add tweeteval

dataset_stats.py

@@ -16,7 +16,7 @@ if __name__ == '__main__':
             'word/validation': mean([count_word(k['text']) for k in _data['validation']]),
             'word/test': mean([count_word(k['text']) for k in _data['test']]),
             'word/train': mean([count_word(k['text']) for k in _data['train']]),
-            'instance/dev': len(_data['validation']),
+            'instance/validation': len(_data['validation']),
             'instance/test': len(_data['test']),
             'instance/train': len(_data['train'])
         }

stats.csv

@@ -2,6 +2,6 @@
 word/validation,32,40,502,26,32,20,132,13,24,15
 word/test,32,42,612,26,32,19,131,14,21,15
 word/train,31,42,536,26,32,19,133,13,20,16
-instance/dev,2427,114,64,30212,455,666,2001,955,1000,374
+instance/validation,2427,114,64,30212,455,666,2001,955,1000,374
 instance/test,3469,139,65,30135,974,1334,2001,784,2970,1421
 instance/train,4169,1688,516,500,3219,8000,6001,2862,9000,3257