fix documentation format (#60)

docs/opt_tutorial.rst

@@ -153,8 +153,9 @@ Conclusion
 ----------
 
 Optimizing code with ``torch.compile`` involves:
+
 1. Using ``depyf`` to decompile the bytecode and understand the performance bottlenecks.
 2. Identifying and addressing unnecessary checks or other sources of overhead.
 3. Using ``depyf`` to directly call the compiled function, bypassing unnecessary steps where appropriate.
 
-By following these steps, you can significantly improve performance, especially in environments where execution time is critical.
+By following these steps, we can significantly improve performance, especially in environments where execution time is critical.