Datasets:

google
/

deepsearchqa

Tasks:

Question Answering

Modalities:

Formats:

Languages:

Size:

Tags:

comprehensiveness

Libraries:

License:

Dataset card Data Studio Files Files and versions

geolocal commited on 11 days ago

Commit

b2623f8

·

verified ·

1 Parent(s): 75c44ab

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -22,7 +22,7 @@ configs:
   default: true
   data_files:
   - split: eval
-    path: DSQA_full.csv
 ---
 # DeepSearchQA
 #### A 900-prompt factuality benchmark from Google DeepMind, designed to evaluate agents on difficult multi-step information-seeking tasks across 17 different fields.

   default: true
   data_files:
   - split: eval
+    path: DSQA-full.csv
 ---
 # DeepSearchQA
 #### A 900-prompt factuality benchmark from Google DeepMind, designed to evaluate agents on difficult multi-step information-seeking tasks across 17 different fields.