Add pure_blsm mixture

README.md

@@ -128,6 +128,22 @@ dataset_info:
     num_examples: 1070
   download_size: 44974608
   dataset_size: 50973209
+- config_name: pure_blsm
+  features:
+  - name: image
+    dtype: image
+  - name: answer
+    dtype: string
+  - name: source_dataset
+    dtype: string
+  - name: split
+    dtype: string
+  splits:
+  - name: train
+    num_bytes: 33808923
+    num_examples: 706
+  download_size: 32741985
+  dataset_size: 33808923
 configs:
 - config_name: balanced_equal
   data_files:
@@ -161,4 +177,8 @@ configs:
   data_files:
   - split: train
     path: mmiq_viz_50_50/train-*
+- config_name: pure_blsm
+  data_files:
+  - split: train
+    path: pure_blsm/train-*
 ---

pure_blsm/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:70642b0d1170be5aa9df00ddd5cad9a2e45979b42c56e35f2240d8990f39ecc7
+size 32741985