Update notes

base/science-tech-maths/programming/languages/python/python.md

@@ -67,6 +67,7 @@
     - [Mutable default arguments](#mutable-default-arguments)
     - [lru\_cache or cache on class methods](#lru_cache-or-cache-on-class-methods)
   - [Timing code](#timing-code)
+  - [Caching code](#caching-code)
   - [Modules import time](#modules-import-time)
   - [Wheels](#wheels)
   - [Virtualenv](#virtualenv)
@@ -1259,6 +1260,16 @@ print(timeit.timeit(my_function, number=100000))
 
 Don't use `time.time()`! It's not monotonic! Use `time.monotonic_ns()` instead to measure a time difference.
 
+## Caching code
+
+On expensive functions, you can use `functools.lru_cache` to cache the result of the function for the same arguments:
+
+```python
+from pydantic import TypeAdapter
+
+TypeAdapter = lru_cache(TypeAdapter)
+```
+
 ## Modules import time
 
 ```bash