Upload README.md with huggingface_hub

README.md

@@ -6,62 +6,6 @@ tags:
 - preference
 - human-feedback
 - rlhf
-dataset_info:
-  features:
-  - name: prompt_id
-    dtype: string
-  - name: annotator_id
-    dtype: string
-  - name: preference_type
-    dtype: string
-  - name: ranking
-    dtype: string
-  - name: rank_gap
-    dtype: int64
-  - name: chosen_label
-    dtype: string
-  - name: rejected_label
-    dtype: string
-  - name: prompt
-    list:
-    - name: role
-      dtype: string
-    - name: content
-      dtype: string
-  - name: chosen
-    list:
-    - name: role
-      dtype: string
-    - name: content
-      dtype: string
-  - name: rejected
-    list:
-    - name: role
-      dtype: string
-    - name: content
-      dtype: string
-  - name: prompt_text
-    dtype: string
-  - name: chosen_text
-    dtype: string
-  - name: rejected_text
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 136028105
-    num_examples: 44311
-  - name: test
-    num_bytes: 13744661
-    num_examples: 4923
-  download_size: 76557161
-  dataset_size: 149772766
-configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
-  - split: test
-    path: data/test-*
 ---
 
 # sumuks/coval-ha
@@ -87,8 +31,15 @@ tie between `C` and `D`.
 ## Build Summary
 
 - Rows: 49,234
+- Train rows: 44,311
+- Test rows: 4,923
+- Train prompts: 969
+- Test prompts: 108
+- Prompt overlap between train and test: 0
 - Source dataset: `openai/coval` / `comparisons`
 - Minimum rank-group gap: `2`
+- Test split target: `0.1`
+- Split seed: `42`
 
 ### Preference Types