Create README.md

README.md

@@ -0,0 +1,11 @@
+Pretraining KoLD Dataset with pretrained "koelectra-v3" model.
+
+dataset : https://github.com/boychaboy/KOLD
+
+pretrained_model : https://huggingface.co/monologg/koelectra-base-v3-discriminator
+
+So you should use tokenizer with "koelectra-base-v3-discriminator".
+
+label maps are like
+>
+    {0: "not_hate_speech", 1: "hate_speech"}