"question": "Thank you in Hausa", "answer": "Na gode"},
+    
+    # AFRICAN INNOVATION
+    {"question": "M-Pesa", "answer": "Mobile money service launched in Kenya in 2007"},
+    {"question": "Andela", "answer": "Trains African software developers for global companies"},
+]
+
+print(f"✅ African Knowledge Base: {len(simba_knowledge_base)} entries")
+
+# =====================
+# SEARCH SYSTEM
+# =====================
+
+try:
+    embedder = SentenceTransformer('all-MiniLM-L6-v2')
+    
+    # Build search index
+    questions = [item["question"] for item in simba_knowledge_base]
+    question_embeddings = embedder.encode(questions)
+
+    dimension = question_embeddings.shape[1]
+    index = faiss.IndexFlatIP(dimension)
+    faiss.normalize_L2(question_embeddings)
+    index.add(question_embeddings)
+
+    print("✅ Smart Search System Ready!")
+except Exception as e:
+    print(f"❌ Search system error: {e}")
+    index = None
+
+def simba_search(query, top_k=2):
+    """Search African knowledge base"""
+    if index is None:
+        return simba_knowledge_base[:top_k]  # Fallback
+    
+    try:
+        query_embedding = embedder.encode([query])
+        faiss.normalize_L2(query_embedding)
+        
+        scores, indices = index.search(query_embedding, top_k)
+        
+        results = []
+        for i, idx in enumerate(indices[0]):
+            if idx < len(simba_knowledge_base):
+                results.append({
+                    "question": simba_knowledge_base[idx]["question"],
+                    "answer": simba_knowledge_base[idx]["answer"],
+                    "score": scores[0][i]
+                })
+        
+        return results
+    except:
+        return simba_knowledge_base[:top_k]  # Fallback
+
+# =====================
+# SIMBA AI CHAT FUNCTION
+# =====================
+
+def simba_ai_chat(message, history):
+    """Main chat function for Simba AI"""
+    
+    try:
+        # Search for relevant knowledge
+        search_results = simba_search(message, top_k=2)
+        
+        # Build context
+        context = "📚 African Knowledge Reference:\n"
+        for i, result in enumerate(search_results, 1):
+            context += f"{i}. {result['question']}: {result['answer']}\n"
+        
+        # Build prompt
+        prompt = f"""<s>[INST] 🦁 You are SIMBA AI - the First African Large Language Model.
+
+You specialize in African languages, coding, mathematics, and African innovation.
+
+Use this knowledge:
+{context}
+
+Question: {message}
+
+Provide an accurate, helpful response that showcases African excellence. [/INST] 🦁 Simba AI:"""
+        
+        # Generate response
+        inputs = tokenizer(prompt, return_tensors="pt", truncation=True, max_length=1024).to(model.device)
+        
+        with torch.no_grad():
+            outputs = model.generate(
+                **inputs,
+                max_new_tokens=150,
+                temperature=0.7,
+                do_sample=True,
+                pad_token_id=tokenizer.eos_token_id,
+                eos_token_id=tokenizer.eos_token_id,
+            )
+        
+        full_response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+        
+        # Extract response
+        if "🦁 Simba AI:" in full_response:
+            response = full_response.split("🦁 Simba AI:")[-1].strip()
+        else:
+            response = full_response
+        
+        return response
+        
+    except Exception as e:
+        return f"🦁 Simba AI is currently learning... (Error: {str(e)})"
+
+# =====================
+# GRADIO INTERFACE
+# =====================
+
+# Custom CSS for African theme
+css = """
+.gradio-container {
+    font-family: 'Arial', sans-serif;
+}
+.header {
+    text-align: center;
+    padding: 20px;
+    background: linear-gradient(135deg, #ff7e5f, #feb47b);
+    color: white;
+    border-radius: 10px;
+    margin-bottom: 20px;
+}
+"""
+
+# Create chat interface
+with gr.Blocks(css=css, theme=gr.themes.Soft()) as demo:
+    
+    gr.HTML("""
+    <div class="header">
+        <h1>🦁 Simba AI - First African LLM</h1>
+        <h3>Specializing in African Languages, Coding & Mathematics</h3>
+        <p>Ask about Yoruba, Swahili, Igbo, Hausa, Python programming, math problems, and African innovation!</p>
+    </div>
+    """)
+    
+    chatbot = gr.Chatbot(
+        label="🦁 Chat with Simba AI",
+        height=500,
+        show_copy_button=True,
+        placeholder="Ask me anything about African languages, coding, or mathematics..."
+    )
+    
+    with gr.Row():
+        msg = gr.Textbox(
+            label="Your message",
+            placeholder="Type your question here...",
+            lines=2,
+            scale=4
+        )
+        send_btn = gr.Button("🚀 Ask Simba AI", variant="primary", scale=1)
+    
+    with gr.Row():
+        clear_btn = gr.Button("🧹 Clear Chat")
+    
+    # Examples
+    gr.Examples(
+        examples=[
+            "Write a Python function to add two numbers",
+            "How do you say hello in Yoruba?",
+            "What is 15 + 27?",
+            "Create a factorial function",
+            "Thank you in Swahili",
+            "Calculate 8 × 7",
+            "What is M-Pesa?"
+        ],
+        inputs=msg,
+        label="💡 Try these examples:"
+    )
+    
+    # Event handlers
+    def respond(message, chat_history):
+        bot_message = simba_ai_chat(message, chat_history)
+        chat_history.append((message, bot_message))
+        return "", chat_history
+    
+    msg.submit(respond, [msg, chatbot], [msg, chatbot])
+    send_btn.click(respond, [msg, chatbot], [msg, chatbot])
+    clear_btn.click(lambda: None, None, chatbot, queue=False)
+
+# =====================
+# LAUNCH
+# =====================
+
+if __name__ == "__main__":
+    demo.launch(debug=True)