Push beir_scidocs (1000 items, 1 splits) from Langfuse

README.md

@@ -1,10 +1,37 @@
 ---
-tags: ["benchmark", "beir", "scidocs", "retrieval"]
+tags:
+- benchmark
+- beir
+- scidocs
+- retrieval
 task_categories:
-  - question-answering
-  - text-retrieval
+- question-answering
+- text-retrieval
 size_categories:
-  - 1K<n<10K
+- 1K<n<10K
+dataset_info:
+  features:
+  - name: id
+    dtype: string
+  - name: input
+    dtype: string
+  - name: expected_output
+    dtype: string
+  - name: metadata.query_id
+    dtype: string
+  - name: metadata.split
+    dtype: string
+  splits:
+  - name: test
+    num_bytes: 487113
+    num_examples: 1000
+  download_size: 326994
+  dataset_size: 487113
+configs:
+- config_name: default
+  data_files:
+  - split: test
+    path: data/test-*
 ---
 
 # beir_scidocs