Randomly shuffle cached data files when loading them in (#1076)

prob-ml · Oct 7, 2024 · 4694457 · 4694457
1 parent 0e565b1
commit 4694457
Showing 1 changed file with 1 addition and 0 deletions.
diff --git a/bliss/cached_dataset.py b/bliss/cached_dataset.py
@@ -269,6 +269,7 @@ def _load_file_paths_and_slices(self):
         file_names = [
             f for f in sorted(os.listdir(str(self.cached_data_path))) if f.endswith(".pt")
         ]
+        random.shuffle(file_names)
         if self.subset_fraction:
             file_names = file_names[: math.ceil(len(file_names) * self.subset_fraction)]
         self.file_paths = [os.path.join(str(self.cached_data_path), f) for f in file_names]