updated llm

core_logic.py

@@ -10,8 +10,12 @@ from huggingface_hub import InferenceClient
 from tools import web_search, parse_file
 
 # Recommended: Qwen2.5-Coder-32B or Llama-3.1-70B-Instruct
-#client = InferenceClient("Qwen/Qwen2.5-Coder-32B-Instruct", token=os.getenv("HF_TOKEN"))
-client = InferenceClient("deepseek-ai/DeepSeek-V4-Pro", token=os.getenv("HF_TOKEN"))
+#client = InferenceClient("deepseek-ai/DeepSeek-V4-Pro", token=os.getenv("HF_TOKEN"))
+client = InferenceClient("Qwen/Qwen2.5-Coder-32B-Instruct", token=os.getenv("HF_TOKEN"))
+#client = InferenceClient("Qwen/Qwen2.5-Coder-7B-Instruct", token=os.getenv("HF_TOKEN"))
+#client = InferenceClient("llama-3.1-8b-instant", token=os.getenv("HF_TOKEN")) "llama-3.1-70b-versatile" -> GROQ API
+#client = InferenceClient("meta-llama/Llama-3.1-8B-Instruct", token=os.getenv("HF_TOKEN")) # Or "Qwen/Qwen2.5-72B-Instruct"
+
 
 SYSTEM_PROMPT = """
 You are the 'Silicon Architect'—a master-stroke creative genius in AI Engineering and Technical Architecture.