Update app.py

app.py

@@ -3,37 +3,44 @@ import threading
 import asyncio
 import discord
 from dotenv import load_dotenv
-from llama_cpp import Llama  # Library for GGUF models
+from huggingface_hub import hf_hub_download
+from llama_cpp import Llama
 
-# Load environment variables (set these via Hugging Face Secrets)
+# Load environment variables (from Hugging Face Secrets and .env if available)
 load_dotenv()
 DISCORD_TOKEN = os.getenv("DISCORD_TOKEN")
 
 if not DISCORD_TOKEN:
     raise ValueError("Discord bot token is missing. Set DISCORD_TOKEN in the environment variables.")
 
-# Set the local path to your quantized model file.
-# Ensure that this file (e.g. DeepScaleR-1.5B-Preview-Q8_0.gguf) is uploaded to your repository.
-MODEL_PATH = "./DeepScaleR-1.5B-Preview-Q8_0.gguf"
+# Set model details
+MODEL_REPO = "bartowski/agentica-org_DeepScaleR-1.5B-Preview-GGUF"
+MODEL_FILENAME = "agentica-org_DeepScaleR-1.5B-Preview-Q8_0.gguf"
+MODEL_PATH = f"./{MODEL_FILENAME}"
 
-# Initialize the model with appropriate settings.
-# Adjust n_threads and other parameters as needed.
+# If the model file doesn't exist locally, download it.
+if not os.path.exists(MODEL_PATH):
+    print("Model file not found, downloading...")
+    MODEL_PATH = hf_hub_download(repo_id=MODEL_REPO, filename=MODEL_FILENAME)
+    print(f"Model downloaded to {MODEL_PATH}")
+
+# Initialize the model using llama-cpp-python.
+# Adjust n_threads if needed.
 llm = Llama(model_path=MODEL_PATH, n_threads=4)
 
+# Define a function to generate a response.
 def generate_response(prompt):
-    # Generate text using llama-cpp's Llama instance.
-    # Adjust parameters (max_tokens, temperature, top_p) for speed/quality tradeoffs.
     output = llm(prompt=prompt, max_tokens=200, temperature=0.7, top_p=0.9, echo=False)
     response = output["text"]
-    # Optionally enforce your bot identity:
+    # Replace the internal model name with your bot's identity.
     response = response.replace("DeepScaleR", "Shiv Yantra AI")
     return response
 
-# ----------------------------
+# ---------------------------
 # Discord Bot Setup
-# ----------------------------
+# ---------------------------
 intents = discord.Intents.default()
-intents.message_content = True  # Enable reading message content
+intents.message_content = True  # Enable access to message content
 client = discord.Client(intents=intents)
 
 @client.event
@@ -49,7 +56,7 @@ async def on_message(message):
     user_input = message.content.strip()
     if user_input:
         try:
-            # Run the generate_response function in a separate thread so as not to block Discord's event loop.
+            # Run generate_response in a separate thread so as not to block the Discord loop.
             ai_response = await asyncio.to_thread(generate_response, user_input)
         except Exception as e:
             print(f"Error during generation: {e}")
@@ -59,11 +66,11 @@ async def on_message(message):
 def run_discord_bot():
     client.run(DISCORD_TOKEN)
 
-# ----------------------------
+# ---------------------------
 # Start the Discord Bot
-# ----------------------------
+# ---------------------------
 if __name__ == "__main__":
-    # Start the Discord bot in a separate daemon thread.
+    # Run the Discord bot on a separate thread.
     threading.Thread(target=run_discord_bot, daemon=True).start()
     
     # Keep the main thread alive.