Fix ConnectionError for gated datasets and unauthenticated users

[albertvillanova]

Fix ConnectionError for gated datasets and unauthenticated users. See:

• CI worker tests are broken for gated datasets: ConnectionError dataset-viewer#3025

Note that a recent change in the Hub returns dataset info for gated datasets and unauthenticated users, instead of raising a GatedRepoError as before. See:

• Gated dataset info is leaked huggingface_hub#2457

This PR adds an additional check (/auth-check) for gated datasets and raises DatasetNotFoundError for unauthenticated users, as it was the case before the change in the Hub.

• Fix suggested by @Pierrci (thanks @Wauplin for pointing it out).

Fix #7109.

[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.

[albertvillanova]

Note that the CI error is unrelated to this PR and should be addressed in another PR. See:

• CI is broken for numpy-2: Failed to fetch wheel: llvmlite==0.34.0 #7111

[Wauplin]

Looks good to me! We might add support for the /auth-check endpoint in huggingface_hub directly in the future.

[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)	0.005354 / 0.011353 (-0.005999)	0.004031 / 0.011008 (-0.006977)	0.062470 / 0.038508 (0.023962)	0.030882 / 0.023109 (0.007773)	0.244816 / 0.275898 (-0.031082)	0.264324 / 0.323480 (-0.059156)	0.004164 / 0.007986 (-0.003822)	0.002858 / 0.004328 (-0.001471)	0.049008 / 0.004250 (0.044758)	0.042139 / 0.037052 (0.005086)	0.279496 / 0.258489 (0.021007)	0.279408 / 0.293841 (-0.014433)	0.029701 / 0.128546 (-0.098845)	0.012501 / 0.075646 (-0.063145)	0.203267 / 0.419271 (-0.216004)	0.035964 / 0.043533 (-0.007569)	0.239361 / 0.255139 (-0.015778)	0.258942 / 0.283200 (-0.024257)	0.017956 / 0.141683 (-0.123727)	1.160468 / 1.452155 (-0.291687)	1.203475 / 1.492716 (-0.289242)

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)	0.004639 / 0.018006 (-0.013367)	0.298020 / 0.000490 (0.297530)	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)	0.019371 / 0.037411 (-0.018040)	0.063311 / 0.014526 (0.048785)	0.076412 / 0.176557 (-0.100145)	0.122574 / 0.737135 (-0.614561)	0.078076 / 0.296338 (-0.218263)

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)	0.275381 / 0.215209 (0.060172)	2.713220 / 2.077655 (0.635565)	1.441940 / 1.504120 (-0.062179)	1.325545 / 1.541195 (-0.215650)	1.363859 / 1.468490 (-0.104631)	0.715147 / 4.584777 (-3.869630)	2.356482 / 3.745712 (-1.389230)	2.882792 / 5.269862 (-2.387069)	1.833399 / 4.565676 (-2.732278)	0.077872 / 0.424275 (-0.346403)	0.005172 / 0.007607 (-0.002435)	0.326361 / 0.226044 (0.100316)	3.239202 / 2.268929 (0.970273)	1.837745 / 55.444624 (-53.606879)	1.517299 / 6.876477 (-5.359178)	1.552938 / 2.142072 (-0.589134)	0.801496 / 4.805227 (-4.003731)	0.133351 / 6.500664 (-6.367314)	0.042052 / 0.075469 (-0.033418)

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)	0.957887 / 1.841788 (-0.883901)	11.625291 / 8.074308 (3.550983)	9.679413 / 10.191392 (-0.511979)	0.140271 / 0.680424 (-0.540153)	0.013991 / 0.534201 (-0.520210)	0.299874 / 0.579283 (-0.279409)	0.267164 / 0.434364 (-0.167200)	0.338143 / 0.540337 (-0.202194)	0.434105 / 1.386936 (-0.952831)

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)	0.005833 / 0.011353 (-0.005520)	0.003761 / 0.011008 (-0.007247)	0.049699 / 0.038508 (0.011191)	0.032786 / 0.023109 (0.009677)	0.265100 / 0.275898 (-0.010798)	0.291045 / 0.323480 (-0.032435)	0.004281 / 0.007986 (-0.003705)	0.002737 / 0.004328 (-0.001591)	0.048524 / 0.004250 (0.044274)	0.040783 / 0.037052 (0.003731)	0.281122 / 0.258489 (0.022633)	0.311349 / 0.293841 (0.017508)	0.032143 / 0.128546 (-0.096403)	0.011747 / 0.075646 (-0.063899)	0.059432 / 0.419271 (-0.359840)	0.034362 / 0.043533 (-0.009171)	0.261061 / 0.255139 (0.005922)	0.279536 / 0.283200 (-0.003663)	0.019172 / 0.141683 (-0.122510)	1.160069 / 1.452155 (-0.292086)	1.224160 / 1.492716 (-0.268556)

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)	0.093596 / 0.018006 (0.075590)	0.302862 / 0.000490 (0.302372)	0.000208 / 0.000200 (0.000008)	0.000047 / 0.000054 (-0.000007)

Benchmark: benchmark_indices_mapping.json

metric	select	shard	shuffle	sort	train_test_split
new / old (diff)	0.022785 / 0.037411 (-0.014626)	0.079263 / 0.014526 (0.064737)	0.091340 / 0.176557 (-0.085216)	0.129453 / 0.737135 (-0.607682)	0.091349 / 0.296338 (-0.204989)

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)	0.298166 / 0.215209 (0.082957)	3.003146 / 2.077655 (0.925491)	1.575903 / 1.504120 (0.071783)	1.445231 / 1.541195 (-0.095963)	1.477116 / 1.468490 (0.008625)	0.726496 / 4.584777 (-3.858281)	0.959827 / 3.745712 (-2.785885)	2.941142 / 5.269862 (-2.328720)	1.878581 / 4.565676 (-2.687096)	0.078475 / 0.424275 (-0.345800)	0.005137 / 0.007607 (-0.002470)	0.352078 / 0.226044 (0.126034)	3.486113 / 2.268929 (1.217184)	1.965024 / 55.444624 (-53.479600)	1.667223 / 6.876477 (-5.209254)	1.665254 / 2.142072 (-0.476819)	0.803543 / 4.805227 (-4.001684)	0.133003 / 6.500664 (-6.367661)	0.041462 / 0.075469 (-0.034008)

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)	1.045534 / 1.841788 (-0.796254)	12.124988 / 8.074308 (4.050680)	10.418723 / 10.191392 (0.227331)	0.142453 / 0.680424 (-0.537971)	0.015686 / 0.534201 (-0.518515)	0.300557 / 0.579283 (-0.278726)	0.119851 / 0.434364 (-0.314512)	0.342297 / 0.540337 (-0.198040)	0.441263 / 1.386936 (-0.945673)

[Pierrci]

lgtm!