We provide here examples of training metrics for a gpt2 style model trained on a multilingual dataset from the paper
