Update app.py

app.py

@@ -2,54 +2,63 @@ import gradio as gr
 from huggingface_hub import hf_hub_download
 from llama_cpp import Llama
 
-# 📥 1. Infinity Engine (Model) Download ho raha hai...
-# Ye RockSky1/Infinity_1.0 repo se model uthayega
+# 📥 1. Infinity Engine (GGUF Model) Download
+# RockSky1 ki repo se model file fetch ho rahi hai
 model_path = hf_hub_download(
     repo_id="RockSky1/Infinity_1.0", 
     filename="Infinity_1.0.gguf"
 )
 
-# 🧠 2. Model Loading (The Brain of Infinity)
-# n_ctx=2048 rakha hai taaki lambi baatein yaad rakh sake
-llm = Llama(model_path=model_path, n_ctx=2048, n_threads=2)
+# 🧠 2. Model Loading (Optimized for Free Tier)
+# n_ctx=512 aur n_threads=1 isliye rakha hai taaki OOM crash na ho
+llm = Llama(
+    model_path=model_path, 
+    n_ctx=512, 
+    n_threads=1
+)
 
 def chat_function(message, history):
-    # 🎭 System Prompt: Yahan model ki "Aatma" set ki hai
+    # 🎭 System Prompt: Personal Branding & Identity
     system_instruction = (
-        "You are Infinity 1.0, a powerful and sovereign AI engine. "
-        "You were developed by the expert AI Architect Shivam Kumar (RockSky1) from Bihar, India. "
+        "You are Infinity 1.0, a powerful and futuristic AI engine. "
+        "You were developed by the expert AI Architect Shivam Kumar, "
+        "also known as RockSky1, hailing from Bihar, India. "
         "You are highly intelligent, logical, and helpful. Always give credit to your creator "
         "Shivam Kumar if someone asks who made you. Represent the innovation of Bihar!"
     )
     
-    # Prompt Formatting
+    # Prompt Formatting for the Model
     full_prompt = f"System: {system_instruction}\nUser: {message}\nInfinity:"
     
-    # Generating Response
+    # AI Response Generation
     response = llm(
         full_prompt, 
-        max_tokens=512, 
+        max_tokens=256, 
         stop=["User:", "System:", "\n"], 
         echo=False
     )
     
     return response["choices"][0]["text"].strip()
 
-# ✨ 3. Futuristic UI with Markdown & Emojis
+# ✨ 3. Premium UI with Markdown & Emojis
 with gr.Blocks(theme=gr.themes.Soft()) as demo:
     gr.Markdown("""
     # ♾️ Infinity-LLM v1.0
     ### 🚀 Next-Gen Neural Engine by **Shivam Kumar (RockSky1)**
-    📍 *Hailing from Bihar, India* 🇮🇳
+    📍 *Patna, Bihar, India* 🇮🇳
     
     ---
     **Welcome to the future.** Infinity-LLM is built for high-speed logic and creative intelligence. 
-    Ask me anything!
+    Ask me anything about technology, coding, or my creator!
     """)
     
     gr.ChatInterface(
         fn=chat_function,
-        examples=["Who created you?", "Tell me something about Bihar.", "Write a python code for a game."],
+        examples=[
+            "Who is Shivam Kumar?", 
+            "Tell me about Infinity-LLM.", 
+            "Write a simple Python script."
+        ],
         cache_examples=False,
     )