Update README.md

README.md

@@ -96,4 +96,3 @@ The model achieves the following results without any fine-tuning (zero-shot):
 
 To get these results, we used the Eleuther AI evaluation harness [here](https://github.com/EleutherAI/lm-evaluation-harness),
 which can produce results different than those reported in the GPT2 paper. The p-values come from the stderr from the evaluation harness, plus a normal distribution assumption.
-We chose these 20 tasks, because they are the tasks that the GPT2 and GPT3 papers report results for.