docs: communicate sample rate in card

Signed-off-by: WoodenMaiden <yann.pomie@laposte.net>

README.md

@@ -7,5 +7,5 @@ size_categories:
 - n<1K
 ---
 
-This contains training data for the [do-2021/activator-keyword-detector](https://huggingface.co/do-2021/activator-keyword-detector) with male models in a split format.
-Audio samples in the "yes" subdirectory would activate the assistant, and those in "no" should not.
+This contains training data for the [do-2021/activator-keyword-detector](https://huggingface.co/do-2021/activator-keyword-detector) with male voices, with a sample rate of 32000HZ, in a split format.
+Audio samples in the "yes" subdirectory would activate the assistant, and those in "no" should not.