Datasets:

pavel-nesterov

/

pavel-github-issues

like 0

[ "Hi! Yes, we should deprecate the `fs` param here. Would you be interested in submitting a PR? ", "> Hi! Yes, we should deprecate the `fs` param here. Would you be interested in submitting a PR?\r\n\r\nYeah I can do that sometime next week. Should the storage_options be a new arg here? I’ll look around for anywhere else where fs is an arg.", "Closed by #5393." ]

[ "_The documentation is not available anymore as the PR was closed or merged._", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.010576 / 0.011353 (-0.000777) | 0.006010 / 0.011008 (-0.004998) | 0.109375 / 0.038508 (0.070867) | 0.037780 / 0.023109 (0.014670) | 0.381552 / 0.275898 (0.105654) | 0.446039 / 0.323480 (0.122559) | 0.009004 / 0.007986 (0.001019) | 0.005653 / 0.004328 (0.001324) | 0.087027 / 0.004250 (0.082776) | 0.040346 / 0.037052 (0.003293) | 0.398827 / 0.258489 (0.140338) | 0.407281 / 0.293841 (0.113440) | 0.051723 / 0.128546 (-0.076824) | 0.020254 / 0.075646 (-0.055392) | 0.376841 / 0.419271 (-0.042430) | 0.055505 / 0.043533 (0.011972) | 0.383464 / 0.255139 (0.128325) | 0.436130 / 0.283200 (0.152930) | 0.117403 / 0.141683 (-0.024280) | 1.569016 / 1.452155 (0.116862) | 1.889831 / 1.492716 (0.397115) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.297962 / 0.018006 (0.279956) | 0.683699 / 0.000490 (0.683210) | 0.000918 / 0.000200 (0.000718) | 0.000100 / 0.000054 (0.000045) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.026742 / 0.037411 (-0.010669) | 0.125293 / 0.014526 (0.110768) | 0.128769 / 0.176557 (-0.047787) | 0.179447 / 0.737135 (-0.557688) | 0.142032 / 0.296338 (-0.154306) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.588389 / 0.215209 (0.373180) | 5.943514 / 2.077655 (3.865859) | 2.631163 / 1.504120 (1.127043) | 1.865446 / 1.541195 (0.324252) | 2.055610 / 1.468490 (0.587120) | 1.090288 / 4.584777 (-3.494489) | 5.457151 / 3.745712 (1.711439) | 5.645614 / 5.269862 (0.375752) | 2.849492 / 4.565676 (-1.716184) | 0.140447 / 0.424275 (-0.283828) | 0.015421 / 0.007607 (0.007813) | 0.735528 / 0.226044 (0.509484) | 7.394097 / 2.268929 (5.125169) | 3.219714 / 55.444624 (-52.224911) | 2.504134 / 6.876477 (-4.372342) | 2.524291 / 2.142072 (0.382219) | 1.452776 / 4.805227 (-3.352452) | 0.256142 / 6.500664 (-6.244522) | 0.093809 / 0.075469 (0.018340) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.570046 / 1.841788 (-0.271742) | 17.360385 / 8.074308 (9.286077) | 20.750595 / 10.191392 (10.559203) | 0.218486 / 0.680424 (-0.461938) | 0.048527 / 0.534201 (-0.485674) | 0.549568 / 0.579283 (-0.029715) | 0.633993 / 0.434364 (0.199629) | 0.632585 / 0.540337 (0.092248) | 0.712817 / 1.386936 (-0.674119) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.010524 / 0.011353 (-0.000829) | 0.006307 / 0.011008 (-0.004701) | 0.129671 / 0.038508 (0.091162) | 0.038952 / 0.023109 (0.015842) | 0.421936 / 0.275898 (0.146038) | 0.489911 / 0.323480 (0.166431) | 0.007661 / 0.007986 (-0.000325) | 0.005430 / 0.004328 (0.001102) | 0.091851 / 0.004250 (0.087600) | 0.059755 / 0.037052 (0.022703) | 0.449810 / 0.258489 (0.191321) | 0.519498 / 0.293841 (0.225657) | 0.061644 / 0.128546 (-0.066902) | 0.018950 / 0.075646 (-0.056696) | 0.399149 / 0.419271 (-0.020122) | 0.067670 / 0.043533 (0.024137) | 0.441091 / 0.255139 (0.185952) | 0.459327 / 0.283200 (0.176128) | 0.122476 / 0.141683 (-0.019207) | 1.760129 / 1.452155 (0.307974) | 1.767945 / 1.492716 (0.275228) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.276675 / 0.018006 (0.258669) | 0.606798 / 0.000490 (0.606308) | 0.000449 / 0.000200 (0.000249) | 0.000078 / 0.000054 (0.000023) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.027762 / 0.037411 (-0.009649) | 0.108330 / 0.014526 (0.093805) | 0.134714 / 0.176557 (-0.041843) | 0.175666 / 0.737135 (-0.561470) | 0.134917 / 0.296338 (-0.161421) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.676756 / 0.215209 (0.461547) | 6.746519 / 2.077655 (4.668864) | 2.660869 / 1.504120 (1.156750) | 2.273688 / 1.541195 (0.732494) | 2.392580 / 1.468490 (0.924090) | 1.127848 / 4.584777 (-3.456929) | 5.356499 / 3.745712 (1.610787) | 2.933006 / 5.269862 (-2.336855) | 1.872877 / 4.565676 (-2.692799) | 0.139504 / 0.424275 (-0.284771) | 0.013501 / 0.007607 (0.005894) | 0.749888 / 0.226044 (0.523843) | 8.157031 / 2.268929 (5.888103) | 3.627751 / 55.444624 (-51.816874) | 2.713152 / 6.876477 (-4.163324) | 2.934585 / 2.142072 (0.792512) | 1.376398 / 4.805227 (-3.428829) | 0.251537 / 6.500664 (-6.249127) | 0.083995 / 0.075469 (0.008526) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.635446 / 1.841788 (-0.206342) | 18.435807 / 8.074308 (10.361498) | 21.395291 / 10.191392 (11.203899) | 0.247238 / 0.680424 (-0.433186) | 0.030503 / 0.534201 (-0.503698) | 0.553096 / 0.579283 (-0.026187) | 0.597583 / 0.434364 (0.163219) | 0.594135 / 0.540337 (0.053797) | 0.673815 / 1.386936 (-0.713122) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png \"CML watermark\")\n" ]

[ "Another example is that `IterableDataset.map` does not have `fn_kwargs`, among other arguments. It makes it harder to convert code from Dataset to IterableDataset.", "Hi! `fn_kwargs` was added to `IterableDataset.map` in `datasets 2.5.0`, so please update your installation (`pip install -U datasets`) to use it.\r\n\r\nRegarding `column_names`, I agree we should add this property to `IterableDataset`. In the meantime, you can use `list(dataset.features.keys())` instead.", "Thanks! That's great news.\n\nOn Thu, Dec 22, 2022, 07:48 Mario Šaško ***@***.***> wrote:\n\n> Hi! fn_kwargs was added to IterableDataset.map in datasets 2.5.0, so\n> please update your installation (pip install -U datasets) to use it.\n>\n> Regarding column_names, I agree we should add this property to\n> IterableDataset. In the meantime, you can use\n> list(dataset.features.keys()) instead.\n>\n> —\n> Reply to this email directly, view it on GitHub\n> <https://github.com/huggingface/datasets/issues/5383#issuecomment-1362993633>,\n> or unsubscribe\n> <https://github.com/notifications/unsubscribe-auth/AAHD6N2EQUFEOUFDW3VHSILWORZ45ANCNFSM6AAAAAATGKWVGM>\n> .\n> You are receiving this because you authored the thread.Message ID:\n> ***@***.***>\n>\n" ]

[ "_The documentation is not available anymore as the PR was closed or merged._", "Could you review this small PR @albertvillanova ? :)", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.011200 / 0.011353 (-0.000153) | 0.006156 / 0.011008 (-0.004852) | 0.119072 / 0.038508 (0.080564) | 0.042616 / 0.023109 (0.019507) | 0.348329 / 0.275898 (0.072431) | 0.418550 / 0.323480 (0.095070) | 0.009302 / 0.007986 (0.001316) | 0.004596 / 0.004328 (0.000267) | 0.090111 / 0.004250 (0.085860) | 0.053341 / 0.037052 (0.016289) | 0.361234 / 0.258489 (0.102745) | 0.400427 / 0.293841 (0.106586) | 0.045601 / 0.128546 (-0.082945) | 0.013806 / 0.075646 (-0.061841) | 0.393178 / 0.419271 (-0.026094) | 0.056809 / 0.043533 (0.013276) | 0.344090 / 0.255139 (0.088951) | 0.370610 / 0.283200 (0.087410) | 0.125728 / 0.141683 (-0.015955) | 1.671931 / 1.452155 (0.219776) | 1.703143 / 1.492716 (0.210427) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.226534 / 0.018006 (0.208527) | 0.496487 / 0.000490 (0.495998) | 0.002235 / 0.000200 (0.002035) | 0.000094 / 0.000054 (0.000039) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.031298 / 0.037411 (-0.006113) | 0.137740 / 0.014526 (0.123214) | 0.153497 / 0.176557 (-0.023059) | 0.204201 / 0.737135 (-0.532934) | 0.162324 / 0.296338 (-0.134014) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.475922 / 0.215209 (0.260712) | 4.682344 / 2.077655 (2.604689) | 2.107387 / 1.504120 (0.603267) | 1.884792 / 1.541195 (0.343597) | 2.003180 / 1.468490 (0.534690) | 0.810212 / 4.584777 (-3.774564) | 4.631047 / 3.745712 (0.885334) | 4.467606 / 5.269862 (-0.802256) | 2.334196 / 4.565676 (-2.231480) | 0.099713 / 0.424275 (-0.324562) | 0.014732 / 0.007607 (0.007125) | 0.604587 / 0.226044 (0.378543) | 5.951679 / 2.268929 (3.682751) | 2.704761 / 55.444624 (-52.739863) | 2.280695 / 6.876477 (-4.595781) | 2.279489 / 2.142072 (0.137417) | 0.962474 / 4.805227 (-3.842753) | 0.195279 / 6.500664 (-6.305385) | 0.071503 / 0.075469 (-0.003966) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.558037 / 1.841788 (-0.283751) | 17.722140 / 8.074308 (9.647832) | 16.229016 / 10.191392 (6.037624) | 0.177148 / 0.680424 (-0.503276) | 0.034162 / 0.534201 (-0.500039) | 0.513945 / 0.579283 (-0.065338) | 0.533542 / 0.434364 (0.099178) | 0.672457 / 0.540337 (0.132119) | 0.762390 / 1.386936 (-0.624546) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.009739 / 0.011353 (-0.001613) | 0.006095 / 0.011008 (-0.004914) | 0.105968 / 0.038508 (0.067460) | 0.046229 / 0.023109 (0.023120) | 0.449156 / 0.275898 (0.173258) | 0.462182 / 0.323480 (0.138702) | 0.006981 / 0.007986 (-0.001004) | 0.004867 / 0.004328 (0.000539) | 0.082142 / 0.004250 (0.077891) | 0.058652 / 0.037052 (0.021600) | 0.454542 / 0.258489 (0.196052) | 0.494910 / 0.293841 (0.201069) | 0.047159 / 0.128546 (-0.081387) | 0.014677 / 0.075646 (-0.060969) | 0.370819 / 0.419271 (-0.048452) | 0.064603 / 0.043533 (0.021070) | 0.441514 / 0.255139 (0.186375) | 0.442802 / 0.283200 (0.159603) | 0.138603 / 0.141683 (-0.003080) | 1.692810 / 1.452155 (0.240655) | 1.894596 / 1.492716 (0.401880) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.281681 / 0.018006 (0.263675) | 0.532693 / 0.000490 (0.532203) | 0.005484 / 0.000200 (0.005284) | 0.000156 / 0.000054 (0.000102) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.032994 / 0.037411 (-0.004417) | 0.134614 / 0.014526 (0.120088) | 0.142286 / 0.176557 (-0.034270) | 0.187220 / 0.737135 (-0.549916) | 0.144897 / 0.296338 (-0.151441) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.519536 / 0.215209 (0.304327) | 5.214429 / 2.077655 (3.136775) | 2.612575 / 1.504120 (1.108455) | 2.369085 / 1.541195 (0.827891) | 2.503157 / 1.468490 (1.034667) | 0.834827 / 4.584777 (-3.749950) | 4.586789 / 3.745712 (0.841077) | 4.472605 / 5.269862 (-0.797257) | 2.314471 / 4.565676 (-2.251205) | 0.095817 / 0.424275 (-0.328458) | 0.014086 / 0.007607 (0.006478) | 0.605875 / 0.226044 (0.379831) | 6.153143 / 2.268929 (3.884214) | 3.187456 / 55.444624 (-52.257169) | 2.755377 / 6.876477 (-4.121100) | 2.777118 / 2.142072 (0.635046) | 0.967285 / 4.805227 (-3.837942) | 0.199202 / 6.500664 (-6.301462) | 0.075979 / 0.075469 (0.000510) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.481758 / 1.841788 (-0.360030) | 18.053769 / 8.074308 (9.979461) | 15.558780 / 10.191392 (5.367388) | 0.226135 / 0.680424 (-0.454288) | 0.021668 / 0.534201 (-0.512533) | 0.562618 / 0.579283 (-0.016666) | 0.518183 / 0.434364 (0.083819) | 0.628580 / 0.540337 (0.088243) | 0.740368 / 1.386936 (-0.646568) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#4e4d46eec24c36799c0efcc1b7231f597039c497 \"CML watermark\")\n" ]

