rename datasets-server to dataset-viewer

[severo]

See huggingface/dataset-viewer#2650

Tell me if it's OK, or if it's a breaking change that must be handled differently.

Also note that the docs page is still https://huggingface.co/docs/datasets-server/, so I didn't change it.

And the API URL is still https://datasets-server.huggingface.co/ (and might always be), so I let it too.

[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.

[mariosasko]

Thanks!

We treat the utils._datasets_server module as private (and don't directly expose the DatasetsServerError), so this shouldn't break users' code.

[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)	0.005224 / 0.011353 (-0.006129)	0.003938 / 0.011008 (-0.007070)	0.063829 / 0.038508 (0.025321)	0.030975 / 0.023109 (0.007865)	0.265090 / 0.275898 (-0.010808)	0.290994 / 0.323480 (-0.032486)	0.003083 / 0.007986 (-0.004902)	0.002810 / 0.004328 (-0.001518)	0.048860 / 0.004250 (0.044609)	0.044663 / 0.037052 (0.007611)	0.272161 / 0.258489 (0.013672)	0.306966 / 0.293841 (0.013125)	0.028028 / 0.128546 (-0.100518)	0.010616 / 0.075646 (-0.065031)	0.211649 / 0.419271 (-0.207623)	0.035906 / 0.043533 (-0.007626)	0.251779 / 0.255139 (-0.003360)	0.275543 / 0.283200 (-0.007657)	0.017710 / 0.141683 (-0.123973)	1.127015 / 1.452155 (-0.325139)	1.173319 / 1.492716 (-0.319397)

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)	0.090625 / 0.018006 (0.072619)	0.301973 / 0.000490 (0.301483)	0.000217 / 0.000200 (0.000017)	0.000053 / 0.000054 (-0.000002)

Benchmark: benchmark_indices_mapping.json

metric	select	shard	shuffle	sort	train_test_split
new / old (diff)	0.018868 / 0.037411 (-0.018543)	0.062402 / 0.014526 (0.047876)	0.074053 / 0.176557 (-0.102504)	0.121484 / 0.737135 (-0.615652)	0.078674 / 0.296338 (-0.217664)

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)	0.277821 / 0.215209 (0.062612)	2.761642 / 2.077655 (0.683987)	1.452735 / 1.504120 (-0.051385)	1.336303 / 1.541195 (-0.204891)	1.343045 / 1.468490 (-0.125445)	0.560917 / 4.584777 (-4.023860)	2.353427 / 3.745712 (-1.392286)	2.699067 / 5.269862 (-2.570795)	1.704752 / 4.565676 (-2.860925)	0.062668 / 0.424275 (-0.361607)	0.005120 / 0.007607 (-0.002487)	0.330455 / 0.226044 (0.104410)	3.264604 / 2.268929 (0.995675)	1.791940 / 55.444624 (-53.652685)	1.526083 / 6.876477 (-5.350394)	1.541429 / 2.142072 (-0.600643)	0.630343 / 4.805227 (-4.174884)	0.115189 / 6.500664 (-6.385475)	0.041716 / 0.075469 (-0.033753)

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)	0.975008 / 1.841788 (-0.866779)	11.326924 / 8.074308 (3.252616)	9.810300 / 10.191392 (-0.381092)	0.141068 / 0.680424 (-0.539356)	0.013950 / 0.534201 (-0.520251)	0.285691 / 0.579283 (-0.293592)	0.257968 / 0.434364 (-0.176396)	0.322976 / 0.540337 (-0.217361)	0.411114 / 1.386936 (-0.975822)

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)	0.005176 / 0.011353 (-0.006177)	0.003631 / 0.011008 (-0.007377)	0.050006 / 0.038508 (0.011498)	0.030622 / 0.023109 (0.007513)	0.277364 / 0.275898 (0.001466)	0.299752 / 0.323480 (-0.023728)	0.004110 / 0.007986 (-0.003876)	0.002694 / 0.004328 (-0.001634)	0.048966 / 0.004250 (0.044715)	0.039634 / 0.037052 (0.002582)	0.289959 / 0.258489 (0.031470)	0.320689 / 0.293841 (0.026848)	0.029285 / 0.128546 (-0.099261)	0.010435 / 0.075646 (-0.065211)	0.057432 / 0.419271 (-0.361840)	0.032554 / 0.043533 (-0.010979)	0.277354 / 0.255139 (0.022215)	0.296872 / 0.283200 (0.013673)	0.017338 / 0.141683 (-0.124344)	1.134174 / 1.452155 (-0.317981)	1.184695 / 1.492716 (-0.308021)

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)	0.089953 / 0.018006 (0.071947)	0.299372 / 0.000490 (0.298882)	0.000212 / 0.000200 (0.000012)	0.000043 / 0.000054 (-0.000012)

Benchmark: benchmark_indices_mapping.json

metric	select	shard	shuffle	sort	train_test_split
new / old (diff)	0.021349 / 0.037411 (-0.016062)	0.075167 / 0.014526 (0.060641)	0.085910 / 0.176557 (-0.090647)	0.124729 / 0.737135 (-0.612406)	0.088313 / 0.296338 (-0.208025)

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)	0.291939 / 0.215209 (0.076730)	2.851077 / 2.077655 (0.773423)	1.609382 / 1.504120 (0.105262)	1.469656 / 1.541195 (-0.071539)	1.490469 / 1.468490 (0.021979)	0.570421 / 4.584777 (-4.014356)	2.441438 / 3.745712 (-1.304274)	2.756514 / 5.269862 (-2.513347)	1.714202 / 4.565676 (-2.851474)	0.063656 / 0.424275 (-0.360619)	0.005640 / 0.007607 (-0.001967)	0.336240 / 0.226044 (0.110196)	3.355434 / 2.268929 (1.086505)	1.947553 / 55.444624 (-53.497072)	1.672776 / 6.876477 (-5.203700)	1.685316 / 2.142072 (-0.456757)	0.638849 / 4.805227 (-4.166378)	0.116304 / 6.500664 (-6.384360)	0.041588 / 0.075469 (-0.033881)

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)	1.026700 / 1.841788 (-0.815088)	12.044628 / 8.074308 (3.970319)	10.464007 / 10.191392 (0.272615)	0.156169 / 0.680424 (-0.524255)	0.015624 / 0.534201 (-0.518577)	0.287233 / 0.579283 (-0.292050)	0.270374 / 0.434364 (-0.163990)	0.325255 / 0.540337 (-0.215083)	0.412021 / 1.386936 (-0.974915)