Update README.md

README.md

@@ -41,6 +41,14 @@ configs:
     path: data/shots-*
   - split: test
     path: data/test-*
+license: cc-by-4.0
+task_categories:
+- visual-question-answering
+language:
+- ru
+pretty_name: ruCLEVR
+size_categories:
+- 1K<n<10K
 ---
 
 
@@ -136,4 +144,4 @@ To create RuCLEVR, we used two strategies: 1) generation of the new samples and
 
 Metrics for aggregated evaluation of responses:
 
-- `Exact match`: Exact match is the average of scores for all processed cases, where a given case score is 1 if the predicted string is the exact same as its reference string, and is 0 otherwise.
+- `Exact match`: Exact match is the average of scores for all processed cases, where a given case score is 1 if the predicted string is the exact same as its reference string, and is 0 otherwise.