update testing requirements

Author: AakashKumarNain

tensorflow_addons/metrics/README.md

@@ -15,14 +15,13 @@
 #### Standard API
 In order to conform with the current API standard, all metrics
 must:
- * Be a `tf.function`.
- * Have the signature `fn(input, axis=-1, name=None)`.
+ * Inherit from `tf.metrics.Metric`.
  * [Register as a keras global object](https://github.com/tensorflow/addons/blob/master/tensorflow_addons/utils/python/keras_utils.py)
   so it can be serialized properly.
  * Add the addon to the `py_library` in this sub-package's BUILD file.
 
 #### Testing Requirements
- * Simple unittests that demonstrate the layer is behaving as expected.
+ * Simple unittests that demonstrate the metric is behaving as expected.
  * When applicable, run all unittests with TensorFlow's
    `@run_in_graph_and_eager_modes` (for test method)
    or `run_all_in_graph_and_eager_modes` (for TestCase subclass)