[MINOR:TYPO] Update save_to_disk docstring (#7575)

* [MINOR:TYPO] Update dataset_dict.py

r/hub/filesystem in save_to_disk

* [MINOR:TYPO] Update arrow_dataset.py

r/hub/filesystem in save_to_disk

Author: cakiki

src/datasets/arrow_dataset.py

@@ -1443,7 +1443,7 @@ def save_to_disk(
                 Path (e.g. `dataset/train`) or remote URI (e.g. `s3://my-bucket/dataset/train`)
                 of the dataset directory where the dataset will be saved to.
             max_shard_size (`int` or `str`, *optional*, defaults to `"500MB"`):
-                The maximum size of the dataset shards to be uploaded to the hub. If expressed as a string, needs to be digits followed by a unit
+                The maximum size of the dataset shards to be saved to the filesystem. If expressed as a string, needs to be digits followed by a unit
                 (like `"50MB"`).
             num_shards (`int`, *optional*):
                 Number of shards to write. By default the number of shards depends on `max_shard_size` and `num_proc`.

src/datasets/dataset_dict.py

@@ -1300,7 +1300,7 @@ def save_to_disk(
                 Path (e.g. `dataset/train`) or remote URI (e.g. `s3://my-bucket/dataset/train`)
                 of the dataset dict directory where the dataset dict will be saved to.
             max_shard_size (`int` or `str`, *optional*, defaults to `"500MB"`):
-                The maximum size of the dataset shards to be uploaded to the hub. If expressed as a string, needs to be digits followed by a unit
+                The maximum size of the dataset shards to be saved to the filesystem. If expressed as a string, needs to be digits followed by a unit
                 (like `"50MB"`).
             num_shards (`Dict[str, int]`, *optional*):
                 Number of shards to write. By default the number of shards depends on `max_shard_size` and `num_proc`.