Update README.md

README.md

@@ -1,30 +1,30 @@
----
-license: apache-2.0
-dataset_info:
-  features:
-  - name: anchor
-    dtype: string
-  - name: positive
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 10784472
-    num_examples: 10136
-  - name: test
-    num_bytes: 2652944
-    num_examples: 2534
-  download_size: 5985227
-  dataset_size: 13437416
-configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
-  - split: test
-    path: data/test-*
-language:
-- en
----
+---
+license: apache-2.0
+dataset_info:
+  features:
+  - name: anchor
+    dtype: string
+  - name: positive
+    dtype: string
+  splits:
+  - name: train
+    num_bytes: 10784472
+    num_examples: 10136
+  - name: test
+    num_bytes: 2652944
+    num_examples: 2534
+  download_size: 5985227
+  dataset_size: 13437416
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
+  - split: test
+    path: data/test-*
+language:
+- en
+---
 
 # Dataset Card for Dataset Name
 
@@ -35,7 +35,7 @@ Dataset for finetuning embedding models for patent language.
 ## Dataset Details
 <!-- Provide a longer summary of what this dataset is. -->
 
-Anchor/Positive pairs extracted from `text` and `text_b` columns in the ultrabalanced version of the [PatentMatch dataset](https://hpi.de/naumann/projects/web-science/paar-patent-analysis-and-retrieval/patentmatch.html)
+Anchor/Positive pairs extracted from `text` and `text_b` columns with `label=1`in the ultrabalanced version of the [PatentMatch dataset](https://hpi.de/naumann/projects/web-science/paar-patent-analysis-and-retrieval/patentmatch.html)
 
 <!-- Provide the basic links for the dataset. -->