Update README.md

README.md

@@ -47,8 +47,8 @@ Be advised that this is a preview model meant to showcase the base model's capab
 
 #### Pre-training
 
-We pre-trained the model in two stages, first training on billions of tokens of mixed audio and text data using a next-token-prediction objective.
-Then, we trained on tens of thousands of hours of German and English speech mixed with a little text instruction data to preserve the text understanding capability of the model.
+We pre-trained the model in two stages, first training on billions of tokens of mixed speech and text data using a next-token-prediction objective.
+Then, we trained on tens of thousands of hours of German and English TTS data mixed with a little text instruction data to preserve the text understanding capability of the model.
 
 We used the following datasets, as well as some in-house datasets:
 - [HuggingFaceFW/fineweb](https://huggingface.co/datasets/HuggingFaceFW/fineweb)