fixes

Signed-off-by: amogkam <[email protected]>

python/ray/data/dataset.py

@@ -2852,7 +2852,7 @@ def write_numpy(
  To control the number of number of blocks, call
  :meth:`~ray.data.Dataset.repartition`.
 
-  By default, the format of the output files will be {uuid}_{block_idx}.npy,
+ By default, the format of the output files will be {uuid}_{block_idx}.npy,
  where ``uuid`` is an unique id for the dataset. To modify this behavior,
  implement a custom
  :class:`~ray.data.datasource.BlockWritePathProvider`

python/ray/data/read_api.py

@@ -1888,7 +1888,7 @@ def from_numpy_refs(
  >>> ray.data.from_numpy(arr_ref)
  MaterializedDataset(num_blocks=1, num_rows=1, schema={data: int64})
 
-  >>> # Create a Ray Dataset from a list of NumPy array references.
+ >>> # Create a Ray Dataset from a list of NumPy array references.
  >>> ray.data.from_numpy([arr, arr])
  MaterializedDataset(num_blocks=2, num_rows=2, schema={data: int64})
 

rllib/offline/dataset_writer.py

@@ -71,9 +71,7 @@ def write(self, sample_batch: SampleBatchType):
  # Todo: We should flush at the end of sampling even if this
  # condition was not reached.
  if len(self.samples) >= self.max_num_samples_per_file:
- ds = data.from_items(self.samples).repartition(
- num_blocks=1, shuffle=False
- )
+ ds = data.from_items(self.samples).repartition(num_blocks=1, shuffle=False)
  if self.format == "json":
  ds.write_json(self.path, try_create_dir=True)
  elif self.format == "parquet":