HP tuning split: 3000 train, 750 val (stratified)

README.md

@@ -0,0 +1,30 @@
+---
+dataset_info:
+  features:
+  - name: id
+    dtype: int64
+  - name: clean_text
+    dtype: string
+  - name: cyberbullying_type
+    dtype: string
+  - name: has_emoji
+    dtype: bool
+  - name: emoji_count
+    dtype: int64
+  splits:
+  - name: train
+    num_bytes: 489878
+    num_examples: 3000
+  - name: validation
+    num_bytes: 117707
+    num_examples: 750
+  download_size: 381818
+  dataset_size: 607585
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
+  - split: validation
+    path: data/validation-*
+---

data/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:683c59f069a970657f466172b9f87728b0ea0f56a24a4427c63e704ec3659eee
+size 305914

data/validation-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f94085394d8345eea6ee14c22ab8f36c9474b2b3940641bf6acda7131dde07e
+size 75904