Convert dataset to Parquet

Convert dataset to Parquet.

README.md

@@ -0,0 +1,49 @@
+---
+dataset_info:
+  config_name: source
+  features:
+  - name: id
+    dtype: string
+  - name: abstract
+    dtype: string
+  - name: specialities
+    sequence:
+      class_label:
+        names:
+          '0': microbiology
+          '1': etiology
+          '2': virology
+          '3': physiology
+          '4': immunology
+          '5': parasitology
+          '6': genetics
+          '7': chemistry
+          '8': veterinary
+          '9': surgery
+          '10': pharmacology
+          '11': psychology
+  - name: specialities_one_hot
+    sequence: float32
+  splits:
+  - name: train
+    num_bytes: 2601475
+    num_examples: 1514
+  - name: validation
+    num_bytes: 1742790
+    num_examples: 1022
+  - name: test
+    num_bytes: 1859985
+    num_examples: 1088
+  download_size: 3407937
+  dataset_size: 6204250
+configs:
+- config_name: source
+  data_files:
+  - split: train
+    path: source/train-*
+  - split: validation
+    path: source/validation-*
+  - split: test
+    path: source/test-*
+  default: true
+---

source/test-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3048c2f17ac0402aaf8b999972cac77ca1131f579805548d64e823185b0f116c
+size 1024481

source/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:798cdc4c7d8c5c82ad2824386a9d98adadb40ea228ebd702a0cfd121fc86740a
+size 1422136

source/validation-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fcd6d15214df38e19ba8ba43f8f214a26aa7b243146d9a1541135574ee77db72
+size 961320