Update README.md

README.md

@@ -45,9 +45,8 @@ A typical data point comprises the path to the audio file, called audio and its
 - sentence: The sentence the user was prompted to speak
 
 ### Data Splits
-The speech material has been subdivided into portions for train and test.
-The train split consists of [TODO] audio clips and the related sentences.
-The test split consists of [TODO] audio clips and the related sentences.
+The speech material has not been subdivided into portions, everything is in the "train" split.
+The train split consists of 82825 audio clips and the related sentences.
 
 
 ### Contributions