upload enhanced dataset

max_score: 0.8, relative_margin: 0.05, absolute_margin: 0.025;
model: sentence-transformers/all-mpnet-base-v2

3HN-biencoder2-lm/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b8ea8106cc7d12282a5f66c4e1cd495fce880583a8cd4c7d682fae2a6f430e52
+size 167705714

README.md

@@ -18,6 +18,24 @@ dataset_info:
     num_examples: 50089
   download_size: 43080341
   dataset_size: 72178253
+- config_name: 3HN-biencoder2-lm
+  features:
+  - name: query
+    dtype: string
+  - name: answer
+    dtype: string
+  - name: negative_1
+    dtype: string
+  - name: negative_2
+    dtype: string
+  - name: negative_3
+    dtype: string
+  splits:
+  - name: train
+    num_bytes: 257668043
+    num_examples: 99241
+  download_size: 167705714
+  dataset_size: 257668043
 - config_name: biencoder
   features:
   - name: query
@@ -67,6 +85,10 @@ configs:
   data_files:
   - split: train
     path: 3HN-biencoder1-lm/train-*
+- config_name: 3HN-biencoder2-lm
+  data_files:
+  - split: train
+    path: 3HN-biencoder2-lm/train-*
 - config_name: biencoder
   data_files:
   - split: train