diff --git a/phases/00-setup-and-tooling/09-data-management/code/data_utils.py b/phases/00-setup-and-tooling/09-data-management/code/data_utils.py
index 25f9dd037..5df6f045e 100644
--- a/phases/00-setup-and-tooling/09-data-management/code/data_utils.py
+++ b/phases/00-setup-and-tooling/09-data-management/code/data_utils.py
@@ -1,4 +1,3 @@
-import os
 import sys
 import json
 import hashlib
@@ -159,10 +158,10 @@ def fingerprint(ds, num_rows: int = 100):
     print("=" * 60)
 
     print("\n--- 1. Load and inspect a dataset ---")
-    ds = load_and_inspect("rotten_tomatoes", split="train")
+    ds = load_and_inspect("cornell-movie-review-data/rotten_tomatoes", split="train")
 
     print("\n--- 2. Stream a dataset ---")
-    rows = stream_dataset("rotten_tomatoes", max_rows=3)
+    rows = stream_dataset("cornell-movie-review-data/rotten_tomatoes", max_rows=3)
     for row in rows:
         print(f"  {row['text'][:80]}...")
 
diff --git a/phases/00-setup-and-tooling/09-data-management/outputs/prompt-data-helper.md b/phases/00-setup-and-tooling/09-data-management/outputs/prompt-data-helper.md
index 8daf3ff31..832b4234e 100644
--- a/phases/00-setup-and-tooling/09-data-management/outputs/prompt-data-helper.md
+++ b/phases/00-setup-and-tooling/09-data-management/outputs/prompt-data-helper.md
@@ -33,15 +33,15 @@ Common task-to-dataset mapping:
 
 | Task | Starter Dataset | HF ID |
 |------|----------------|-------|
-| Text classification | Rotten Tomatoes | `rotten_tomatoes` |
-| Sentiment analysis | IMDB | `imdb` |
-| Natural language inference | MNLI | `glue/mnli` |
-| Question answering | SQuAD | `squad` |
-| Summarization | CNN/DailyMail | `cnn_dailymail` |
-| Translation | WMT | `wmt16` |
-| Language modeling | WikiText | `wikitext` |
-| Token classification | CoNLL-2003 | `conll2003` |
-| Image classification | MNIST / CIFAR-10 | `mnist` / `cifar10` |
+| Text classification | Rotten Tomatoes | `cornell-movie-review-data/rotten_tomatoes` |
+| Sentiment analysis | IMDB | `stanfordnlp/imdb` |
+| Natural language inference | MNLI | `nyu-mll/glue` (config:`mnli`) |
+| Question answering | SQuAD | `rajpurkar/squad` |
+| Summarization | CNN/DailyMail | `abisee/cnn_dailymail`(config: `3.0.0`) |
+| Translation | WMT | `wmt/wmt16`(config: `cs-en`) |
+| Language modeling | WikiText | `Salesforce/wikitext` |
+| Token classification | CoNLL-2003 | `lhoestq/conll2003` |
+| Image classification | MNIST / CIFAR-10 | `ylecun/mnist` / `uoft-cs/cifar10` |
 | Object detection | COCO | `detection-datasets/coco` |
 
 When recommending, prefer smaller datasets for learning and prototyping. Suggest larger datasets only when the user is ready to train at scale.