Added limitations

README.md

@@ -81,4 +81,9 @@ elif label == "LABEL_0":
 print(f"Comment includes mention {has_mention}")
 ```
 
+## Limitations
+
+Clearly, the amount of training data was to small for a state of the art result. This can be seen in the evaluation chapter. Future rounds of retraining have to be performed. For the sake of completeness we publish this model here within [the projects documentation](https://interaktiv.br.de/ai-detect-newsroom-mentions-in-comments/).
+
+An analysis of possible biases reproduced by the present model, regardless of whether they originate from our finetuning or the underlying gpt2 model, is beyond the scope of this work. We assume that biases exist within the model and an analysis will be a task for future work