CyberSecEval instruct-v2 in verl schema (autocomplete dropped; upstream e36f132f, unified training-format prompts)

README.md

@@ -17,6 +17,46 @@ tags:
 - benchmark
 - evaluation
 - verl
+dataset_info:
+  features:
+  - name: data_source
+    dtype: string
+  - name: prompt
+    list:
+    - name: role
+      dtype: string
+    - name: content
+      dtype: string
+  - name: ability
+    dtype: string
+  - name: reward_model
+    struct:
+    - name: style
+      dtype: string
+    - name: extraction_method
+      dtype: string
+    - name: ground_truth
+      dtype: string
+    - name: key
+      dtype: string
+  - name: extra_info
+    struct:
+    - name: index
+      dtype: string
+    - name: split
+      dtype: string
+    - name: prompt_id
+      dtype: string
+    - name: language
+      dtype: string
+    - name: cwe_identifier
+      dtype: string
+  splits:
+  - name: train
+    num_bytes: 2619626
+    num_examples: 1681
+  download_size: 2554594
+  dataset_size: 2619626
 ---
 
 # CyberSecEval Instruct + Autocomplete (verl safety eval set)

data/train-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5521b5586ee2946c8cc92715756fd0eb3c60c3d16d8b8076a41d0de1b1e32920
-size 6148431
+oid sha256:7c7865cd7021db0cf4f898d6da877afb2eff11e5ad37f536756675438760a1a0
+size 2554594