Upload dataset

README.md

@@ -21,13 +21,18 @@ dataset_info:
   - name: train
     num_bytes: 10485274
     num_examples: 12500
-  download_size: 5296023
-  dataset_size: 10485274
+  - name: test_500
+    num_bytes: 419353
+    num_examples: 500
+  download_size: 5514764
+  dataset_size: 10904627
 configs:
 - config_name: default
   data_files:
   - split: train
     path: data/train-*
+  - split: test_500
+    path: data/test_500-*
 ---
 
 Difficulty is estimated by the number of correct responses from 10 samples per question using [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct).

data/test_500-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7588d4569a885ee3395eab28418688c31c5bb8933a52967e345a21cf528c88d2
+size 218741