Add train3 train dataset

README.md

@@ -121,6 +121,30 @@ dataset_info:
     num_examples: 46908
   download_size: 7770000
   dataset_size: 19684539
+- config_name: train3
+  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: 19687922
+    num_examples: 46908
+  download_size: 7749093
+  dataset_size: 19687922
 configs:
 - config_name: train0
   data_files:
@@ -142,4 +166,8 @@ configs:
   data_files:
   - split: train
     path: train2/train-*
+- config_name: train3
+  data_files:
+  - split: train
+    path: train3/train-*
 ---

train3/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a51fdbfd82ca973e6126c7ff855198f730446778c20e19de5997266362a3029
+size 7749093