Fix missing info when loading some datasets from Parquet export

[lhoestq]

Fix getting the info for script-based datasets with Parquet export with a single config not named "default".

E.g.

from datasets import load_dataset_builder

b = load_dataset_builder("bookcorpus")
print(b.info.features)
# should print {'text': Value(dtype='string', id=None)}

I fixed this by setting the default config name when there is only one config.

[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]

Thanks for the fix! Cool.

Just a nit formal comment below, but feel free to ignore it.

[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)	0.005577 / 0.011353 (-0.005776)	0.004452 / 0.011008 (-0.006556)	0.067849 / 0.038508 (0.029341)	0.032328 / 0.023109 (0.009219)	0.256924 / 0.275898 (-0.018974)	0.273410 / 0.323480 (-0.050070)	0.004359 / 0.007986 (-0.003626)	0.003484 / 0.004328 (-0.000845)	0.053880 / 0.004250 (0.049630)	0.058142 / 0.037052 (0.021089)	0.268863 / 0.258489 (0.010374)	0.307977 / 0.293841 (0.014136)	0.028840 / 0.128546 (-0.099707)	0.011808 / 0.075646 (-0.063839)	0.216277 / 0.419271 (-0.202995)	0.039245 / 0.043533 (-0.004288)	0.250420 / 0.255139 (-0.004719)	0.273642 / 0.283200 (-0.009557)	0.019340 / 0.141683 (-0.122342)	1.176734 / 1.452155 (-0.275421)	1.250643 / 1.492716 (-0.242074)

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)	0.181210 / 0.018006 (0.163204)	1.070750 / 0.000490 (1.070261)	0.000315 / 0.000200 (0.000115)	0.000045 / 0.000054 (-0.000009)

Benchmark: benchmark_indices_mapping.json

metric	select	shard	shuffle	sort	train_test_split
new / old (diff)	0.022905 / 0.037411 (-0.014507)	0.064549 / 0.014526 (0.050023)	0.077113 / 0.176557 (-0.099443)	0.131976 / 0.737135 (-0.605159)	0.081266 / 0.296338 (-0.215072)

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)	0.291363 / 0.215209 (0.076154)	2.851691 / 2.077655 (0.774036)	1.592815 / 1.504120 (0.088695)	1.494550 / 1.541195 (-0.046645)	1.516464 / 1.468490 (0.047974)	0.583244 / 4.584777 (-4.001532)	2.504907 / 3.745712 (-1.240805)	3.183490 / 5.269862 (-2.086371)	1.932854 / 4.565676 (-2.632823)	0.067564 / 0.424275 (-0.356711)	0.006587 / 0.007607 (-0.001020)	0.346368 / 0.226044 (0.120324)	3.428256 / 2.268929 (1.159327)	1.994176 / 55.444624 (-53.450448)	1.688116 / 6.876477 (-5.188360)	1.767653 / 2.142072 (-0.374420)	0.673867 / 4.805227 (-4.131360)	0.125582 / 6.500664 (-6.375082)	0.047198 / 0.075469 (-0.028271)

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)	1.002895 / 1.841788 (-0.838893)	16.332893 / 8.074308 (8.258585)	10.781993 / 10.191392 (0.590601)	0.153919 / 0.680424 (-0.526505)	0.015528 / 0.534201 (-0.518673)	0.306182 / 0.579283 (-0.273101)	0.296380 / 0.434364 (-0.137984)	0.341432 / 0.540337 (-0.198905)	0.455900 / 1.386936 (-0.931036)

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)	0.006442 / 0.011353 (-0.004911)	0.004433 / 0.011008 (-0.006576)	0.053327 / 0.038508 (0.014819)	0.035966 / 0.023109 (0.012856)	0.280913 / 0.275898 (0.005015)	0.308419 / 0.323480 (-0.015061)	0.005842 / 0.007986 (-0.002144)	0.003789 / 0.004328 (-0.000539)	0.053983 / 0.004250 (0.049732)	0.069052 / 0.037052 (0.032000)	0.299225 / 0.258489 (0.040736)	0.336470 / 0.293841 (0.042629)	0.068170 / 0.128546 (-0.060377)	0.012259 / 0.075646 (-0.063388)	0.064166 / 0.419271 (-0.355106)	0.037291 / 0.043533 (-0.006241)	0.281318 / 0.255139 (0.026179)	0.297093 / 0.283200 (0.013893)	0.021358 / 0.141683 (-0.120324)	1.189584 / 1.452155 (-0.262571)	1.256985 / 1.492716 (-0.235731)

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)	0.216726 / 0.018006 (0.198720)	2.496957 / 0.000490 (2.496467)	0.000336 / 0.000200 (0.000136)	0.000070 / 0.000054 (0.000016)

Benchmark: benchmark_indices_mapping.json

metric	select	shard	shuffle	sort	train_test_split
new / old (diff)	0.026604 / 0.037411 (-0.010807)	0.080398 / 0.014526 (0.065873)	0.094475 / 0.176557 (-0.082082)	0.136263 / 0.737135 (-0.600873)	0.097898 / 0.296338 (-0.198440)

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)	0.295171 / 0.215209 (0.079962)	2.947530 / 2.077655 (0.869875)	1.607531 / 1.504120 (0.103411)	1.485045 / 1.541195 (-0.056150)	1.524899 / 1.468490 (0.056409)	0.572934 / 4.584777 (-4.011843)	2.544320 / 3.745712 (-1.201393)	3.292630 / 5.269862 (-1.977232)	1.927138 / 4.565676 (-2.638539)	0.068560 / 0.424275 (-0.355715)	0.005982 / 0.007607 (-0.001625)	0.345833 / 0.226044 (0.119789)	3.424253 / 2.268929 (1.155324)	2.195017 / 55.444624 (-53.249608)	1.712037 / 6.876477 (-5.164440)	1.763899 / 2.142072 (-0.378174)	0.653776 / 4.805227 (-4.151451)	0.123056 / 6.500664 (-6.377609)	0.044572 / 0.075469 (-0.030897)

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)	1.033400 / 1.841788 (-0.808388)	15.409887 / 8.074308 (7.335579)	11.220990 / 10.191392 (1.029597)	0.153603 / 0.680424 (-0.526821)	0.016866 / 0.534201 (-0.517335)	0.311945 / 0.579283 (-0.267338)	0.307048 / 0.434364 (-0.127316)	0.350422 / 0.540337 (-0.189915)	0.447308 / 1.386936 (-0.939628)