fix: repair notebook loading and Colab CI (#1013)

* fix: load tutorial images from Parquet

Signed-off-by: Andre Manoel <amanoel@nvidia.com>

* fix: limit Colab CI to docs dependencies

Signed-off-by: Andre Manoel <amanoel@nvidia.com>

---------

Signed-off-by: Andre Manoel <amanoel@nvidia.com>

Author: andreatnvidia

.github/workflows/check-colab-notebooks.yml

@@ -32,7 +32,7 @@ jobs:
           enable-cache: true
 
       - name: Install dependencies
-        run: uv sync --all-packages --group notebooks --group docs
+        run: uv sync --frozen --only-group docs
 
       - name: Get changed notebook sources
         id: changed

docs/colab_notebooks/4-providing-images-as-context.ipynb

@@ -195,8 +195,8 @@
     "IMG_COUNT = 512  # Number of images to process\n",
     "BASE64_IMAGE_HEIGHT = 512  # Standardized height for model input\n",
     "\n",
-    "# Load the pets dataset (train split, ~23 MB total)\n",
-    "img_dataset_cfg = {\"path\": \"rokmr/pets\", \"split\": \"train\"}"
+    "# Load the Parquet export to avoid ZIP label inference in datasets 5.x.\n",
+    "img_dataset_cfg = {\"path\": \"rokmr/pets\", \"revision\": \"refs/convert/parquet\", \"split\": \"train\"}"
    ]
   },
   {

docs/notebook_source/4-providing-images-as-context.py

@@ -99,8 +99,8 @@
 IMG_COUNT = 512  # Number of images to process
 BASE64_IMAGE_HEIGHT = 512  # Standardized height for model input
 
-# Load the pets dataset (train split, ~23 MB total)
-img_dataset_cfg = {"path": "rokmr/pets", "split": "train"}
+# Load the Parquet export to avoid ZIP label inference in datasets 5.x.
+img_dataset_cfg = {"path": "rokmr/pets", "revision": "refs/convert/parquet", "split": "train"}
 
 
 # %%