Expanduser in save_to_disk()

[Unknown3141592]

Fixes #5651. The same problem occurs when loading from disk so I fixed it there too.

I am not sure why the case distinction between local and remote filesystems is even necessary for DatasetDict when saving to disk. Imo this could be removed (leaving only fs.makedirs(dataset_dict_path, exist_ok=True)).

[HuggingFaceDocBuilderDev]

The documentation is not available anymore as the PR was closed or merged.

[mariosasko]

I renamed resolved_path to save_path. Looks all good now. Thanks!

[mariosasko]

I am not sure why the case distinction between local and remote filesystems is even necessary for DatasetDict when saving to disk. Imo this could be removed (leaving only fs.makedirs(dataset_dict_path, exist_ok=True)).

Indeed. But it's better to address this in a separate PR.

[github-actions]

Show benchmarks

PyArrow==8.0.0

Show updated benchmarks!

Benchmark: benchmark_array_xd.json

metric	read_batch_formatted_as_numpy after write_array2d	read_batch_formatted_as_numpy after write_flattened_sequence	read_batch_formatted_as_numpy after write_nested_sequence	read_batch_unformated after write_array2d	read_batch_unformated after write_flattened_sequence	read_batch_unformated after write_nested_sequence	read_col_formatted_as_numpy after write_array2d	read_col_formatted_as_numpy after write_flattened_sequence	read_col_formatted_as_numpy after write_nested_sequence	read_col_unformated after write_array2d	read_col_unformated after write_flattened_sequence	read_col_unformated after write_nested_sequence	read_formatted_as_numpy after write_array2d	read_formatted_as_numpy after write_flattened_sequence	read_formatted_as_numpy after write_nested_sequence	read_unformated after write_array2d	read_unformated after write_flattened_sequence	read_unformated after write_nested_sequence	write_array2d	write_flattened_sequence	write_nested_sequence
new / old (diff)	0.007696 / 0.011353 (-0.003656)	0.004497 / 0.011008 (-0.006511)	0.099302 / 0.038508 (0.060794)	0.083360 / 0.023109 (0.060251)	0.393483 / 0.275898 (0.117585)	0.450505 / 0.323480 (0.127025)	0.004610 / 0.007986 (-0.003376)	0.003637 / 0.004328 (-0.000692)	0.075752 / 0.004250 (0.071501)	0.064034 / 0.037052 (0.026982)	0.397785 / 0.258489 (0.139296)	0.462948 / 0.293841 (0.169107)	0.035902 / 0.128546 (-0.092644)	0.009640 / 0.075646 (-0.066007)	0.342299 / 0.419271 (-0.076973)	0.059586 / 0.043533 (0.016053)	0.404918 / 0.255139 (0.149779)	0.440889 / 0.283200 (0.157690)	0.028981 / 0.141683 (-0.112702)	1.775380 / 1.452155 (0.323226)	1.866663 / 1.492716 (0.373946)

Benchmark: benchmark_getitem_100B.json

metric	get_batch_of_1024_random_rows	get_batch_of_1024_rows	get_first_row	get_last_row
new / old (diff)	0.249080 / 0.018006 (0.231074)	0.456460 / 0.000490 (0.455970)	0.028145 / 0.000200 (0.027945)	0.000402 / 0.000054 (0.000347)

Benchmark: benchmark_indices_mapping.json

metric	select	shard	shuffle	sort	train_test_split
new / old (diff)	0.030373 / 0.037411 (-0.007038)	0.088562 / 0.014526 (0.074036)	0.122837 / 0.176557 (-0.053720)	0.167122 / 0.737135 (-0.570014)	0.103953 / 0.296338 (-0.192385)

Benchmark: benchmark_iterating.json

metric	read 5000	read 50000	read_batch 50000 10	read_batch 50000 100	read_batch 50000 1000	read_formatted numpy 5000	read_formatted pandas 5000	read_formatted tensorflow 5000	read_formatted torch 5000	read_formatted_batch numpy 5000 10	read_formatted_batch numpy 5000 1000	shuffled read 5000	shuffled read 50000	shuffled read_batch 50000 10	shuffled read_batch 50000 100	shuffled read_batch 50000 1000	shuffled read_formatted numpy 5000	shuffled read_formatted_batch numpy 5000 10	shuffled read_formatted_batch numpy 5000 1000
new / old (diff)	0.431714 / 0.215209 (0.216505)	4.182224 / 2.077655 (2.104570)	2.025650 / 1.504120 (0.521530)	1.838905 / 1.541195 (0.297710)	1.868710 / 1.468490 (0.400219)	0.538422 / 4.584777 (-4.046355)	4.038941 / 3.745712 (0.293228)	3.717695 / 5.269862 (-1.552166)	2.313197 / 4.565676 (-2.252479)	0.061060 / 0.424275 (-0.363215)	0.008248 / 0.007607 (0.000641)	0.497438 / 0.226044 (0.271394)	4.946663 / 2.268929 (2.677734)	2.571841 / 55.444624 (-52.872784)	2.155894 / 6.876477 (-4.720583)	2.183180 / 2.142072 (0.041107)	0.639810 / 4.805227 (-4.165417)	0.153273 / 6.500664 (-6.347391)	0.068606 / 0.075469 (-0.006863)

Benchmark: benchmark_map_filter.json

metric	filter	map fast-tokenizer batched	map identity	map identity batched	map no-op batched	map no-op batched numpy	map no-op batched pandas	map no-op batched pytorch	map no-op batched tensorflow
new / old (diff)	1.376152 / 1.841788 (-0.465635)	20.747088 / 8.074308 (12.672780)	15.200311 / 10.191392 (5.008919)	0.166380 / 0.680424 (-0.514043)	0.021417 / 0.534201 (-0.512784)	0.435677 / 0.579283 (-0.143606)	0.460412 / 0.434364 (0.026048)	0.509978 / 0.540337 (-0.030359)	0.702506 / 1.386936 (-0.684430)

PyArrow==latest

Show updated benchmarks!

