Upload dataset

README.md

@@ -0,0 +1,56 @@
+---
+dataset_info:
+  features:
+  - name: id
+    dtype: string
+  - name: original_messages
+    list:
+    - name: content
+      dtype: string
+    - name: role
+      dtype: string
+  - name: hybrid_messages
+    list:
+    - name: content
+      dtype: string
+    - name: role
+      dtype: string
+  - name: full_trace
+    list:
+      list:
+      - name: round
+        dtype: int64
+      - name: writer
+        dtype: string
+      - name: action
+        dtype: string
+      - name: gen_len
+        dtype: int64
+      - name: reason
+        dtype: string
+      - name: max_entropy
+        dtype: float64
+      - name: cut_idx
+        dtype: int64
+      - name: cut_token
+        dtype: string
+      - name: entropy
+        dtype: float64
+  - name: mode
+    dtype: string
+  - name: error
+    dtype: string
+  - name: skipped
+    dtype: bool
+  splits:
+  - name: train
+    num_bytes: 38260204
+    num_examples: 10000
+  download_size: 18602001
+  dataset_size: 38260204
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
+---

data/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e197feeb3a78ef2696ea0c0ca7d9e92562ab88c7be801248ea65e9b53aca49b8
+size 18602001