From 852dff564ed28684b30d611cdd2f3c8830d392ab Mon Sep 17 00:00:00 2001
From: imagineer99 <samleejackson0@gmail.com>
Date: Thu, 26 Feb 2026 06:32:45 +0000
Subject: [PATCH] feat: added datasets of size 5M and 10M to pretraining size
 category

---
 studio/frontend/src/hooks/use-hf-dataset-search.ts | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/studio/frontend/src/hooks/use-hf-dataset-search.ts b/studio/frontend/src/hooks/use-hf-dataset-search.ts
index 32879fd015..ec64ffd7e6 100644
--- a/studio/frontend/src/hooks/use-hf-dataset-search.ts
+++ b/studio/frontend/src/hooks/use-hf-dataset-search.ts
@@ -141,6 +141,8 @@ const PRETRAINING_PLAIN_TAGS = new Set(["pretraining", "pre-training"]);
 const OCR_PLAIN_TAGS = new Set(["ocr", "document-ocr"]);
 
 const PRETRAINING_SIZE_CATEGORIES = new Set([
+  "5M<n<10M",
+  "10M<n<100M",
   "100M<n<1B",
   "1B<n<10B",
   "10B<n<100B",