[ "Hi! This error can happen if there is a local file/folder with the same name as the requested dataset. And to avoid it, rename the local file/folder.\r\n\r\nSoon, it will be possible to explicitly request a Hub dataset as follows:https://github.com/huggingface/datasets/issues/5228#issuecomment-1313494020" ]

[ "Hi! I agree `skip` can be inefficient to use in the current state.\r\n\r\nTo make it fast, we could use \"statistics\" stored in Parquet metadata and read only the chunks needed to form a dataset. \r\n\r\nAnd thanks to the \"datasets-server\" project, which aims to store the Parquet versions of the Hub datasets (only the smaller datasets are covered currently), this solution can also be applied to datasets stored in formats other than Parquet. (cc @severo)", "@mariosasko do the current parquet files created by the datasets-server already have the required \"statistics\"? If not, please open an issue on https://github.com/huggingface/datasets-server with some details to make sure we implement it.", "Yes, nothing has to be changed on the datasets-server side. What I mean by \"statistics\" is that we can use the \"row_group\" metadata embedded in a Parquet file (by default) to fetch the requested rows more efficiently.", "Glad to see the feature could be of interest. \r\n\r\nI'm sure there are many possible ways to implement this feature. I don't know enough about the datasets-server, but I guess that it is not instantaneous, in the sense that user-owned private datasets might need hours or days until they are ported to the datasets-server (if at all), which could be cumbersome. Having optionally that information in the `dataset_infos.json` file would make it easier for users to control the skip process a bit.", "re: statistics:\r\n\r\n- https://arrow.apache.org/docs/python/generated/pyarrow.parquet.FileMetaData.html\r\n- https://arrow.apache.org/docs/python/generated/pyarrow.parquet.RowGroupMetaData.html\r\n\r\n```python\r\n>>> import pyarrow.parquet as pq\r\n>>> import hffs\r\n>>> fs = hffs.HfFileSystem(\"glue\", repo_type=\"dataset\", revision=\"refs/convert/parquet\")\r\n>>> metadata = pq.read_metadata(\"ax/glue-test.parquet\", filesystem=fs)\r\n>>> metadata\r\n<pyarrow._parquet.FileMetaData object at 0x7f4537cec400>\r\n created_by: parquet-cpp-arrow version 7.0.0\r\n num_columns: 4\r\n num_rows: 1104\r\n num_row_groups: 2\r\n format_version: 1.0\r\n serialized_size: 2902\r\n>>> metadata.row_group(0)\r\n<pyarrow._parquet.RowGroupMetaData object at 0x7f45564bcbd0>\r\n num_columns: 4\r\n num_rows: 1000\r\n total_byte_size: 164474\r\n>>> metadata.row_group(1)\r\n<pyarrow._parquet.RowGroupMetaData object at 0x7f455005c400>\r\n num_columns: 4\r\n num_rows: 104\r\n total_byte_size: 13064\r\n```", "> user-owned private datasets might need hours or days until they are ported to the datasets-server (if at all)\r\n\r\nprivate datasets are not supported yet (https://github.com/huggingface/datasets-server/issues/39)", "@versae `Dataset.push_to_hub` writes shards in Parquet, so this solution would also work for such datasets (immediately after the push). ", "@mariosasko that is right. However, there are still a good amount of datasets for which the shards are created manually. In our very specific case, we create medium-sized datasets (rarely over 100-200GB) of both text and audio, we prepare the shards by hand and then upload then. It would be great to have immediate access to this download skipping feature for them too." ]

[ "_The documentation is not available anymore as the PR was closed or merged._", "Thanks for the changes, looks good to me!", "@stevhliu I have pushed some quality improvements both in terms of code and content. Would you be able to re-review? ", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008325 / 0.011353 (-0.003028) | 0.004432 / 0.011008 (-0.006576) | 0.099794 / 0.038508 (0.061286) | 0.029469 / 0.023109 (0.006360) | 0.306554 / 0.275898 (0.030656) | 0.367373 / 0.323480 (0.043893) | 0.007532 / 0.007986 (-0.000454) | 0.003310 / 0.004328 (-0.001018) | 0.077453 / 0.004250 (0.073203) | 0.034836 / 0.037052 (-0.002216) | 0.311696 / 0.258489 (0.053207) | 0.349683 / 0.293841 (0.055842) | 0.033089 / 0.128546 (-0.095457) | 0.011339 / 0.075646 (-0.064307) | 0.321699 / 0.419271 (-0.097573) | 0.040213 / 0.043533 (-0.003320) | 0.304741 / 0.255139 (0.049602) | 0.331569 / 0.283200 (0.048369) | 0.090397 / 0.141683 (-0.051285) | 1.526001 / 1.452155 (0.073847) | 1.558863 / 1.492716 (0.066146) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.179446 / 0.018006 (0.161440) | 0.416308 / 0.000490 (0.415818) | 0.002390 / 0.000200 (0.002190) | 0.000075 / 0.000054 (0.000021) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.023641 / 0.037411 (-0.013770) | 0.096672 / 0.014526 (0.082147) | 0.104330 / 0.176557 (-0.072227) | 0.146338 / 0.737135 (-0.590797) | 0.108278 / 0.296338 (-0.188060) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.420194 / 0.215209 (0.204985) | 4.196981 / 2.077655 (2.119326) | 1.861206 / 1.504120 (0.357086) | 1.658748 / 1.541195 (0.117554) | 1.704309 / 1.468490 (0.235819) | 0.691639 / 4.584777 (-3.893138) | 3.346303 / 3.745712 (-0.399409) | 1.932962 / 5.269862 (-3.336900) | 1.299395 / 4.565676 (-3.266281) | 0.081869 / 0.424275 (-0.342406) | 0.012415 / 0.007607 (0.004808) | 0.530805 / 0.226044 (0.304761) | 5.293486 / 2.268929 (3.024558) | 2.328327 / 55.444624 (-53.116297) | 1.964956 / 6.876477 (-4.911521) | 2.002793 / 2.142072 (-0.139280) | 0.813380 / 4.805227 (-3.991847) | 0.150030 / 6.500664 (-6.350634) | 0.065194 / 0.075469 (-0.010275) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.259421 / 1.841788 (-0.582367) | 13.667796 / 8.074308 (5.593488) | 13.819121 / 10.191392 (3.627729) | 0.136718 / 0.680424 (-0.543706) | 0.028510 / 0.534201 (-0.505691) | 0.402246 / 0.579283 (-0.177037) | 0.405279 / 0.434364 (-0.029085) | 0.467185 / 0.540337 (-0.073153) | 0.554213 / 1.386936 (-0.832723) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.006738 / 0.011353 (-0.004615) | 0.004616 / 0.011008 (-0.006393) | 0.096978 / 0.038508 (0.058470) | 0.027750 / 0.023109 (0.004640) | 0.411505 / 0.275898 (0.135607) | 0.441796 / 0.323480 (0.118316) | 0.005073 / 0.007986 (-0.002913) | 0.003360 / 0.004328 (-0.000968) | 0.074445 / 0.004250 (0.070194) | 0.040654 / 0.037052 (0.003602) | 0.414277 / 0.258489 (0.155788) | 0.448665 / 0.293841 (0.154824) | 0.032346 / 0.128546 (-0.096200) | 0.011533 / 0.075646 (-0.064114) | 0.317349 / 0.419271 (-0.101923) | 0.041934 / 0.043533 (-0.001599) | 0.409102 / 0.255139 (0.153963) | 0.429977 / 0.283200 (0.146777) | 0.089459 / 0.141683 (-0.052224) | 1.518127 / 1.452155 (0.065973) | 1.569902 / 1.492716 (0.077186) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.232648 / 0.018006 (0.214642) | 0.413751 / 0.000490 (0.413261) | 0.000404 / 0.000200 (0.000204) | 0.000057 / 0.000054 (0.000003) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.025468 / 0.037411 (-0.011943) | 0.098195 / 0.014526 (0.083669) | 0.108882 / 0.176557 (-0.067674) | 0.150059 / 0.737135 (-0.587076) | 0.110742 / 0.296338 (-0.185597) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.445326 / 0.215209 (0.230117) | 4.449200 / 2.077655 (2.371545) | 2.098939 / 1.504120 (0.594819) | 1.861207 / 1.541195 (0.320012) | 1.901385 / 1.468490 (0.432894) | 0.695287 / 4.584777 (-3.889490) | 3.461775 / 3.745712 (-0.283938) | 2.998566 / 5.269862 (-2.271296) | 1.555036 / 4.565676 (-3.010641) | 0.082789 / 0.424275 (-0.341486) | 0.012772 / 0.007607 (0.005165) | 0.564855 / 0.226044 (0.338811) | 5.631049 / 2.268929 (3.362120) | 2.543771 / 55.444624 (-52.900854) | 2.194378 / 6.876477 (-4.682099) | 2.267168 / 2.142072 (0.125095) | 0.803330 / 4.805227 (-4.001898) | 0.151336 / 6.500664 (-6.349328) | 0.067015 / 0.075469 (-0.008454) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.298422 / 1.841788 (-0.543366) | 13.933637 / 8.074308 (5.859329) | 13.570848 / 10.191392 (3.379456) | 0.150787 / 0.680424 (-0.529637) | 0.016911 / 0.534201 (-0.517290) | 0.384771 / 0.579283 (-0.194512) | 0.397505 / 0.434364 (-0.036858) | 0.450931 / 0.540337 (-0.089406) | 0.534501 / 1.386936 (-0.852435) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png \"CML watermark\")\n", "@lhoestq @nateraw made some changes as per the comments. PTAL and approve as necessary. ", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.009037 / 0.011353 (-0.002316) | 0.004970 / 0.011008 (-0.006038) | 0.099223 / 0.038508 (0.060715) | 0.034935 / 0.023109 (0.011826) | 0.297027 / 0.275898 (0.021129) | 0.352861 / 0.323480 (0.029382) | 0.007558 / 0.007986 (-0.000427) | 0.003903 / 0.004328 (-0.000425) | 0.075663 / 0.004250 (0.071413) | 0.042577 / 0.037052 (0.005524) | 0.307182 / 0.258489 (0.048693) | 0.344237 / 0.293841 (0.050396) | 0.041438 / 0.128546 (-0.087108) | 0.012159 / 0.075646 (-0.063487) | 0.333771 / 0.419271 (-0.085501) | 0.047847 / 0.043533 (0.004314) | 0.290797 / 0.255139 (0.035658) | 0.320517 / 0.283200 (0.037318) | 0.098334 / 0.141683 (-0.043349) | 1.446187 / 1.452155 (-0.005968) | 1.495506 / 1.492716 (0.002789) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.203704 / 0.018006 (0.185698) | 0.441325 / 0.000490 (0.440835) | 0.001173 / 0.000200 (0.000973) | 0.000080 / 0.000054 (0.000026) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.026694 / 0.037411 (-0.010718) | 0.103819 / 0.014526 (0.089294) | 0.116377 / 0.176557 (-0.060179) | 0.158280 / 0.737135 (-0.578856) | 0.119797 / 0.296338 (-0.176541) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.405723 / 0.215209 (0.190514) | 4.047633 / 2.077655 (1.969979) | 1.805652 / 1.504120 (0.301532) | 1.611382 / 1.541195 (0.070187) | 1.663117 / 1.468490 (0.194627) | 0.692589 / 4.584777 (-3.892188) | 3.689970 / 3.745712 (-0.055742) | 2.089760 / 5.269862 (-3.180101) | 1.450576 / 4.565676 (-3.115101) | 0.085276 / 0.424275 (-0.338999) | 0.012042 / 0.007607 (0.004434) | 0.513159 / 0.226044 (0.287115) | 5.123235 / 2.268929 (2.854306) | 2.281864 / 55.444624 (-53.162761) | 1.926170 / 6.876477 (-4.950307) | 2.035093 / 2.142072 (-0.106979) | 0.857457 / 4.805227 (-3.947770) | 0.166088 / 6.500664 (-6.334576) | 0.062115 / 0.075469 (-0.013354) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.197776 / 1.841788 (-0.644012) | 14.674452 / 8.074308 (6.600144) | 14.275990 / 10.191392 (4.084598) | 0.170848 / 0.680424 (-0.509576) | 0.028613 / 0.534201 (-0.505588) | 0.438650 / 0.579283 (-0.140633) | 0.439323 / 0.434364 (0.004959) | 0.515090 / 0.540337 (-0.025247) | 0.614216 / 1.386936 (-0.772720) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.007159 / 0.011353 (-0.004194) | 0.005142 / 0.011008 (-0.005866) | 0.096953 / 0.038508 (0.058445) | 0.033036 / 0.023109 (0.009927) | 0.391790 / 0.275898 (0.115892) | 0.427120 / 0.323480 (0.103640) | 0.005691 / 0.007986 (-0.002294) | 0.004848 / 0.004328 (0.000519) | 0.072258 / 0.004250 (0.068008) | 0.049017 / 0.037052 (0.011965) | 0.387267 / 0.258489 (0.128778) | 0.437112 / 0.293841 (0.143272) | 0.036360 / 0.128546 (-0.092186) | 0.012249 / 0.075646 (-0.063397) | 0.336246 / 0.419271 (-0.083025) | 0.048777 / 0.043533 (0.005244) | 0.397872 / 0.255139 (0.142733) | 0.399768 / 0.283200 (0.116568) | 0.101283 / 0.141683 (-0.040400) | 1.443999 / 1.452155 (-0.008156) | 1.575496 / 1.492716 (0.082779) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.220952 / 0.018006 (0.202946) | 0.442220 / 0.000490 (0.441730) | 0.000406 / 0.000200 (0.000206) | 0.000058 / 0.000054 (0.000004) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.028626 / 0.037411 (-0.008786) | 0.109929 / 0.014526 (0.095403) | 0.120989 / 0.176557 (-0.055568) | 0.157377 / 0.737135 (-0.579758) | 0.125522 / 0.296338 (-0.170816) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.436565 / 0.215209 (0.221356) | 4.380771 / 2.077655 (2.303117) | 2.200003 / 1.504120 (0.695883) | 2.013289 / 1.541195 (0.472094) | 2.052658 / 1.468490 (0.584168) | 0.703706 / 4.584777 (-3.881071) | 3.823289 / 3.745712 (0.077577) | 2.064882 / 5.269862 (-3.204980) | 1.330834 / 4.565676 (-3.234842) | 0.085945 / 0.424275 (-0.338330) | 0.012511 / 0.007607 (0.004904) | 0.544171 / 0.226044 (0.318127) | 5.476059 / 2.268929 (3.207130) | 2.695586 / 55.444624 (-52.749039) | 2.330239 / 6.876477 (-4.546238) | 2.429290 / 2.142072 (0.287218) | 0.843154 / 4.805227 (-3.962073) | 0.169334 / 6.500664 (-6.331330) | 0.064261 / 0.075469 (-0.011209) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.268344 / 1.841788 (-0.573444) | 14.934342 / 8.074308 (6.860034) | 13.555389 / 10.191392 (3.363997) | 0.142725 / 0.680424 (-0.537699) | 0.017891 / 0.534201 (-0.516310) | 0.424833 / 0.579283 (-0.154450) | 0.420035 / 0.434364 (-0.014329) | 0.491009 / 0.540337 (-0.049329) | 0.586953 / 1.386936 (-0.799983) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png \"CML watermark\")\n", "Merging this PR with approvals from @stevhliu @lhoestq. ", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008586 / 0.011353 (-0.002767) | 0.004659 / 0.011008 (-0.006350) | 0.100343 / 0.038508 (0.061835) | 0.029861 / 0.023109 (0.006751) | 0.301090 / 0.275898 (0.025192) | 0.369528 / 0.323480 (0.046048) | 0.006920 / 0.007986 (-0.001065) | 0.003513 / 0.004328 (-0.000815) | 0.078514 / 0.004250 (0.074263) | 0.035285 / 0.037052 (-0.001767) | 0.311257 / 0.258489 (0.052768) | 0.353995 / 0.293841 (0.060154) | 0.033733 / 0.128546 (-0.094813) | 0.011489 / 0.075646 (-0.064157) | 0.323095 / 0.419271 (-0.096176) | 0.040808 / 0.043533 (-0.002725) | 0.301779 / 0.255139 (0.046640) | 0.348517 / 0.283200 (0.065318) | 0.086962 / 0.141683 (-0.054721) | 1.496270 / 1.452155 (0.044115) | 1.514260 / 1.492716 (0.021544) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.189502 / 0.018006 (0.171496) | 0.419326 / 0.000490 (0.418837) | 0.002160 / 0.000200 (0.001960) | 0.000084 / 0.000054 (0.000029) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.023669 / 0.037411 (-0.013742) | 0.096574 / 0.014526 (0.082048) | 0.105970 / 0.176557 (-0.070587) | 0.148531 / 0.737135 (-0.588605) | 0.109948 / 0.296338 (-0.186391) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.424968 / 0.215209 (0.209759) | 4.246292 / 2.077655 (2.168637) | 1.911062 / 1.504120 (0.406943) | 1.700733 / 1.541195 (0.159538) | 1.760756 / 1.468490 (0.292266) | 0.696966 / 4.584777 (-3.887811) | 3.372320 / 3.745712 (-0.373392) | 2.886281 / 5.269862 (-2.383581) | 1.553082 / 4.565676 (-3.012594) | 0.082835 / 0.424275 (-0.341440) | 0.012688 / 0.007607 (0.005081) | 0.536352 / 0.226044 (0.310308) | 5.382510 / 2.268929 (3.113582) | 2.365664 / 55.444624 (-53.078960) | 1.995631 / 6.876477 (-4.880845) | 2.073865 / 2.142072 (-0.068207) | 0.819109 / 4.805227 (-3.986118) | 0.150278 / 6.500664 (-6.350386) | 0.065201 / 0.075469 (-0.010268) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.239835 / 1.841788 (-0.601953) | 13.911847 / 8.074308 (5.837539) | 13.500433 / 10.191392 (3.309041) | 0.137153 / 0.680424 (-0.543271) | 0.028451 / 0.534201 (-0.505750) | 0.394659 / 0.579283 (-0.184625) | 0.404915 / 0.434364 (-0.029449) | 0.458944 / 0.540337 (-0.081394) | 0.542288 / 1.386936 (-0.844648) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.006791 / 0.011353 (-0.004562) | 0.004590 / 0.011008 (-0.006419) | 0.098697 / 0.038508 (0.060189) | 0.027634 / 0.023109 (0.004525) | 0.344383 / 0.275898 (0.068485) | 0.385607 / 0.323480 (0.062127) | 0.005413 / 0.007986 (-0.002573) | 0.003447 / 0.004328 (-0.000881) | 0.077268 / 0.004250 (0.073018) | 0.041823 / 0.037052 (0.004770) | 0.342904 / 0.258489 (0.084414) | 0.399371 / 0.293841 (0.105530) | 0.032668 / 0.128546 (-0.095879) | 0.011598 / 0.075646 (-0.064048) | 0.319973 / 0.419271 (-0.099299) | 0.041760 / 0.043533 (-0.001773) | 0.340510 / 0.255139 (0.085371) | 0.377929 / 0.283200 (0.094730) | 0.090889 / 0.141683 (-0.050793) | 1.496068 / 1.452155 (0.043913) | 1.574884 / 1.492716 (0.082168) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.230489 / 0.018006 (0.212483) | 0.425234 / 0.000490 (0.424745) | 0.000406 / 0.000200 (0.000206) | 0.000059 / 0.000054 (0.000004) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.024650 / 0.037411 (-0.012761) | 0.102706 / 0.014526 (0.088180) | 0.108017 / 0.176557 (-0.068539) | 0.143645 / 0.737135 (-0.593490) | 0.110556 / 0.296338 (-0.185782) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.468038 / 0.215209 (0.252829) | 4.670514 / 2.077655 (2.592860) | 2.446620 / 1.504120 (0.942500) | 2.241255 / 1.541195 (0.700060) | 2.286409 / 1.468490 (0.817919) | 0.698923 / 4.584777 (-3.885854) | 3.401121 / 3.745712 (-0.344592) | 1.892399 / 5.269862 (-3.377462) | 1.163101 / 4.565676 (-3.402575) | 0.082567 / 0.424275 (-0.341708) | 0.012662 / 0.007607 (0.005055) | 0.571262 / 0.226044 (0.345218) | 5.731740 / 2.268929 (3.462812) | 2.879649 / 55.444624 (-52.564975) | 2.533846 / 6.876477 (-4.342631) | 2.654789 / 2.142072 (0.512717) | 0.811345 / 4.805227 (-3.993882) | 0.152495 / 6.500664 (-6.348169) | 0.067748 / 0.075469 (-0.007721) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.267852 / 1.841788 (-0.573935) | 14.114920 / 8.074308 (6.040612) | 14.355403 / 10.191392 (4.164011) | 0.150393 / 0.680424 (-0.530031) | 0.016855 / 0.534201 (-0.517346) | 0.378710 / 0.579283 (-0.200573) | 0.385380 / 0.434364 (-0.048984) | 0.439054 / 0.540337 (-0.101284) | 0.524343 / 1.386936 (-0.862593) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png \"CML watermark\")\n" ]

[ "Thanks for reporting @shaoyuta. We are investigating it.\r\n\r\nWe are transferring the issue to \"the_pile\" Community tab on the Hub: https://huggingface.co/datasets/the_pile/discussions/4" ]

