Add train20 train dataset

README.md

@@ -289,6 +289,30 @@ dataset_info:
     num_examples: 46908
   download_size: 7770000
   dataset_size: 19684539
+- config_name: train20
+  features:
+  - name: id
+    dtype: string
+  - name: images
+    list:
+    - name: bytes
+      dtype: binary
+    - name: path
+      dtype: string
+  - name: conversation
+    list:
+    - name: role
+      dtype: string
+    - name: content
+      dtype: string
+  - name: source
+    dtype: string
+  splits:
+  - name: train
+    num_bytes: 19759826
+    num_examples: 46890
+  download_size: 7742268
+  dataset_size: 19759826
 - config_name: train4
   features:
   - name: id
@@ -482,6 +506,10 @@ configs:
   data_files:
   - split: train
     path: train2/train-*
+- config_name: train20
+  data_files:
+  - split: train
+    path: train20/train-*
 - config_name: train4
   data_files:
   - split: train

train20/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6df3cbb1472a89ba126e6c331c20e4c89bc257ba389ef6a5bfb5bd13fd01ab18
+size 7742268