diff --git a/e2e-challenge/README.md b/e2e-challenge/README.md
index f22b84f..a856e64 100644
--- a/e2e-challenge/README.md
+++ b/e2e-challenge/README.md
@@ -33,7 +33,7 @@ To train and evaluate the baseline, you need to:
    the setup that was best on the development data
 
 ```
-../run_tgen.py seq2seq_train config/seq2seq.py \
+../run_tgen.py seq2seq_train config/config.yaml \
     input/train-das.txt input/train-text.txt \
     model.pickle.gz
 ```