add tests

tests/test_serialization.py

@@ -1,4 +1,5 @@
 from huggingface_hub.serialization import split_state_dict_into_shards_factory
+from huggingface_hub.serialization._base import convert_file_size_to_int
 from huggingface_hub.serialization._numpy import get_tensor_size as get_tensor_size_numpy
 from huggingface_hub.serialization._tensorflow import get_tensor_size as get_tensor_size_tensorflow
 from huggingface_hub.serialization._torch import get_tensor_size as get_tensor_size_torch
@@ -123,3 +124,13 @@ def test_get_tensor_size_torch():
 
     assert get_tensor_size_torch(torch.tensor([1, 2, 3, 4, 5], dtype=torch.float64)) == 5 * 8
     assert get_tensor_size_torch(torch.tensor([1, 2, 3, 4, 5], dtype=torch.float16)) == 5 * 2
+
+
+def test_convert_file_size_to_int():
+    assert convert_file_size_to_int("1KiB") == 2**10
+    assert convert_file_size_to_int("1KB") == 10**3
+    assert convert_file_size_to_int("1MiB") == 2**20
+    assert convert_file_size_to_int("1MB") == 10**6
+    assert convert_file_size_to_int("1GiB") == 2**30
+    assert convert_file_size_to_int("1GB") == 10**9
+    assert convert_file_size_to_int("5GB") == 5 * 10**9