Benchmark: benchmark_array_xd.json

metric	read_batch_formatted_as_numpy after write_array2d	read_batch_formatted_as_numpy after write_flattened_sequence	read_batch_formatted_as_numpy after write_nested_sequence	read_batch_unformated after write_array2d	read_batch_unformated after write_flattened_sequence	read_batch_unformated after write_nested_sequence	read_col_formatted_as_numpy after write_array2d	read_col_formatted_as_numpy after write_flattened_sequence	read_col_formatted_as_numpy after write_nested_sequence	read_col_unformated after write_array2d	read_col_unformated after write_flattened_sequence	read_col_unformated after write_nested_sequence	read_formatted_as_numpy after write_array2d	read_formatted_as_numpy after write_flattened_sequence	read_formatted_as_numpy after write_nested_sequence	read_unformated after write_array2d	read_unformated after write_flattened_sequence	read_unformated after write_nested_sequence	write_array2d	write_flattened_sequence	write_nested_sequence
new / old (diff)	0.007378 / 0.011353 (-0.003975)	0.003938 / 0.011008 (-0.007070)	0.067095 / 0.038508 (0.028587)	0.082252 / 0.023109 (0.059143)	0.420317 / 0.275898 (0.144419)	0.477496 / 0.323480 (0.154017)	0.006259 / 0.007986 (-0.001727)	0.003513 / 0.004328 (-0.000816)	0.072107 / 0.004250 (0.067856)	0.061737 / 0.037052 (0.024684)	0.444142 / 0.258489 (0.185653)	0.488926 / 0.293841 (0.195085)	0.033623 / 0.128546 (-0.094923)	0.008091 / 0.075646 (-0.067555)	0.073997 / 0.419271 (-0.345274)	0.051295 / 0.043533 (0.007762)	0.442551 / 0.255139 (0.187412)	0.462713 / 0.283200 (0.179513)	0.023115 / 0.141683 (-0.118568)	1.645759 / 1.452155 (0.193604)	1.758121 / 1.492716 (0.265405)

Benchmark: benchmark_getitem_100B.json

metric	get_batch_of_1024_random_rows	get_batch_of_1024_rows	get_first_row	get_last_row
new / old (diff)	0.233450 / 0.018006 (0.215444)	0.445384 / 0.000490 (0.444894)	0.006412 / 0.000200 (0.006212)	0.000111 / 0.000054 (0.000056)

Benchmark: benchmark_indices_mapping.json

metric	select	shard	shuffle	sort	train_test_split
new / old (diff)	0.032446 / 0.037411 (-0.004965)	0.098515 / 0.014526 (0.083989)	0.109095 / 0.176557 (-0.067462)	0.167645 / 0.737135 (-0.569490)	0.110403 / 0.296338 (-0.185936)

Benchmark: benchmark_iterating.json

metric	read 5000	read 50000	read_batch 50000 10	read_batch 50000 100	read_batch 50000 1000	read_formatted numpy 5000	read_formatted pandas 5000	read_formatted tensorflow 5000	read_formatted torch 5000	read_formatted_batch numpy 5000 10	read_formatted_batch numpy 5000 1000	shuffled read 5000	shuffled read 50000	shuffled read_batch 50000 10	shuffled read_batch 50000 100	shuffled read_batch 50000 1000	shuffled read_formatted numpy 5000	shuffled read_formatted_batch numpy 5000 10	shuffled read_formatted_batch numpy 5000 1000
new / old (diff)	0.470189 / 0.215209 (0.254980)	4.663224 / 2.077655 (2.585569)	2.504474 / 1.504120 (1.000354)	2.282867 / 1.541195 (0.741673)	2.331598 / 1.468490 (0.863108)	0.554421 / 4.584777 (-4.030356)	4.078657 / 3.745712 (0.332945)	3.516339 / 5.269862 (-1.753523)	2.239134 / 4.565676 (-2.326542)	0.062690 / 0.424275 (-0.361585)	0.008406 / 0.007607 (0.000799)	0.533827 / 0.226044 (0.307782)	5.423984 / 2.268929 (3.155055)	2.972784 / 55.444624 (-52.471840)	2.699056 / 6.876477 (-4.177421)	2.844403 / 2.142072 (0.702331)	0.639194 / 4.805227 (-4.166033)	0.142097 / 6.500664 (-6.358567)	0.064646 / 0.075469 (-0.010823)

Benchmark: benchmark_map_filter.json

metric	filter	map fast-tokenizer batched	map identity	map identity batched	map no-op batched	map no-op batched numpy	map no-op batched pandas	map no-op batched pytorch	map no-op batched tensorflow
new / old (diff)	1.544640 / 1.841788 (-0.297148)	21.453429 / 8.074308 (13.379121)	15.610723 / 10.191392 (5.419331)	0.207796 / 0.680424 (-0.472628)	0.021912 / 0.534201 (-0.512289)	0.430472 / 0.579283 (-0.148811)	0.467530 / 0.434364 (0.033166)	0.541339 / 0.540337 (0.001002)	0.721976 / 1.386936 (-0.664960)