Set temporary numpy upper version < 2.0.0 to fix CI

[albertvillanova]

Set temporary numpy upper version < 2.0.0 to fix CI. See: https://github.com/huggingface/datasets/actions/runs/9546031216/job/26308072017

A module that was compiled using NumPy 1.x cannot be run in
NumPy 2.0.0 as it may crash. To support both 1.x and 2.x
versions of NumPy, modules must be compiled with NumPy 2.0.
Some module may need to rebuild instead e.g. with 'pybind11>=2.12'.

If you are a user of the module, the easiest solution will be to
downgrade to 'numpy<2' or try to upgrade the affected module.
We expect that some modules will need time to support NumPy 2.

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

[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)	0.005168 / 0.011353 (-0.006185)	0.003720 / 0.011008 (-0.007288)	0.063347 / 0.038508 (0.024839)	0.031474 / 0.023109 (0.008364)	0.243233 / 0.275898 (-0.032665)	0.276695 / 0.323480 (-0.046785)	0.004109 / 0.007986 (-0.003877)	0.002689 / 0.004328 (-0.001639)	0.049522 / 0.004250 (0.045271)	0.043477 / 0.037052 (0.006425)	0.258578 / 0.258489 (0.000088)	0.288134 / 0.293841 (-0.005707)	0.027836 / 0.128546 (-0.100710)	0.010677 / 0.075646 (-0.064969)	0.206412 / 0.419271 (-0.212860)	0.036204 / 0.043533 (-0.007329)	0.250588 / 0.255139 (-0.004551)	0.272354 / 0.283200 (-0.010846)	0.018359 / 0.141683 (-0.123324)	1.118867 / 1.452155 (-0.333288)	1.157318 / 1.492716 (-0.335399)

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)	0.092927 / 0.018006 (0.074921)	0.298252 / 0.000490 (0.297762)	0.000228 / 0.000200 (0.000028)	0.000042 / 0.000054 (-0.000013)

Benchmark: benchmark_indices_mapping.json

metric	select	shard	shuffle	sort	train_test_split
new / old (diff)	0.018824 / 0.037411 (-0.018588)	0.069304 / 0.014526 (0.054778)	0.075094 / 0.176557 (-0.101462)	0.122546 / 0.737135 (-0.614590)	0.076453 / 0.296338 (-0.219885)

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)	0.287131 / 0.215209 (0.071922)	2.838945 / 2.077655 (0.761291)	1.473578 / 1.504120 (-0.030542)	1.351214 / 1.541195 (-0.189981)	1.354924 / 1.468490 (-0.113566)	0.577092 / 4.584777 (-4.007685)	2.348072 / 3.745712 (-1.397640)	2.762130 / 5.269862 (-2.507732)	1.725195 / 4.565676 (-2.840482)	0.063596 / 0.424275 (-0.360679)	0.004921 / 0.007607 (-0.002686)	0.335422 / 0.226044 (0.109377)	3.340398 / 2.268929 (1.071469)	1.789390 / 55.444624 (-53.655234)	1.516247 / 6.876477 (-5.360229)	1.529653 / 2.142072 (-0.612420)	0.643547 / 4.805227 (-4.161680)	0.116491 / 6.500664 (-6.384173)	0.042404 / 0.075469 (-0.033065)

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)	0.959839 / 1.841788 (-0.881948)	11.269778 / 8.074308 (3.195470)	9.574898 / 10.191392 (-0.616494)	0.128979 / 0.680424 (-0.551444)	0.013901 / 0.534201 (-0.520300)	0.280778 / 0.579283 (-0.298505)	0.256511 / 0.434364 (-0.177853)	0.319361 / 0.540337 (-0.220977)	0.411803 / 1.386936 (-0.975133)

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)	0.005453 / 0.011353 (-0.005899)	0.003478 / 0.011008 (-0.007530)	0.050055 / 0.038508 (0.011547)	0.031415 / 0.023109 (0.008306)	0.275057 / 0.275898 (-0.000841)	0.296690 / 0.323480 (-0.026789)	0.004253 / 0.007986 (-0.003732)	0.002777 / 0.004328 (-0.001551)	0.049553 / 0.004250 (0.045303)	0.039843 / 0.037052 (0.002791)	0.286938 / 0.258489 (0.028449)	0.318579 / 0.293841 (0.024738)	0.029773 / 0.128546 (-0.098774)	0.010404 / 0.075646 (-0.065242)	0.057915 / 0.419271 (-0.361356)	0.033486 / 0.043533 (-0.010047)	0.273293 / 0.255139 (0.018154)	0.293155 / 0.283200 (0.009955)	0.017843 / 0.141683 (-0.123839)	1.131130 / 1.452155 (-0.321024)	1.167412 / 1.492716 (-0.325304)

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)	0.092553 / 0.018006 (0.074547)	0.298888 / 0.000490 (0.298399)	0.000201 / 0.000200 (0.000001)	0.000043 / 0.000054 (-0.000011)

Benchmark: benchmark_indices_mapping.json

metric	select	shard	shuffle	sort	train_test_split
new / old (diff)	0.022646 / 0.037411 (-0.014765)	0.076921 / 0.014526 (0.062395)	0.089238 / 0.176557 (-0.087318)	0.128793 / 0.737135 (-0.608342)	0.089190 / 0.296338 (-0.207148)

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)	0.292552 / 0.215209 (0.077343)	2.884277 / 2.077655 (0.806622)	1.568798 / 1.504120 (0.064678)	1.441819 / 1.541195 (-0.099375)	1.435766 / 1.468490 (-0.032724)	0.572435 / 4.584777 (-4.012342)	0.957387 / 3.745712 (-2.788326)	2.650843 / 5.269862 (-2.619019)	1.727424 / 4.565676 (-2.838252)	0.063470 / 0.424275 (-0.360805)	0.005314 / 0.007607 (-0.002293)	0.345881 / 0.226044 (0.119836)	3.395463 / 2.268929 (1.126535)	1.921340 / 55.444624 (-53.523285)	1.621563 / 6.876477 (-5.254914)	1.742561 / 2.142072 (-0.399512)	0.639948 / 4.805227 (-4.165279)	0.116091 / 6.500664 (-6.384573)	0.041218 / 0.075469 (-0.034251)

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)	0.991506 / 1.841788 (-0.850281)	11.897462 / 8.074308 (3.823154)	10.083008 / 10.191392 (-0.108384)	0.140626 / 0.680424 (-0.539798)	0.015454 / 0.534201 (-0.518747)	0.283856 / 0.579283 (-0.295427)	0.125935 / 0.434364 (-0.308429)	0.323884 / 0.540337 (-0.216454)	0.438348 / 1.386936 (-0.948588)