bug(dataset): fix dataset label size calculate (#1027)

fix dataset label size calculate
star-whale · Aug 28, 2022 · 113a513 · 113a513
1 parent bc1ed6c
commit 113a513
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 3 deletions.
diff --git a/client/starwhale/api/_impl/dataset/builder.py b/client/starwhale/api/_impl/dataset/builder.py
@@ -378,7 +378,7 @@ def make_swds(self) -> DatasetSummary:
             )
 
             total_data_size += data_size
-            total_label_size += len(label)
+            total_label_size += sys.getsizeof(label)
             increased_rows += 1
 
         self._copy_files(map_path_sign)

diff --git a/example/mnist/dataset.yaml b/example/mnist/dataset.yaml
@@ -3,8 +3,8 @@ name: mnist
 data_dir: data
 data_filter: "t10k-image*"
 label_filter: "t10k-label*"
-process: mnist.process:DataSetProcessExecutor
-#process: mnist.process:RawDataSetProcessExecutor
+#process: mnist.process:DataSetProcessExecutor
+process: mnist.process:RawDataSetProcessExecutor
 #process: mnist.process:LinkRawDataSetProcessExecutor
 
 desc: MNIST data and label test dataset