Upload dataset

README.md

@@ -13,17 +13,19 @@ dataset_info:
     dtype: int64
   - name: answer_prefix
     dtype: string
+  - name: token_position_answer
+    dtype: int64
   - name: question_id
     dtype: int64
   splits:
-  - name: 16k_qa_hotpotqa
-    num_bytes: 33089358
+  - name: 32k_niah_single_noise_words_numbers
+    num_bytes: 61231446
     num_examples: 500
-  download_size: 20740312
-  dataset_size: 33089358
+  download_size: 3087127
+  dataset_size: 61231446
 configs:
 - config_name: default
   data_files:
-  - split: 16k_qa_hotpotqa
-    path: data/16k_qa_hotpotqa-*
+  - split: 32k_niah_single_noise_words_numbers
+    path: data/32k_niah_single_noise_words_numbers-*
 ---

data/32k_niah_single_noise_words_numbers-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:668ace161565e5dd75b18f6e5432c95471d4b3025ad8d822d36cec37ec240350
+size 3087127