Nit (notebook lint test) in the Better performance with tf.function guide

Author: 8bitmp3

site/en/guide/function.ipynb

@@ -272,7 +272,7 @@
         "id": "K7scSzLx662f"
       },
       "source": [
-        "When we pass arguments of different types into a `Function`, both stages are run:\n"
+        "When you pass arguments of different types into a `Function`, both stages are run:\n"
       ]
     },
     {
@@ -1301,7 +1301,7 @@
         "\n",
         "`Function` creates a new `ConcreteFunction` when called with a new value of a Python argument. However, it does not do that for the Python closure, globals, or nonlocals of that `Function`. If their value changes in between calls to the `Function`, the `Function` will still use the values they had when it was traced. This is different from how regular Python functions work.\n",
         "\n",
-        "For that reason, we recommend a functional programming style that uses arguments instead of closing over outer names."
+        "For that reason, you should follow a functional programming style that uses arguments instead of closing over outer names."
       ]
     },
     {
@@ -1439,7 +1439,7 @@
       "source": [
         "Using the same `Function` to evaluate the updated instance of the model will be buggy since the updated model has the [same cache key](#rules_of_tracing) as the original model.\n",
         "\n",
-        "For that reason, we recommend that you write your `Function` to avoid depending on mutable object attributes or create new objects.\n",
+        "For that reason, you're recommended to write your `Function` to avoid depending on mutable object attributes or create new objects.\n",
         "\n",
         "If that is not possible, one workaround is to make new `Function`s each time you modify your object to force retracing:"
       ]