Update README.md

README.md

@@ -45,6 +45,8 @@ The model performs exceptionally well on writing, explanation and discussion tas
     "average": 6.8875
 }
 ```
+Have a look at some examples [in this Google Doc](https://docs.google.com/document/d/1SAAikkPAF4oLoFISqE0P1mRL5OUk8l2pI90zZC4bP1E/edit?usp=sharing).
+
 
 ## Model Details
 
@@ -114,6 +116,9 @@ in advance, and the model may in some instances produce inaccurate, biased or ot
 to user prompts. Therefore, before deploying any applications of `LeoLM/leo-hessianai-70b-chat`, developers should
 perform safety testing and tuning tailored to their specific applications of the model.
 
+We are aware of the model refusing to answer more often than desired. This will be adressed in future versions. For now, the training
+dataset is equal to that used for our smaller chat variants.
+
 Please see Meta's [Responsible Use Guide](https://ai.meta.com/llama/responsible-use-guide/).
 
 ## Finetuning Details