This model is a fine-tuned version of XLM Roberta Base on the Amazone Review Multi dataset. 
The model is trained om the English data and reviews about groceries. 
It achieves the following results:


Loss: 1.13
Mae: 0.56