Update README.md

README.md

@@ -32,7 +32,7 @@ model-index:
 
 This is the model for Irish ASR. It has been trained on the Common-voice dataset and living Irish audio dataset. The Common-voice code for the Irish language is ga-IE. From the Common voice dataset, all the Validated audio clips and all the living audio clips were taken into account and after a random train-test split, 90% percent of the total dataset (5156 utterances) were taken for training, and the rest of the 10% of real data (579 utterances) were taken for testing. 
 
-This dataset was finetuned on wav2vec2-large-xls-r-300m. On the testing dataset, 25.96% of WER could be achieved. 
+This dataset was finetuned on wav2vec2-large-xls-r-300m. On the testing dataset, 25.94% of WER could be achieved. 
 
 ### How to use
 Example of transcribing the Common Voice audio clip from the invalidated dataset, using GPU if available. The model expects 16kHz audio.