Convert dataset to Parquet

Convert dataset to Parquet.

README.md

@@ -0,0 +1,30 @@
+---
+dataset_info:
+  config_name: induction
+  features:
+  - name: source
+    dtype: string
+  - name: target
+    dtype: string
+  splits:
+  - name: train
+    num_bytes: 98736
+    num_examples: 1000
+  - name: test
+    num_bytes: 490070
+    num_examples: 5000
+  - name: validation
+    num_bytes: 491609
+    num_examples: 5000
+  download_size: 370911
+  dataset_size: 1080415
+configs:
+- config_name: induction
+  data_files:
+  - split: train
+    path: induction/train-*
+  - split: test
+    path: induction/test-*
+  - split: validation
+    path: induction/validation-*
+---

induction/test-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ccef06a6590fc25e7072d05733e60da2c69ef136865da26e022e1e1498d7d2ed
+size 168102

induction/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62acb7417ec9c9b484f16893a835a074a7ae0c14a27447e4b48b26f1cd606c40
+size 34367

induction/validation-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec2c0d0c7fb86dbcace2773223fa3827505bd471e87729078aa8cf5815134dac
+size 168442