Zero-shot wer plot added

README.md

@@ -101,6 +101,7 @@ To evaluate generalizability, the model was evaluated against multiple open-sour
 **OBS!** The vocab used for training incudes numerals (0,1,2,..,9), which are translated to text in a post-processing step. If the model misses spaces the numbers are interpreted as one, which especially affects the NST score as this dataset contains many numerals.
 
 <img src="https://huggingface.co/CoRal-project/roest-wav2vec2-2B-v2/resolve/main/images/cer_comparison_zero-shot_roest.png">
+<img src="https://huggingface.co/CoRal-project/roest-wav2vec2-2B-v2/resolve/main/images/wer_comparison_zero-shot_roest.png">
 
 
 ### Conversational CoRal-v2 Performance