Update README.md

BERT model pretrained from scratch on the brabant-xvii dataset.
Pretraining was done for both the MLM and NSP objectives (at once).

README.md

@@ -1,3 +1,9 @@
 ---
 license: mit
----
+pipeline_tag: fill-mask
+datasets:
+- arch-be/brabant-xvii
+language:
+- nl
+- fr
+---