Help With MLRecommender / CreateML Recommendation

So, I'm playing around a little bit with CreateML and MLRecommender and I'm having a hard time figuring out how to create and train my model. I've got a very simple data set for experimenation but the tools keep kicking things back. Clearly I'm missing something.

Here's my data set, movie reviews in CSV format:

user_id,movie_id,rating_id
1,Inception,4
1,Interstellar,5
2,Titanic,4
3,Titanic,1

With CreateML, I've got a recommender for that file with 'Users' mapped to 'user_id', 'Items' mapped to 'movie_id', 'Ratings' mapped to 'rating_id'.

When I ask it to train, I get a training error:

"Number of string item ids specified (3) does not equal the number of items given (2)"

Hmmmm, not sure what that means. Can anyone help? I feel like it must be something really obvious tht I'm missing but I've tried a bunch of variations without any luck.

Any help appreciated.