add arxiv paper

README.md

@@ -21,7 +21,7 @@ co2_eq_emissions:
       emissions: 23660
 ---
 
-`dant5-small` is a 60M parameter model with architecture identical to `t5-small`. It was trained for 10 epochs on the Danigh GigaWord Corpus ([official website](https://gigaword.dk), [paper](https://aclanthology.org/2021.nodalida-main.46/)).
+`dant5-small` is a 60M parameter model with architecture identical to `t5-small`. Training details are given in the paper [Training a T5 Using Lab-sized Resources](https://arxiv.org/abs/2208.12097).  It was trained for 10 epochs on the Danigh GigaWord Corpus ([official website](https://gigaword.dk), [paper](https://aclanthology.org/2021.nodalida-main.46/)).
 
 ## To use the model