ahkamboh commited on
Commit
045363d
·
verified ·
1 Parent(s): eb6136a

Link the test questions dataset, search-reader-bench

Browse files
Files changed (1) hide show
  1. README.md +3 -0
README.md CHANGED
@@ -183,6 +183,9 @@ strings from questions and results if they could ever contain them.
183
 
184
  Full tables, the test questions and the grading rules are in [`eval/results.md`](eval/results.md).
185
 
 
 
 
186
  v3.1a (this release), PyTorch fp32 on CPU:
187
 
188
  | question set | questions | accuracy |
 
183
 
184
  Full tables, the test questions and the grading rules are in [`eval/results.md`](eval/results.md).
185
 
186
+ The 258 test questions are also a dataset, [search-reader-bench](https://huggingface.co/datasets/ahkamboh/search-reader-bench):
187
+ gold answers, this model's answers and the grader, so you can grade your own search system the same way.
188
+
189
  v3.1a (this release), PyTorch fp32 on CPU:
190
 
191
  | question set | questions | accuracy |