fix(labels): init labels

train/label_split_data/.amlignore

@@ -5,3 +5,4 @@
 *.ipynb
 *.lock
 *.toml
+/dataset-cats-dogs-others-labels

train/label_split_data/.dockerignore

@@ -2,3 +2,4 @@
 *.ipynb
 *.lock
 *.toml
+/dataset-cats-dogs-others-labels

train/label_split_data/.gitignore

@@ -1,3 +1,4 @@
 .idea/
 output_images/
 output_integration/
+dataset-cats-dogs-others-labels/

train/label_split_data/dataset-cats-dogs-others-labels.yml

@@ -0,0 +1,11 @@
+$schema: https://azuremlschemas.azureedge.net/latest/data.schema.json
+
+# Supported paths include:
+# local: ./<path>
+# blob:  https://<account_name>.blob.core.windows.net/<container_name>/<path>
+# ADLS gen2: abfss://<file_system>@<account_name>.dfs.core.windows.net/<path>/
+# Datastore: azureml://datastores/<data_store_name>/paths/<path>
+type: uri_folder
+name: cats_dogs_others_labels
+description: Cats Dogs and Others Labels
+path: ./dataset-cats-dogs-others-labels

train/label_split_data/init_dataset.sh

@@ -0,0 +1,19 @@
+resource_group=$1
+workspace_name=$2
+dataset_name=cats_dogs_others_labels
+dataset_version=1
+
+echo "Init label dataset"
+
+RESULT=$(az ml data show --resource-group $resource_group --workspace-name $workspace_name --version $dataset_version --name $dataset_name > /dev/null 2>&1; echo $?)
+if [ $RESULT -eq 0 ]; then
+  echo "Dataset already exist"
+else
+  echo "Dataset not found so create it"
+  git clone https://github.com/guillaume-chervet/dataset-cats-dogs-others-labels
+  rm -rf dataset-cats-dogs-others-labels/.git
+  az ml data create -f dataset-cats-dogs-others-labels.yml --resource-group $resource_group --workspace-name $workspace_name
+
+fi
+
+echo "Init label dataset done"