Update app.py

app.py

@@ -1,62 +1,47 @@
-import os
 import gradio as gr
-from duckduckgo_search import DDGS
-from huggingface_hub import InferenceClient
+import torch
+from transformers import pipeline
 
-# 🔱 HF Token එක ලබා ගැනීම
-HF_TOKEN = os.getenv("HF_TOKEN")
+# Model configuration
+# Note: Use "google/gemma-2-2b-it" as it is highly efficient for Free Tier Spaces
+model_id = "google/gemma-2-2b-it"
 
-# 🔱 Provider කෙනෙක්ව සෘජුවම තෝරා ගැනීම
-# මෙහිදී 'provider=' කොටස මඟින් සෘජුවම අදාළ Server එකට Request එක යොමු කරයි.
-client = InferenceClient(
-    token=HF_TOKEN
+# Specialist Inference Engine initialization
+pipe = pipeline(
+    "text-generation",
+    model=model_id,
+    model_kwargs={"torch_dtype": torch.bfloat16},
+    device_map="auto",
 )
 
-def web_search(query):
-    try:
-        with DDGS() as ddgs:
-            results = ddgs.text(query, max_results=3)
-            search_data = ""
-            if results:
-                for r in results:
-                    search_data += f"\nSource: {r['title']}\nContent: {r['body']}\n"
-                return search_data
-            return "No web results found."
-    except Exception:
-        return "Web search unavailable."
-
-def inachi_ai_response(user_input, history):
-    context = web_search(user_input)
+def specialist_respond(message, history):
+    # System prompt to maintain identity
+    system_prompt = "You are a highly advanced AI assistant developed under the INACHI AI project. Be precise and technical."
+    
+    # Building the conversation format for Gemma
+    messages = [{"role": "system", "content": system_prompt}]
+    for val in history:
+        if val[0]: messages.append({"role": "user", "content": val[0]})
+        if val[1]: messages.append({"role": "assistant", "content": val[1]})
+    
+    messages.append({"role": "user", "content": message})
     
-    messages = [
-        {"role": "system", "content": "You are INACHI-AI, a professional assistant. Use web context to help the user."},
-        {"role": "user", "content": f"Context: {context}\n\nQuestion: {user_input}"}
-    ]
+    # Generating the response
+    prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
+    outputs = pipe(prompt, max_new_tokens=512, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
     
-    response = ""
-    try:
-        # 🔱 මෙතනදී තමයි Model එක සහ Provider එක සෘජුවම සම්බන්ධ කරන්නේ
-        # උදා: Novita පාවිච්චි කිරීමට 'provider="novita"' ලෙස එක් කරන්න
-        for message in client.chat_completion(
-            model="google/gemma-4-E2B",
-            messages=messages,
-            max_tokens=1024,
-            stream=True,
-            # provider="novita" # <--- ඔන්න මෙතනට Provider දාන්න පුළුවන්
-        ):
-            token = message.choices[0].delta.content
-            if token:
-                response += token
-                yield response
-    except Exception as e:
-        yield f"Error Specialist: {str(e)}"
+    return outputs[0]["generated_text"][len(prompt):]
 
-# 🔱 Gradio Interface
-demo = gr.ChatInterface(
-    fn=inachi_ai_response,
-    title="INACHI-CORE (Provider-Locked Edition)",
-    description="Specialist, මම දැන් සෘජු Provider සම්බන්ධතාවයක් හරහා ක්‍රියාත්මක වෙමි."
-)
+# 🔱 INACHI AI UI Design
+with gr.Blocks(theme=gr.themes.Soft()) as demo:
+    gr.Markdown("""# 🔱 INACHI AI: SPECIALIST INTERFACE
+    ### Powered by Gemma 2B | Authorized Access: MINZO-PRIME""")
+    
+    chatbot = gr.ChatInterface(
+        fn=specialist_respond,
+        title="INACHI-GEMMA V1",
+        description="Advanced neural processing unit for system research and development.",
+    )
 
 if __name__ == "__main__":
     demo.launch()