Add train4 train dataset

README.md

@@ -145,6 +145,30 @@ dataset_info:
     num_examples: 46908
   download_size: 7770000
   dataset_size: 19684539
+- config_name: train4
+  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: 19685028
+    num_examples: 46908
+  download_size: 7754432
+  dataset_size: 19685028
 configs:
 - config_name: train0
   data_files:
@@ -170,4 +194,8 @@ configs:
   data_files:
   - split: train
     path: train2/train-*
+- config_name: train4
+  data_files:
+  - split: train
+    path: train4/train-*
 ---

train4/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:58ed6ffd1457b720aede619d882c85d5c443e51bf3cc1c9cf86428be4d5522bf
+size 7754432