Initial commit of the multimodal Punjabi Visual Analogies Dataset

README.md

@@ -0,0 +1,41 @@
+---
+dataset_info:
+  features:
+  - name: id
+    dtype: string
+  - name: category
+    dtype: string
+  - name: image
+    dtype: image
+  - name: reasoning
+    dtype: string
+  - name: scripts_gurmukhi_question
+    dtype: string
+  - name: scripts_gurmukhi_options
+    sequence: string
+  - name: scripts_gurmukhi_answer
+    dtype: string
+  - name: scripts_shahmukhi_question
+    dtype: string
+  - name: scripts_shahmukhi_options
+    sequence: string
+  - name: scripts_shahmukhi_answer
+    dtype: string
+  - name: scripts_roman_question
+    dtype: string
+  - name: scripts_roman_options
+    sequence: string
+  - name: scripts_roman_answer
+    dtype: string
+  splits:
+  - name: train
+    num_bytes: 541053211.0
+    num_examples: 375
+  download_size: 540957323
+  dataset_size: 541053211.0
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
+---

data/train-00000-of-00002.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41457298da16bfc664590f8ed7d84fbc5a66f71670b421cd78ffe6d12cba2ec3
+size 264080436

data/train-00001-of-00002.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a718d00fa316490868bfcf4a67a7edc8203d1660c5cdfa384d5a721425f3feb1
+size 276876887