more stuff

app.py

@@ -4,10 +4,9 @@ import json
 import os
 
 # Initialize the client
-# We use a powerful instruction-tuned model. Mistral-7B-Instruct-v0.3 is excellent for reasoning.
-# If that fails, we can fallback to others, but let's start strong.
+# Switching to Qwen/Qwen2.5-72B-Instruct as requested. It is SOTA and supported for chat.
 token = os.getenv("HUGGING_FACE_HUB_TOKEN")
-client = InferenceClient("mistralai/Mistral-7B-Instruct-v0.3", token=token)
+client = InferenceClient("Qwen/Qwen2.5-72B-Instruct", token=token)
 
 SYSTEM_PROMPT = """You are ModelForge, an elite AI Solutions Architect with deep expertise in MLOps, research-level machine learning, and software engineering.
 
@@ -88,7 +87,7 @@ def generate_solution(description):
     ]
     
     try:
-        # Use chat_completion for conversational models (Zephyr, Mistral, Llama)
+        # Use chat_completion for conversational models
         response = client.chat_completion(
             messages,
             max_tokens=2500,
@@ -118,7 +117,7 @@ demo = gr.Interface(
     inputs=gr.Textbox(lines=5, placeholder="Describe your complex ML problem..."),
     outputs=gr.JSON(label="Recommendations"),
     title="ModelForge AI Backend v2",
-    description="Advanced ML Solutions Architect (Powered by Mistral-7B-Instruct)"
+    description="Advanced ML Solutions Architect (Powered by Qwen2.5-72B-Instruct)"
 )
 
 if __name__ == "__main__":