dataset

Author: Borda

.actions/assistant.py

@@ -289,7 +289,7 @@ def _bash_download_data(folder: str) -> List[str]:
             if ext not in AssistantCLI._EXT_ARCHIVE:
                 continue
             if ext in AssistantCLI._EXT_ARCHIVE_ZIP:
-                cmd += [f"unzip -o {fn} -d {AssistantCLI.DATASETS_FOLDER}/{name}/ {UNZIP_PROGRESS_BAR}"]
+                cmd += [f"unzip -o {fn} -d {AssistantCLI.DATASETS_FOLDER}/{name} {UNZIP_PROGRESS_BAR}"]
             else:
                 cmd += [f"tar -zxvf {fn} --overwrite"]
             cmd += [f"rm {fn}"]

.github/workflows/ci_testing.yml

@@ -74,7 +74,7 @@ jobs:
           KAGGLE_KEY: ${{ secrets.KAGGLE_KEY }}
         run: |
           set -e
-          apt install -y tree
+          sudo apt install -y tree
           while IFS= read -r line; do
               python .actions/assistant.py bash-test $line
               cat .actions/_ipynb-test.sh

templates/img-classify/classify.py

@@ -9,8 +9,9 @@
 from pytorch_lightning.loggers import CSVLogger
 
 PATH_DATASETS = os.environ.get("PATH_DATASETS", ".")
-# this dataset is automaticaly downloaded and extracted based on meta link
-DATA_HYMENOPLERA = os.path.join(PATH_DATASETS, "hymenoptera_data")
+# this dataset is automatically downloaded and extracted based on meta link
+# this archive includes the one more level - folder with the same name
+DATA_HYMENOPLERA = os.path.join(PATH_DATASETS, "hymenoptera_data", "hymenoptera_data")
 
 # %% [markdown]
 # ## 1. Create the DataModule