[ "_The documentation is not available anymore as the PR was closed or merged._", "Failing because the test server uses Py3.7 but the `SharedMemory` features require Py3.8! I forgot we still support 3.7 for another couple of months. I'm not sure exactly how to proceed, whether I should leave this PR until then, or just gate the feature behind a version check and skip the tests until the Python version catches up.", "I haven't played with `NumpyMultiprocessingGenerator` so I can't really help here, but this sounds promising :) Otherwise I think it's also fine to allow `num_workers` only for py>=3.8 for now. You can skip the test on 3.7 and make sure to raise an informative error if someone wants to use `num_workers` with 3.7", "Lots of comments here - I'll reply to the specific code comments underneath them, but in response to the general comments:\r\n\r\n@gante: I think this approach is much more performant than a `multiprocessing.Pool`. The reason is that when results are returned from a process `Pool`, the returned Python objects are pickled by the child processes, sent down a pipe and unpickled by the parent process. This creates a huge single-process bottleneck as the parent has to unpickle lots of large NumPy arrays, which is quite slow.\r\n\r\nWhen you use a `SharedMemory` approach, the data is just **there** for the parent process - the child and the parent are writing to exactly the same array in memory, and no pickling or unpickling occurs. This means the parent can just immediately copy the array (which is much faster than unpickling) and yield it to `tf.data`. We're taking advantage of the fact that we know the data is just big NumPy arrays and we don't need the full generality of `pickle`.\r\n\r\n@lhoestq: Sounds good! I'll add a clear error and skip the tests on Py<=3.7.", "Also, an extra technicality, just for information in case anyone looks at this PR later: Recent versions of Python allow [pickled objects to store out-of-band data](https://peps.python.org/pep-0574/). This allows for very efficient zero-copy unpickling of objects like NumPy arrays, with the unpickled object having a view on the same memory as the original. \r\n\r\nHowever, this explicitly does **not** work when the object is unpickled by a different process than the one that created it. For this to work you must explicitly allocate shared memory and create the array there, which pickle cannot handle for you. As a result, if you just benchmark unpickling vs copying of NumPy arrays it can seem like unpickling is very fast - but this is only true when the pickle was created in the unpickling process!", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008666 / 0.011353 (-0.002687) | 0.004624 / 0.011008 (-0.006384) | 0.099247 / 0.038508 (0.060739) | 0.029766 / 0.023109 (0.006657) | 0.303347 / 0.275898 (0.027449) | 0.370022 / 0.323480 (0.046542) | 0.007128 / 0.007986 (-0.000857) | 0.003446 / 0.004328 (-0.000883) | 0.076670 / 0.004250 (0.072420) | 0.038892 / 0.037052 (0.001840) | 0.313035 / 0.258489 (0.054546) | 0.350503 / 0.293841 (0.056662) | 0.033732 / 0.128546 (-0.094815) | 0.011644 / 0.075646 (-0.064003) | 0.323295 / 0.419271 (-0.095977) | 0.040336 / 0.043533 (-0.003196) | 0.302253 / 0.255139 (0.047114) | 0.337199 / 0.283200 (0.053999) | 0.089454 / 0.141683 (-0.052229) | 1.624906 / 1.452155 (0.172752) | 1.546187 / 1.492716 (0.053470) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.184614 / 0.018006 (0.166608) | 0.427397 / 0.000490 (0.426907) | 0.003342 / 0.000200 (0.003142) | 0.000079 / 0.000054 (0.000025) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.023684 / 0.037411 (-0.013727) | 0.100095 / 0.014526 (0.085569) | 0.104996 / 0.176557 (-0.071560) | 0.144719 / 0.737135 (-0.592416) | 0.110759 / 0.296338 (-0.185579) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.421108 / 0.215209 (0.205899) | 4.214094 / 2.077655 (2.136440) | 1.906231 / 1.504120 (0.402111) | 1.698000 / 1.541195 (0.156806) | 1.744856 / 1.468490 (0.276366) | 0.693671 / 4.584777 (-3.891106) | 3.362522 / 3.745712 (-0.383190) | 1.878470 / 5.269862 (-3.391392) | 1.167563 / 4.565676 (-3.398113) | 0.082455 / 0.424275 (-0.341820) | 0.012261 / 0.007607 (0.004654) | 0.525196 / 0.226044 (0.299152) | 5.257553 / 2.268929 (2.988624) | 2.298286 / 55.444624 (-53.146339) | 1.956106 / 6.876477 (-4.920371) | 2.006308 / 2.142072 (-0.135764) | 0.811069 / 4.805227 (-3.994158) | 0.150368 / 6.500664 (-6.350296) | 0.065699 / 0.075469 (-0.009771) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.224516 / 1.841788 (-0.617272) | 13.619084 / 8.074308 (5.544776) | 14.096666 / 10.191392 (3.905274) | 0.151068 / 0.680424 (-0.529356) | 0.028819 / 0.534201 (-0.505382) | 0.402071 / 0.579283 (-0.177212) | 0.408647 / 0.434364 (-0.025717) | 0.466605 / 0.540337 (-0.073733) | 0.547094 / 1.386936 (-0.839842) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.006935 / 0.011353 (-0.004418) | 0.004590 / 0.011008 (-0.006419) | 0.099398 / 0.038508 (0.060890) | 0.028145 / 0.023109 (0.005036) | 0.426582 / 0.275898 (0.150684) | 0.465712 / 0.323480 (0.142233) | 0.005254 / 0.007986 (-0.002731) | 0.004956 / 0.004328 (0.000627) | 0.075616 / 0.004250 (0.071365) | 0.039871 / 0.037052 (0.002819) | 0.428859 / 0.258489 (0.170370) | 0.470839 / 0.293841 (0.176998) | 0.032150 / 0.128546 (-0.096396) | 0.011778 / 0.075646 (-0.063868) | 0.322358 / 0.419271 (-0.096913) | 0.041974 / 0.043533 (-0.001559) | 0.427459 / 0.255139 (0.172320) | 0.446685 / 0.283200 (0.163485) | 0.092000 / 0.141683 (-0.049683) | 1.509231 / 1.452155 (0.057076) | 1.578950 / 1.492716 (0.086234) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.168047 / 0.018006 (0.150041) | 0.418993 / 0.000490 (0.418503) | 0.002855 / 0.000200 (0.002655) | 0.000080 / 0.000054 (0.000026) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.025652 / 0.037411 (-0.011759) | 0.100141 / 0.014526 (0.085616) | 0.107293 / 0.176557 (-0.069264) | 0.142857 / 0.737135 (-0.594278) | 0.110933 / 0.296338 (-0.185406) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.477556 / 0.215209 (0.262347) | 4.777951 / 2.077655 (2.700296) | 2.461885 / 1.504120 (0.957765) | 2.252307 / 1.541195 (0.711112) | 2.307983 / 1.468490 (0.839493) | 0.697570 / 4.584777 (-3.887207) | 3.370323 / 3.745712 (-0.375389) | 3.131333 / 5.269862 (-2.138529) | 1.594839 / 4.565676 (-2.970838) | 0.082333 / 0.424275 (-0.341942) | 0.012574 / 0.007607 (0.004967) | 0.583704 / 0.226044 (0.357660) | 5.817675 / 2.268929 (3.548746) | 2.927054 / 55.444624 (-52.517570) | 2.582929 / 6.876477 (-4.293548) | 2.634275 / 2.142072 (0.492202) | 0.806407 / 4.805227 (-3.998821) | 0.151438 / 6.500664 (-6.349226) | 0.067429 / 0.075469 (-0.008040) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.267011 / 1.841788 (-0.574776) | 13.989515 / 8.074308 (5.915207) | 14.087968 / 10.191392 (3.896576) | 0.142130 / 0.680424 (-0.538293) | 0.017201 / 0.534201 (-0.517000) | 0.383394 / 0.579283 (-0.195889) | 0.381921 / 0.434364 (-0.052443) | 0.439169 / 0.540337 (-0.101168) | 0.524215 / 1.386936 (-0.862721) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#be2ebc8f3cfeb532c933be2443094603bafcab04 \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008489 / 0.011353 (-0.002864) | 0.004617 / 0.011008 (-0.006391) | 0.102035 / 0.038508 (0.063527) | 0.029850 / 0.023109 (0.006741) | 0.296789 / 0.275898 (0.020891) | 0.367270 / 0.323480 (0.043790) | 0.006934 / 0.007986 (-0.001052) | 0.004923 / 0.004328 (0.000595) | 0.079150 / 0.004250 (0.074900) | 0.036884 / 0.037052 (-0.000169) | 0.305747 / 0.258489 (0.047258) | 0.348510 / 0.293841 (0.054669) | 0.034074 / 0.128546 (-0.094472) | 0.011650 / 0.075646 (-0.063997) | 0.324226 / 0.419271 (-0.095045) | 0.041763 / 0.043533 (-0.001770) | 0.300887 / 0.255139 (0.045748) | 0.333393 / 0.283200 (0.050193) | 0.093838 / 0.141683 (-0.047844) | 1.499801 / 1.452155 (0.047646) | 1.505988 / 1.492716 (0.013272) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.198610 / 0.018006 (0.180604) | 0.407380 / 0.000490 (0.406891) | 0.000367 / 0.000200 (0.000167) | 0.000059 / 0.000054 (0.000005) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.022858 / 0.037411 (-0.014554) | 0.095727 / 0.014526 (0.081202) | 0.104014 / 0.176557 (-0.072543) | 0.138764 / 0.737135 (-0.598371) | 0.105860 / 0.296338 (-0.190478) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.416352 / 0.215209 (0.201143) | 4.150007 / 2.077655 (2.072352) | 1.878727 / 1.504120 (0.374607) | 1.678978 / 1.541195 (0.137783) | 1.711990 / 1.468490 (0.243500) | 0.691722 / 4.584777 (-3.893055) | 3.386466 / 3.745712 (-0.359246) | 1.835730 / 5.269862 (-3.434132) | 1.149975 / 4.565676 (-3.415702) | 0.081914 / 0.424275 (-0.342362) | 0.012238 / 0.007607 (0.004631) | 0.522945 / 0.226044 (0.296900) | 5.251793 / 2.268929 (2.982864) | 2.306907 / 55.444624 (-53.137717) | 1.968400 / 6.876477 (-4.908076) | 1.981154 / 2.142072 (-0.160919) | 0.810126 / 4.805227 (-3.995101) | 0.147876 / 6.500664 (-6.352788) | 0.064042 / 0.075469 (-0.011428) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.199150 / 1.841788 (-0.642637) | 13.913473 / 8.074308 (5.839165) | 14.079132 / 10.191392 (3.887740) | 0.137387 / 0.680424 (-0.543037) | 0.028456 / 0.534201 (-0.505745) | 0.394162 / 0.579283 (-0.185122) | 0.402051 / 0.434364 (-0.032313) | 0.461944 / 0.540337 (-0.078394) | 0.542648 / 1.386936 (-0.844288) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.006393 / 0.011353 (-0.004960) | 0.004599 / 0.011008 (-0.006409) | 0.097389 / 0.038508 (0.058881) | 0.027719 / 0.023109 (0.004610) | 0.341060 / 0.275898 (0.065162) | 0.379604 / 0.323480 (0.056124) | 0.004955 / 0.007986 (-0.003030) | 0.003369 / 0.004328 (-0.000959) | 0.075390 / 0.004250 (0.071139) | 0.038518 / 0.037052 (0.001466) | 0.347085 / 0.258489 (0.088596) | 0.393468 / 0.293841 (0.099627) | 0.031482 / 0.128546 (-0.097064) | 0.011585 / 0.075646 (-0.064061) | 0.317969 / 0.419271 (-0.101302) | 0.041389 / 0.043533 (-0.002144) | 0.343812 / 0.255139 (0.088673) | 0.371047 / 0.283200 (0.087848) | 0.090020 / 0.141683 (-0.051663) | 1.461690 / 1.452155 (0.009536) | 1.552458 / 1.492716 (0.059741) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.188691 / 0.018006 (0.170684) | 0.415635 / 0.000490 (0.415145) | 0.005285 / 0.000200 (0.005085) | 0.000087 / 0.000054 (0.000033) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.024695 / 0.037411 (-0.012716) | 0.098939 / 0.014526 (0.084413) | 0.108472 / 0.176557 (-0.068085) | 0.152635 / 0.737135 (-0.584501) | 0.109947 / 0.296338 (-0.186391) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.471975 / 0.215209 (0.256766) | 4.716437 / 2.077655 (2.638782) | 2.420148 / 1.504120 (0.916028) | 2.219864 / 1.541195 (0.678669) | 2.238647 / 1.468490 (0.770157) | 0.697628 / 4.584777 (-3.887149) | 3.530720 / 3.745712 (-0.214993) | 3.327354 / 5.269862 (-1.942508) | 1.665877 / 4.565676 (-2.899800) | 0.082650 / 0.424275 (-0.341625) | 0.012593 / 0.007607 (0.004986) | 0.576109 / 0.226044 (0.350065) | 5.744691 / 2.268929 (3.475762) | 2.863473 / 55.444624 (-52.581152) | 2.529616 / 6.876477 (-4.346861) | 2.562802 / 2.142072 (0.420730) | 0.805631 / 4.805227 (-3.999597) | 0.150788 / 6.500664 (-6.349876) | 0.065743 / 0.075469 (-0.009726) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.295134 / 1.841788 (-0.546654) | 14.096046 / 8.074308 (6.021738) | 13.901399 / 10.191392 (3.710007) | 0.127481 / 0.680424 (-0.552943) | 0.016666 / 0.534201 (-0.517535) | 0.381819 / 0.579283 (-0.197464) | 0.382629 / 0.434364 (-0.051735) | 0.439354 / 0.540337 (-0.100984) | 0.527662 / 1.386936 (-0.859274) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#0fe2ad43f59e65d39f2f2ce7442c76990493deb7 \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008509 / 0.011353 (-0.002844) | 0.004523 / 0.011008 (-0.006485) | 0.100616 / 0.038508 (0.062108) | 0.029573 / 0.023109 (0.006464) | 0.306414 / 0.275898 (0.030516) | 0.377034 / 0.323480 (0.053554) | 0.007621 / 0.007986 (-0.000365) | 0.003335 / 0.004328 (-0.000993) | 0.078598 / 0.004250 (0.074348) | 0.036902 / 0.037052 (-0.000150) | 0.318146 / 0.258489 (0.059657) | 0.355626 / 0.293841 (0.061785) | 0.033441 / 0.128546 (-0.095105) | 0.011552 / 0.075646 (-0.064094) | 0.322973 / 0.419271 (-0.096299) | 0.040564 / 0.043533 (-0.002968) | 0.306451 / 0.255139 (0.051312) | 0.337591 / 0.283200 (0.054392) | 0.086822 / 0.141683 (-0.054861) | 1.484601 / 1.452155 (0.032447) | 1.542777 / 1.492716 (0.050061) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.201711 / 0.018006 (0.183705) | 0.418387 / 0.000490 (0.417898) | 0.002753 / 0.000200 (0.002553) | 0.000263 / 0.000054 (0.000209) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.023016 / 0.037411 (-0.014395) | 0.097313 / 0.014526 (0.082787) | 0.103435 / 0.176557 (-0.073122) | 0.142665 / 0.737135 (-0.594470) | 0.107397 / 0.296338 (-0.188942) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.422739 / 0.215209 (0.207530) | 4.220126 / 2.077655 (2.142471) | 1.865447 / 1.504120 (0.361327) | 1.649647 / 1.541195 (0.108453) | 1.711655 / 1.468490 (0.243165) | 0.704269 / 4.584777 (-3.880508) | 3.407390 / 3.745712 (-0.338322) | 1.929224 / 5.269862 (-3.340638) | 1.281225 / 4.565676 (-3.284452) | 0.082924 / 0.424275 (-0.341351) | 0.012588 / 0.007607 (0.004981) | 0.531025 / 0.226044 (0.304980) | 5.339441 / 2.268929 (3.070512) | 2.298969 / 55.444624 (-53.145656) | 1.952145 / 6.876477 (-4.924332) | 2.034754 / 2.142072 (-0.107318) | 0.823672 / 4.805227 (-3.981555) | 0.151465 / 6.500664 (-6.349199) | 0.066663 / 0.075469 (-0.008807) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.258981 / 1.841788 (-0.582807) | 13.791640 / 8.074308 (5.717332) | 14.001514 / 10.191392 (3.810122) | 0.149805 / 0.680424 (-0.530619) | 0.028614 / 0.534201 (-0.505587) | 0.400266 / 0.579283 (-0.179017) | 0.405891 / 0.434364 (-0.028473) | 0.471903 / 0.540337 (-0.068435) | 0.563656 / 1.386936 (-0.823280) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.006751 / 0.011353 (-0.004601) | 0.004665 / 0.011008 (-0.006343) | 0.098362 / 0.038508 (0.059854) | 0.027451 / 0.023109 (0.004342) | 0.421859 / 0.275898 (0.145961) | 0.458089 / 0.323480 (0.134609) | 0.004885 / 0.007986 (-0.003101) | 0.003459 / 0.004328 (-0.000870) | 0.075871 / 0.004250 (0.071621) | 0.036591 / 0.037052 (-0.000462) | 0.423307 / 0.258489 (0.164818) | 0.467040 / 0.293841 (0.173199) | 0.031837 / 0.128546 (-0.096710) | 0.011604 / 0.075646 (-0.064042) | 0.321132 / 0.419271 (-0.098140) | 0.041806 / 0.043533 (-0.001727) | 0.421653 / 0.255139 (0.166514) | 0.445896 / 0.283200 (0.162696) | 0.087998 / 0.141683 (-0.053685) | 1.475818 / 1.452155 (0.023664) | 1.559487 / 1.492716 (0.066770) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.203096 / 0.018006 (0.185090) | 0.401381 / 0.000490 (0.400892) | 0.004037 / 0.000200 (0.003837) | 0.000080 / 0.000054 (0.000026) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.023757 / 0.037411 (-0.013654) | 0.099919 / 0.014526 (0.085393) | 0.108384 / 0.176557 (-0.068173) | 0.143780 / 0.737135 (-0.593355) | 0.111528 / 0.296338 (-0.184811) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.475896 / 0.215209 (0.260686) | 4.754567 / 2.077655 (2.676912) | 2.444986 / 1.504120 (0.940866) | 2.231055 / 1.541195 (0.689860) | 2.283646 / 1.468490 (0.815156) | 0.701303 / 4.584777 (-3.883474) | 3.381597 / 3.745712 (-0.364115) | 1.878714 / 5.269862 (-3.391148) | 1.171566 / 4.565676 (-3.394111) | 0.083106 / 0.424275 (-0.341169) | 0.012575 / 0.007607 (0.004967) | 0.582570 / 0.226044 (0.356526) | 5.813677 / 2.268929 (3.544748) | 2.908578 / 55.444624 (-52.536046) | 2.548459 / 6.876477 (-4.328017) | 2.581211 / 2.142072 (0.439139) | 0.807925 / 4.805227 (-3.997302) | 0.153516 / 6.500664 (-6.347148) | 0.068763 / 0.075469 (-0.006706) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.249595 / 1.841788 (-0.592193) | 14.208573 / 8.074308 (6.134265) | 14.179174 / 10.191392 (3.987781) | 0.156005 / 0.680424 (-0.524419) | 0.017045 / 0.534201 (-0.517156) | 0.377414 / 0.579283 (-0.201869) | 0.395291 / 0.434364 (-0.039073) | 0.444642 / 0.540337 (-0.095695) | 0.531626 / 1.386936 (-0.855311) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#52888645daa6854928474df6308bd997c8878ced \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008871 / 0.011353 (-0.002482) | 0.004616 / 0.011008 (-0.006392) | 0.100910 / 0.038508 (0.062402) | 0.030381 / 0.023109 (0.007272) | 0.304636 / 0.275898 (0.028737) | 0.384258 / 0.323480 (0.060778) | 0.007019 / 0.007986 (-0.000966) | 0.004262 / 0.004328 (-0.000066) | 0.077082 / 0.004250 (0.072832) | 0.035235 / 0.037052 (-0.001817) | 0.318293 / 0.258489 (0.059804) | 0.356578 / 0.293841 (0.062737) | 0.033568 / 0.128546 (-0.094978) | 0.011583 / 0.075646 (-0.064063) | 0.322442 / 0.419271 (-0.096830) | 0.041941 / 0.043533 (-0.001592) | 0.310469 / 0.255139 (0.055330) | 0.335626 / 0.283200 (0.052427) | 0.088195 / 0.141683 (-0.053487) | 1.466778 / 1.452155 (0.014623) | 1.512459 / 1.492716 (0.019743) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.184126 / 0.018006 (0.166120) | 0.413392 / 0.000490 (0.412902) | 0.002191 / 0.000200 (0.001992) | 0.000072 / 0.000054 (0.000018) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.023426 / 0.037411 (-0.013985) | 0.096240 / 0.014526 (0.081715) | 0.105908 / 0.176557 (-0.070648) | 0.146331 / 0.737135 (-0.590804) | 0.107441 / 0.296338 (-0.188898) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.420018 / 0.215209 (0.204809) | 4.198129 / 2.077655 (2.120474) | 1.998726 / 1.504120 (0.494606) | 1.870410 / 1.541195 (0.329215) | 1.925160 / 1.468490 (0.456670) | 0.688790 / 4.584777 (-3.895987) | 3.430629 / 3.745712 (-0.315083) | 2.875616 / 5.269862 (-2.394246) | 1.566269 / 4.565676 (-2.999408) | 0.082431 / 0.424275 (-0.341844) | 0.012409 / 0.007607 (0.004802) | 0.536178 / 0.226044 (0.310134) | 5.342918 / 2.268929 (3.073989) | 2.410814 / 55.444624 (-53.033811) | 2.056518 / 6.876477 (-4.819958) | 2.240148 / 2.142072 (0.098075) | 0.804848 / 4.805227 (-4.000379) | 0.147325 / 6.500664 (-6.353340) | 0.064217 / 0.075469 (-0.011252) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.285725 / 1.841788 (-0.556063) | 13.909739 / 8.074308 (5.835431) | 14.025774 / 10.191392 (3.834382) | 0.142413 / 0.680424 (-0.538011) | 0.028390 / 0.534201 (-0.505811) | 0.402345 / 0.579283 (-0.176939) | 0.404341 / 0.434364 (-0.030023) | 0.463055 / 0.540337 (-0.077282) | 0.556811 / 1.386936 (-0.830125) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.006557 / 0.011353 (-0.004795) | 0.004668 / 0.011008 (-0.006340) | 0.098839 / 0.038508 (0.060331) | 0.027618 / 0.023109 (0.004508) | 0.409338 / 0.275898 (0.133440) | 0.444048 / 0.323480 (0.120568) | 0.004881 / 0.007986 (-0.003105) | 0.003434 / 0.004328 (-0.000895) | 0.076497 / 0.004250 (0.072247) | 0.038932 / 0.037052 (0.001880) | 0.411419 / 0.258489 (0.152930) | 0.451167 / 0.293841 (0.157326) | 0.031649 / 0.128546 (-0.096897) | 0.011691 / 0.075646 (-0.063955) | 0.321586 / 0.419271 (-0.097685) | 0.041984 / 0.043533 (-0.001549) | 0.407717 / 0.255139 (0.152578) | 0.434687 / 0.283200 (0.151487) | 0.086419 / 0.141683 (-0.055264) | 1.491755 / 1.452155 (0.039601) | 1.569081 / 1.492716 (0.076364) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.231746 / 0.018006 (0.213739) | 0.412271 / 0.000490 (0.411781) | 0.000403 / 0.000200 (0.000203) | 0.000063 / 0.000054 (0.000008) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.024264 / 0.037411 (-0.013147) | 0.100478 / 0.014526 (0.085952) | 0.107065 / 0.176557 (-0.069491) | 0.140724 / 0.737135 (-0.596412) | 0.110631 / 0.296338 (-0.185707) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.472476 / 0.215209 (0.257267) | 4.738919 / 2.077655 (2.661265) | 2.438049 / 1.504120 (0.933929) | 2.237855 / 1.541195 (0.696660) | 2.282885 / 1.468490 (0.814395) | 0.690420 / 4.584777 (-3.894357) | 3.426487 / 3.745712 (-0.319225) | 1.842443 / 5.269862 (-3.427418) | 1.154466 / 4.565676 (-3.411210) | 0.082166 / 0.424275 (-0.342109) | 0.012309 / 0.007607 (0.004701) | 0.574730 / 0.226044 (0.348686) | 5.737566 / 2.268929 (3.468638) | 2.882405 / 55.444624 (-52.562220) | 2.540276 / 6.876477 (-4.336201) | 2.552356 / 2.142072 (0.410283) | 0.796413 / 4.805227 (-4.008815) | 0.152705 / 6.500664 (-6.347959) | 0.068273 / 0.075469 (-0.007196) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.244423 / 1.841788 (-0.597365) | 13.827750 / 8.074308 (5.753442) | 14.074083 / 10.191392 (3.882691) | 0.140291 / 0.680424 (-0.540133) | 0.017337 / 0.534201 (-0.516864) | 0.389314 / 0.579283 (-0.189969) | 0.390914 / 0.434364 (-0.043450) | 0.450333 / 0.540337 (-0.090004) | 0.543860 / 1.386936 (-0.843076) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#2cdcddc51d3cda24c2d79ad137af9e55d0a38044 \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.009490 / 0.011353 (-0.001863) | 0.005211 / 0.011008 (-0.005798) | 0.100884 / 0.038508 (0.062376) | 0.035834 / 0.023109 (0.012725) | 0.293623 / 0.275898 (0.017724) | 0.378118 / 0.323480 (0.054638) | 0.008106 / 0.007986 (0.000120) | 0.005339 / 0.004328 (0.001010) | 0.076311 / 0.004250 (0.072061) | 0.045954 / 0.037052 (0.008902) | 0.308163 / 0.258489 (0.049674) | 0.353470 / 0.293841 (0.059629) | 0.038539 / 0.128546 (-0.090008) | 0.012174 / 0.075646 (-0.063472) | 0.334875 / 0.419271 (-0.084396) | 0.048602 / 0.043533 (0.005069) | 0.295803 / 0.255139 (0.040664) | 0.318894 / 0.283200 (0.035695) | 0.105487 / 0.141683 (-0.036195) | 1.433628 / 1.452155 (-0.018526) | 1.466843 / 1.492716 (-0.025873) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.203426 / 0.018006 (0.185419) | 0.456877 / 0.000490 (0.456387) | 0.001452 / 0.000200 (0.001252) | 0.000088 / 0.000054 (0.000033) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.028308 / 0.037411 (-0.009103) | 0.108965 / 0.014526 (0.094439) | 0.119552 / 0.176557 (-0.057005) | 0.156371 / 0.737135 (-0.580765) | 0.124141 / 0.296338 (-0.172197) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.400183 / 0.215209 (0.184973) | 3.990983 / 2.077655 (1.913329) | 1.806729 / 1.504120 (0.302609) | 1.611944 / 1.541195 (0.070750) | 1.740019 / 1.468490 (0.271529) | 0.699600 / 4.584777 (-3.885177) | 3.868711 / 3.745712 (0.122999) | 3.249758 / 5.269862 (-2.020103) | 1.832213 / 4.565676 (-2.733463) | 0.085282 / 0.424275 (-0.338993) | 0.012726 / 0.007607 (0.005119) | 0.509385 / 0.226044 (0.283341) | 5.066913 / 2.268929 (2.797984) | 2.325710 / 55.444624 (-53.118914) | 1.962238 / 6.876477 (-4.914239) | 2.017576 / 2.142072 (-0.124496) | 0.839444 / 4.805227 (-3.965783) | 0.166936 / 6.500664 (-6.333728) | 0.064546 / 0.075469 (-0.010923) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.196396 / 1.841788 (-0.645392) | 15.077063 / 8.074308 (7.002755) | 14.268103 / 10.191392 (4.076711) | 0.163782 / 0.680424 (-0.516642) | 0.028794 / 0.534201 (-0.505407) | 0.440564 / 0.579283 (-0.138719) | 0.439826 / 0.434364 (0.005463) | 0.514786 / 0.540337 (-0.025551) | 0.603353 / 1.386936 (-0.783583) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.007874 / 0.011353 (-0.003479) | 0.005347 / 0.011008 (-0.005661) | 0.099461 / 0.038508 (0.060953) | 0.034010 / 0.023109 (0.010901) | 0.384650 / 0.275898 (0.108752) | 0.423827 / 0.323480 (0.100347) | 0.006201 / 0.007986 (-0.001784) | 0.004212 / 0.004328 (-0.000117) | 0.074354 / 0.004250 (0.070104) | 0.051675 / 0.037052 (0.014623) | 0.392488 / 0.258489 (0.133999) | 0.425828 / 0.293841 (0.131987) | 0.037444 / 0.128546 (-0.091103) | 0.012388 / 0.075646 (-0.063258) | 0.334482 / 0.419271 (-0.084789) | 0.050715 / 0.043533 (0.007182) | 0.378323 / 0.255139 (0.123184) | 0.395450 / 0.283200 (0.112250) | 0.108403 / 0.141683 (-0.033280) | 1.426803 / 1.452155 (-0.025352) | 1.532417 / 1.492716 (0.039701) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.219989 / 0.018006 (0.201982) | 0.454101 / 0.000490 (0.453611) | 0.000407 / 0.000200 (0.000207) | 0.000056 / 0.000054 (0.000002) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.030590 / 0.037411 (-0.006822) | 0.113483 / 0.014526 (0.098957) | 0.122603 / 0.176557 (-0.053954) | 0.161031 / 0.737135 (-0.576104) | 0.128039 / 0.296338 (-0.168300) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.430458 / 0.215209 (0.215249) | 4.286594 / 2.077655 (2.208940) | 2.056666 / 1.504120 (0.552546) | 1.861142 / 1.541195 (0.319948) | 1.937185 / 1.468490 (0.468695) | 0.701881 / 4.584777 (-3.882896) | 3.970144 / 3.745712 (0.224432) | 2.107118 / 5.269862 (-3.162744) | 1.351561 / 4.565676 (-3.214115) | 0.085470 / 0.424275 (-0.338805) | 0.012366 / 0.007607 (0.004759) | 0.525212 / 0.226044 (0.299168) | 5.301553 / 2.268929 (3.032625) | 2.593862 / 55.444624 (-52.850763) | 2.287315 / 6.876477 (-4.589161) | 2.368249 / 2.142072 (0.226176) | 0.855656 / 4.805227 (-3.949571) | 0.167846 / 6.500664 (-6.332818) | 0.064521 / 0.075469 (-0.010948) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.237008 / 1.841788 (-0.604779) | 15.784303 / 8.074308 (7.709995) | 14.613081 / 10.191392 (4.421689) | 0.161012 / 0.680424 (-0.519412) | 0.017928 / 0.534201 (-0.516273) | 0.423905 / 0.579283 (-0.155378) | 0.428316 / 0.434364 (-0.006048) | 0.500226 / 0.540337 (-0.040112) | 0.606725 / 1.386936 (-0.780211) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#08473e2ee66acb7e6f82d3591bb9b03924a661ed \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008874 / 0.011353 (-0.002479) | 0.004581 / 0.011008 (-0.006428) | 0.100180 / 0.038508 (0.061672) | 0.029990 / 0.023109 (0.006880) | 0.301616 / 0.275898 (0.025718) | 0.343662 / 0.323480 (0.020183) | 0.007111 / 0.007986 (-0.000875) | 0.003428 / 0.004328 (-0.000900) | 0.078031 / 0.004250 (0.073780) | 0.037332 / 0.037052 (0.000279) | 0.301977 / 0.258489 (0.043488) | 0.345581 / 0.293841 (0.051740) | 0.034305 / 0.128546 (-0.094241) | 0.011660 / 0.075646 (-0.063986) | 0.322289 / 0.419271 (-0.096982) | 0.041488 / 0.043533 (-0.002045) | 0.301612 / 0.255139 (0.046473) | 0.328174 / 0.283200 (0.044974) | 0.085561 / 0.141683 (-0.056122) | 1.482114 / 1.452155 (0.029959) | 1.556194 / 1.492716 (0.063478) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.186989 / 0.018006 (0.168983) | 0.421499 / 0.000490 (0.421009) | 0.001193 / 0.000200 (0.000993) | 0.000070 / 0.000054 (0.000016) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.023551 / 0.037411 (-0.013861) | 0.099868 / 0.014526 (0.085343) | 0.105233 / 0.176557 (-0.071324) | 0.141628 / 0.737135 (-0.595507) | 0.109004 / 0.296338 (-0.187335) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.415189 / 0.215209 (0.199979) | 4.145716 / 2.077655 (2.068061) | 1.837917 / 1.504120 (0.333797) | 1.635043 / 1.541195 (0.093848) | 1.683299 / 1.468490 (0.214809) | 0.688538 / 4.584777 (-3.896239) | 3.412628 / 3.745712 (-0.333084) | 1.877456 / 5.269862 (-3.392405) | 1.154129 / 4.565676 (-3.411547) | 0.081850 / 0.424275 (-0.342425) | 0.012309 / 0.007607 (0.004702) | 0.522830 / 0.226044 (0.296785) | 5.238685 / 2.268929 (2.969756) | 2.277840 / 55.444624 (-53.166784) | 1.941787 / 6.876477 (-4.934690) | 1.999688 / 2.142072 (-0.142385) | 0.807590 / 4.805227 (-3.997637) | 0.148157 / 6.500664 (-6.352507) | 0.064898 / 0.075469 (-0.010571) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.253859 / 1.841788 (-0.587929) | 13.676097 / 8.074308 (5.601789) | 14.237837 / 10.191392 (4.046444) | 0.137178 / 0.680424 (-0.543246) | 0.028971 / 0.534201 (-0.505230) | 0.400380 / 0.579283 (-0.178903) | 0.409990 / 0.434364 (-0.024374) | 0.462552 / 0.540337 (-0.077786) | 0.552153 / 1.386936 (-0.834783) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.006831 / 0.011353 (-0.004522) | 0.004627 / 0.011008 (-0.006381) | 0.099883 / 0.038508 (0.061375) | 0.028072 / 0.023109 (0.004962) | 0.343556 / 0.275898 (0.067658) | 0.386792 / 0.323480 (0.063312) | 0.005080 / 0.007986 (-0.002906) | 0.003508 / 0.004328 (-0.000820) | 0.077803 / 0.004250 (0.073552) | 0.040038 / 0.037052 (0.002985) | 0.345089 / 0.258489 (0.086600) | 0.396078 / 0.293841 (0.102238) | 0.032241 / 0.128546 (-0.096305) | 0.011711 / 0.075646 (-0.063935) | 0.320531 / 0.419271 (-0.098740) | 0.043658 / 0.043533 (0.000125) | 0.344696 / 0.255139 (0.089557) | 0.389847 / 0.283200 (0.106648) | 0.092328 / 0.141683 (-0.049355) | 1.477290 / 1.452155 (0.025136) | 1.548698 / 1.492716 (0.055982) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.236073 / 0.018006 (0.218067) | 0.422113 / 0.000490 (0.421624) | 0.000431 / 0.000200 (0.000231) | 0.000060 / 0.000054 (0.000005) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.024738 / 0.037411 (-0.012673) | 0.100546 / 0.014526 (0.086020) | 0.107550 / 0.176557 (-0.069006) | 0.146056 / 0.737135 (-0.591079) | 0.112665 / 0.296338 (-0.183674) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.490259 / 0.215209 (0.275050) | 4.907994 / 2.077655 (2.830339) | 2.547175 / 1.504120 (1.043055) | 2.344419 / 1.541195 (0.803224) | 2.403985 / 1.468490 (0.935495) | 0.696011 / 4.584777 (-3.888766) | 3.442426 / 3.745712 (-0.303286) | 1.878702 / 5.269862 (-3.391159) | 1.158280 / 4.565676 (-3.407396) | 0.082300 / 0.424275 (-0.341975) | 0.012513 / 0.007607 (0.004906) | 0.602696 / 0.226044 (0.376651) | 6.014592 / 2.268929 (3.745663) | 3.014466 / 55.444624 (-52.430159) | 2.669376 / 6.876477 (-4.207101) | 2.724485 / 2.142072 (0.582412) | 0.799795 / 4.805227 (-4.005432) | 0.151220 / 6.500664 (-6.349444) | 0.067486 / 0.075469 (-0.007983) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.281265 / 1.841788 (-0.560523) | 14.362284 / 8.074308 (6.287976) | 14.313690 / 10.191392 (4.122298) | 0.142870 / 0.680424 (-0.537554) | 0.017206 / 0.534201 (-0.516995) | 0.380084 / 0.579283 (-0.199199) | 0.388161 / 0.434364 (-0.046203) | 0.442617 / 0.540337 (-0.097721) | 0.528487 / 1.386936 (-0.858449) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#452b7f8ae78967dc662f5436e751233d46c62e78 \"CML watermark\")\n", "@lhoestq @amyeroberts @gante I did a substantial rewrite and all tests are passing now (Windows seems to time out or something and I can't figure out why - not sure if that's related to this PR!). I also confirmed tests are passing locally with Py==3.10. \r\n\r\nAside from incorporating everyone's comments, I also made a context manager to create and handle shared memory - this ensures that shared memory is cleaned up even if execution is interrupted. Also, shared memory names include a UUID string now to avoid collisions. Finally, string arrays are now split up into fixed-width character arrays in the workers so that they can be passed through shared memory, and the parent process reconstructs them into string arrays.", "Update: `test_arrow_dataset.py` ran fine in this branch on my Windows machine (Py 3.10), so I have no idea what's up with those tests", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008852 / 0.011353 (-0.002500) | 0.004545 / 0.011008 (-0.006464) | 0.099814 / 0.038508 (0.061306) | 0.030314 / 0.023109 (0.007205) | 0.310426 / 0.275898 (0.034528) | 0.366893 / 0.323480 (0.043413) | 0.007183 / 0.007986 (-0.000802) | 0.003476 / 0.004328 (-0.000853) | 0.077566 / 0.004250 (0.073315) | 0.038269 / 0.037052 (0.001217) | 0.319133 / 0.258489 (0.060644) | 0.352399 / 0.293841 (0.058558) | 0.033847 / 0.128546 (-0.094700) | 0.011568 / 0.075646 (-0.064078) | 0.321355 / 0.419271 (-0.097917) | 0.040719 / 0.043533 (-0.002814) | 0.304812 / 0.255139 (0.049673) | 0.329512 / 0.283200 (0.046312) | 0.088045 / 0.141683 (-0.053638) | 1.514182 / 1.452155 (0.062027) | 1.529459 / 1.492716 (0.036742) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.216749 / 0.018006 (0.198743) | 0.409909 / 0.000490 (0.409419) | 0.002790 / 0.000200 (0.002590) | 0.000081 / 0.000054 (0.000027) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.023390 / 0.037411 (-0.014021) | 0.095955 / 0.014526 (0.081430) | 0.104749 / 0.176557 (-0.071807) | 0.143414 / 0.737135 (-0.593721) | 0.109011 / 0.296338 (-0.187328) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.420410 / 0.215209 (0.205201) | 4.185745 / 2.077655 (2.108090) | 1.910207 / 1.504120 (0.406087) | 1.679330 / 1.541195 (0.138135) | 1.727134 / 1.468490 (0.258644) | 0.692379 / 4.584777 (-3.892398) | 3.358731 / 3.745712 (-0.386982) | 2.914657 / 5.269862 (-2.355205) | 1.506083 / 4.565676 (-3.059594) | 0.081922 / 0.424275 (-0.342353) | 0.012691 / 0.007607 (0.005084) | 0.530942 / 0.226044 (0.304897) | 5.357642 / 2.268929 (3.088714) | 2.387347 / 55.444624 (-53.057277) | 2.030001 / 6.876477 (-4.846476) | 2.026405 / 2.142072 (-0.115667) | 0.809406 / 4.805227 (-3.995821) | 0.149003 / 6.500664 (-6.351661) | 0.066910 / 0.075469 (-0.008559) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.278160 / 1.841788 (-0.563627) | 13.632742 / 8.074308 (5.558434) | 13.995537 / 10.191392 (3.804145) | 0.136507 / 0.680424 (-0.543917) | 0.028817 / 0.534201 (-0.505384) | 0.394842 / 0.579283 (-0.184441) | 0.399526 / 0.434364 (-0.034838) | 0.459174 / 0.540337 (-0.081163) | 0.536877 / 1.386936 (-0.850059) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.006814 / 0.011353 (-0.004539) | 0.004456 / 0.011008 (-0.006552) | 0.098386 / 0.038508 (0.059878) | 0.028124 / 0.023109 (0.005015) | 0.409004 / 0.275898 (0.133106) | 0.446746 / 0.323480 (0.123266) | 0.005108 / 0.007986 (-0.002877) | 0.004807 / 0.004328 (0.000479) | 0.075751 / 0.004250 (0.071500) | 0.039297 / 0.037052 (0.002244) | 0.413198 / 0.258489 (0.154709) | 0.452124 / 0.293841 (0.158283) | 0.032534 / 0.128546 (-0.096012) | 0.011689 / 0.075646 (-0.063957) | 0.325465 / 0.419271 (-0.093806) | 0.041347 / 0.043533 (-0.002185) | 0.411489 / 0.255139 (0.156350) | 0.447120 / 0.283200 (0.163920) | 0.093058 / 0.141683 (-0.048625) | 1.489903 / 1.452155 (0.037748) | 1.580771 / 1.492716 (0.088055) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.192619 / 0.018006 (0.174613) | 0.399201 / 0.000490 (0.398711) | 0.002894 / 0.000200 (0.002694) | 0.000071 / 0.000054 (0.000017) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.025120 / 0.037411 (-0.012292) | 0.100126 / 0.014526 (0.085600) | 0.108669 / 0.176557 (-0.067887) | 0.148687 / 0.737135 (-0.588448) | 0.112286 / 0.296338 (-0.184052) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.438866 / 0.215209 (0.223657) | 4.382418 / 2.077655 (2.304764) | 2.106450 / 1.504120 (0.602330) | 1.885105 / 1.541195 (0.343910) | 1.922948 / 1.468490 (0.454458) | 0.693145 / 4.584777 (-3.891632) | 3.378206 / 3.745712 (-0.367506) | 1.867295 / 5.269862 (-3.402566) | 1.164999 / 4.565676 (-3.400678) | 0.081918 / 0.424275 (-0.342357) | 0.012225 / 0.007607 (0.004618) | 0.547114 / 0.226044 (0.321069) | 5.454208 / 2.268929 (3.185279) | 2.532112 / 55.444624 (-52.912512) | 2.192573 / 6.876477 (-4.683904) | 2.225364 / 2.142072 (0.083291) | 0.797165 / 4.805227 (-4.008062) | 0.151185 / 6.500664 (-6.349480) | 0.067512 / 0.075469 (-0.007957) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.303905 / 1.841788 (-0.537883) | 14.107678 / 8.074308 (6.033370) | 14.147630 / 10.191392 (3.956238) | 0.156597 / 0.680424 (-0.523827) | 0.017037 / 0.534201 (-0.517164) | 0.383202 / 0.579283 (-0.196081) | 0.385340 / 0.434364 (-0.049024) | 0.443338 / 0.540337 (-0.097000) | 0.542345 / 1.386936 (-0.844591) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#38228533a03767aab713a3806aac0e8503668c68 \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.009982 / 0.011353 (-0.001371) | 0.005327 / 0.011008 (-0.005681) | 0.099092 / 0.038508 (0.060584) | 0.035824 / 0.023109 (0.012715) | 0.303258 / 0.275898 (0.027360) | 0.335379 / 0.323480 (0.011899) | 0.008192 / 0.007986 (0.000207) | 0.004242 / 0.004328 (-0.000087) | 0.076277 / 0.004250 (0.072026) | 0.043851 / 0.037052 (0.006799) | 0.307750 / 0.258489 (0.049261) | 0.348459 / 0.293841 (0.054618) | 0.038943 / 0.128546 (-0.089604) | 0.012128 / 0.075646 (-0.063519) | 0.334143 / 0.419271 (-0.085128) | 0.047865 / 0.043533 (0.004332) | 0.300909 / 0.255139 (0.045770) | 0.320879 / 0.283200 (0.037680) | 0.103812 / 0.141683 (-0.037871) | 1.468646 / 1.452155 (0.016491) | 1.557660 / 1.492716 (0.064944) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.244108 / 0.018006 (0.226102) | 0.554895 / 0.000490 (0.554405) | 0.005311 / 0.000200 (0.005111) | 0.000120 / 0.000054 (0.000065) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.028771 / 0.037411 (-0.008640) | 0.108133 / 0.014526 (0.093608) | 0.120098 / 0.176557 (-0.056458) | 0.159815 / 0.737135 (-0.577320) | 0.125437 / 0.296338 (-0.170901) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.397675 / 0.215209 (0.182466) | 3.975839 / 2.077655 (1.898184) | 1.797803 / 1.504120 (0.293683) | 1.612517 / 1.541195 (0.071322) | 1.659086 / 1.468490 (0.190596) | 0.679822 / 4.584777 (-3.904955) | 3.688321 / 3.745712 (-0.057391) | 2.155285 / 5.269862 (-3.114576) | 1.466453 / 4.565676 (-3.099223) | 0.084102 / 0.424275 (-0.340173) | 0.012074 / 0.007607 (0.004467) | 0.503744 / 0.226044 (0.277699) | 5.075599 / 2.268929 (2.806670) | 2.312149 / 55.444624 (-53.132476) | 1.975028 / 6.876477 (-4.901449) | 2.069554 / 2.142072 (-0.072519) | 0.828329 / 4.805227 (-3.976898) | 0.162816 / 6.500664 (-6.337849) | 0.063813 / 0.075469 (-0.011656) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.173327 / 1.841788 (-0.668461) | 15.281584 / 8.074308 (7.207276) | 14.450851 / 10.191392 (4.259459) | 0.165621 / 0.680424 (-0.514802) | 0.028779 / 0.534201 (-0.505422) | 0.438483 / 0.579283 (-0.140800) | 0.438477 / 0.434364 (0.004113) | 0.517703 / 0.540337 (-0.022634) | 0.615119 / 1.386936 (-0.771817) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.007013 / 0.011353 (-0.004340) | 0.005272 / 0.011008 (-0.005736) | 0.097203 / 0.038508 (0.058695) | 0.033103 / 0.023109 (0.009994) | 0.380203 / 0.275898 (0.104305) | 0.414868 / 0.323480 (0.091388) | 0.006326 / 0.007986 (-0.001659) | 0.005433 / 0.004328 (0.001104) | 0.074299 / 0.004250 (0.070049) | 0.049418 / 0.037052 (0.012366) | 0.388771 / 0.258489 (0.130282) | 0.435169 / 0.293841 (0.141328) | 0.036170 / 0.128546 (-0.092377) | 0.012452 / 0.075646 (-0.063195) | 0.331215 / 0.419271 (-0.088056) | 0.048577 / 0.043533 (0.005044) | 0.381491 / 0.255139 (0.126352) | 0.396731 / 0.283200 (0.113531) | 0.106435 / 0.141683 (-0.035248) | 1.446437 / 1.452155 (-0.005718) | 1.542337 / 1.492716 (0.049621) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.216714 / 0.018006 (0.198707) | 0.562460 / 0.000490 (0.561970) | 0.003636 / 0.000200 (0.003436) | 0.000100 / 0.000054 (0.000045) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.028726 / 0.037411 (-0.008686) | 0.111993 / 0.014526 (0.097467) | 0.125325 / 0.176557 (-0.051232) | 0.157779 / 0.737135 (-0.579356) | 0.130633 / 0.296338 (-0.165705) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.440520 / 0.215209 (0.225311) | 4.396283 / 2.077655 (2.318628) | 2.204714 / 1.504120 (0.700594) | 2.011667 / 1.541195 (0.470473) | 2.050518 / 1.468490 (0.582028) | 0.695204 / 4.584777 (-3.889573) | 3.779699 / 3.745712 (0.033987) | 2.096064 / 5.269862 (-3.173798) | 1.325446 / 4.565676 (-3.240230) | 0.085315 / 0.424275 (-0.338960) | 0.012178 / 0.007607 (0.004570) | 0.550478 / 0.226044 (0.324434) | 5.471872 / 2.268929 (3.202943) | 2.687147 / 55.444624 (-52.757478) | 2.348465 / 6.876477 (-4.528011) | 2.409700 / 2.142072 (0.267628) | 0.839468 / 4.805227 (-3.965760) | 0.167030 / 6.500664 (-6.333635) | 0.063243 / 0.075469 (-0.012226) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.257347 / 1.841788 (-0.584441) | 15.157821 / 8.074308 (7.083512) | 14.646381 / 10.191392 (4.454989) | 0.185550 / 0.680424 (-0.494874) | 0.018441 / 0.534201 (-0.515760) | 0.423330 / 0.579283 (-0.155954) | 0.426204 / 0.434364 (-0.008160) | 0.498985 / 0.540337 (-0.041352) | 0.608432 / 1.386936 (-0.778504) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#0f96e349ec5665e1e4135b5a108ba5db227bd3b1 \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.010856 / 0.011353 (-0.000497) | 0.005897 / 0.011008 (-0.005111) | 0.117826 / 0.038508 (0.079317) | 0.041899 / 0.023109 (0.018790) | 0.353804 / 0.275898 (0.077906) | 0.431021 / 0.323480 (0.107541) | 0.009288 / 0.007986 (0.001303) | 0.004556 / 0.004328 (0.000227) | 0.089344 / 0.004250 (0.085094) | 0.052224 / 0.037052 (0.015172) | 0.373242 / 0.258489 (0.114753) | 0.420667 / 0.293841 (0.126826) | 0.044191 / 0.128546 (-0.084355) | 0.014083 / 0.075646 (-0.061564) | 0.400373 / 0.419271 (-0.018898) | 0.056119 / 0.043533 (0.012586) | 0.363302 / 0.255139 (0.108163) | 0.382073 / 0.283200 (0.098873) | 0.118646 / 0.141683 (-0.023037) | 1.696576 / 1.452155 (0.244422) | 1.756518 / 1.492716 (0.263802) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.216388 / 0.018006 (0.198382) | 0.485732 / 0.000490 (0.485242) | 0.004012 / 0.000200 (0.003812) | 0.000104 / 0.000054 (0.000050) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.032095 / 0.037411 (-0.005316) | 0.128954 / 0.014526 (0.114429) | 0.137564 / 0.176557 (-0.038993) | 0.184315 / 0.737135 (-0.552820) | 0.144707 / 0.296338 (-0.151631) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.472792 / 0.215209 (0.257583) | 4.723044 / 2.077655 (2.645390) | 2.115075 / 1.504120 (0.610955) | 1.898993 / 1.541195 (0.357798) | 1.972894 / 1.468490 (0.504404) | 0.807210 / 4.584777 (-3.777567) | 4.493139 / 3.745712 (0.747427) | 2.501053 / 5.269862 (-2.768808) | 1.686121 / 4.565676 (-2.879556) | 0.099545 / 0.424275 (-0.324730) | 0.014360 / 0.007607 (0.006753) | 0.596235 / 0.226044 (0.370191) | 5.944285 / 2.268929 (3.675357) | 2.654944 / 55.444624 (-52.789681) | 2.281451 / 6.876477 (-4.595026) | 2.448407 / 2.142072 (0.306334) | 1.000512 / 4.805227 (-3.804716) | 0.196413 / 6.500664 (-6.304251) | 0.075810 / 0.075469 (0.000341) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.435707 / 1.841788 (-0.406081) | 17.931070 / 8.074308 (9.856762) | 16.635522 / 10.191392 (6.444130) | 0.189119 / 0.680424 (-0.491304) | 0.034392 / 0.534201 (-0.499809) | 0.519041 / 0.579283 (-0.060242) | 0.516159 / 0.434364 (0.081795) | 0.601180 / 0.540337 (0.060843) | 0.713180 / 1.386936 (-0.673756) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008741 / 0.011353 (-0.002612) | 0.006102 / 0.011008 (-0.004906) | 0.114787 / 0.038508 (0.076279) | 0.039610 / 0.023109 (0.016501) | 0.451730 / 0.275898 (0.175832) | 0.488820 / 0.323480 (0.165340) | 0.006979 / 0.007986 (-0.001006) | 0.006458 / 0.004328 (0.002130) | 0.086505 / 0.004250 (0.082254) | 0.057684 / 0.037052 (0.020632) | 0.451354 / 0.258489 (0.192865) | 0.523143 / 0.293841 (0.229302) | 0.043224 / 0.128546 (-0.085323) | 0.014671 / 0.075646 (-0.060975) | 0.398030 / 0.419271 (-0.021241) | 0.063650 / 0.043533 (0.020117) | 0.448324 / 0.255139 (0.193185) | 0.476560 / 0.283200 (0.193361) | 0.125772 / 0.141683 (-0.015911) | 1.801051 / 1.452155 (0.348896) | 1.872736 / 1.492716 (0.380020) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.256146 / 0.018006 (0.238139) | 0.486915 / 0.000490 (0.486425) | 0.000513 / 0.000200 (0.000313) | 0.000067 / 0.000054 (0.000012) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.035242 / 0.037411 (-0.002170) | 0.134322 / 0.014526 (0.119797) | 0.144786 / 0.176557 (-0.031770) | 0.188786 / 0.737135 (-0.548349) | 0.151737 / 0.296338 (-0.144602) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.506047 / 0.215209 (0.290838) | 5.028253 / 2.077655 (2.950598) | 2.393070 / 1.504120 (0.888950) | 2.157847 / 1.541195 (0.616652) | 2.229412 / 1.468490 (0.760922) | 0.828973 / 4.584777 (-3.755804) | 4.741470 / 3.745712 (0.995758) | 4.048118 / 5.269862 (-1.221744) | 2.573818 / 4.565676 (-1.991859) | 0.101019 / 0.424275 (-0.323256) | 0.014640 / 0.007607 (0.007033) | 0.632591 / 0.226044 (0.406546) | 6.289153 / 2.268929 (4.020224) | 2.977261 / 55.444624 (-52.467363) | 2.554396 / 6.876477 (-4.322081) | 2.619446 / 2.142072 (0.477374) | 0.988376 / 4.805227 (-3.816851) | 0.196895 / 6.500664 (-6.303769) | 0.076355 / 0.075469 (0.000886) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.493570 / 1.841788 (-0.348218) | 18.422758 / 8.074308 (10.348449) | 17.007352 / 10.191392 (6.815960) | 0.191903 / 0.680424 (-0.488521) | 0.020974 / 0.534201 (-0.513227) | 0.500573 / 0.579283 (-0.078710) | 0.489381 / 0.434364 (0.055017) | 0.580765 / 0.540337 (0.040428) | 0.698907 / 1.386936 (-0.688029) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#fa9baa268a6d285ab0a61cc37413392c94cfe2e8 \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008979 / 0.011353 (-0.002374) | 0.004497 / 0.011008 (-0.006511) | 0.102227 / 0.038508 (0.063719) | 0.031302 / 0.023109 (0.008193) | 0.298488 / 0.275898 (0.022590) | 0.372589 / 0.323480 (0.049109) | 0.007261 / 0.007986 (-0.000725) | 0.003542 / 0.004328 (-0.000786) | 0.078503 / 0.004250 (0.074253) | 0.039474 / 0.037052 (0.002422) | 0.310991 / 0.258489 (0.052502) | 0.353245 / 0.293841 (0.059404) | 0.033798 / 0.128546 (-0.094749) | 0.011634 / 0.075646 (-0.064012) | 0.321141 / 0.419271 (-0.098131) | 0.041264 / 0.043533 (-0.002268) | 0.300900 / 0.255139 (0.045761) | 0.326255 / 0.283200 (0.043055) | 0.092477 / 0.141683 (-0.049205) | 1.478921 / 1.452155 (0.026766) | 1.514915 / 1.492716 (0.022198) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.184415 / 0.018006 (0.166408) | 0.428986 / 0.000490 (0.428497) | 0.002590 / 0.000200 (0.002390) | 0.000072 / 0.000054 (0.000018) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.023730 / 0.037411 (-0.013681) | 0.099846 / 0.014526 (0.085320) | 0.107075 / 0.176557 (-0.069482) | 0.147475 / 0.737135 (-0.589661) | 0.111802 / 0.296338 (-0.184537) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.413704 / 0.215209 (0.198494) | 4.144498 / 2.077655 (2.066843) | 1.855900 / 1.504120 (0.351780) | 1.647958 / 1.541195 (0.106763) | 1.712437 / 1.468490 (0.243947) | 0.688382 / 4.584777 (-3.896395) | 3.432136 / 3.745712 (-0.313576) | 2.837211 / 5.269862 (-2.432651) | 1.519004 / 4.565676 (-3.046672) | 0.082429 / 0.424275 (-0.341846) | 0.012610 / 0.007607 (0.005003) | 0.525078 / 0.226044 (0.299034) | 5.272932 / 2.268929 (3.004003) | 2.340482 / 55.444624 (-53.104143) | 2.007372 / 6.876477 (-4.869104) | 2.060567 / 2.142072 (-0.081506) | 0.806476 / 4.805227 (-3.998752) | 0.149421 / 6.500664 (-6.351243) | 0.066252 / 0.075469 (-0.009218) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.235078 / 1.841788 (-0.606710) | 13.870758 / 8.074308 (5.796450) | 14.104582 / 10.191392 (3.913190) | 0.159375 / 0.680424 (-0.521049) | 0.029233 / 0.534201 (-0.504968) | 0.392184 / 0.579283 (-0.187099) | 0.407909 / 0.434364 (-0.026455) | 0.458757 / 0.540337 (-0.081581) | 0.547681 / 1.386936 (-0.839255) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.007194 / 0.011353 (-0.004159) | 0.004578 / 0.011008 (-0.006431) | 0.098936 / 0.038508 (0.060428) | 0.029639 / 0.023109 (0.006530) | 0.347241 / 0.275898 (0.071343) | 0.378838 / 0.323480 (0.055358) | 0.005632 / 0.007986 (-0.002353) | 0.003469 / 0.004328 (-0.000860) | 0.075536 / 0.004250 (0.071285) | 0.043301 / 0.037052 (0.006249) | 0.348091 / 0.258489 (0.089602) | 0.388595 / 0.293841 (0.094754) | 0.033512 / 0.128546 (-0.095034) | 0.011754 / 0.075646 (-0.063892) | 0.321003 / 0.419271 (-0.098268) | 0.044634 / 0.043533 (0.001101) | 0.346688 / 0.255139 (0.091549) | 0.366346 / 0.283200 (0.083147) | 0.093650 / 0.141683 (-0.048033) | 1.509913 / 1.452155 (0.057759) | 1.596414 / 1.492716 (0.103698) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.230466 / 0.018006 (0.212459) | 0.417106 / 0.000490 (0.416617) | 0.000959 / 0.000200 (0.000759) | 0.000070 / 0.000054 (0.000015) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.025581 / 0.037411 (-0.011830) | 0.105246 / 0.014526 (0.090720) | 0.108997 / 0.176557 (-0.067560) | 0.144342 / 0.737135 (-0.592794) | 0.113911 / 0.296338 (-0.182427) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.479608 / 0.215209 (0.264399) | 4.766081 / 2.077655 (2.688426) | 2.446597 / 1.504120 (0.942477) | 2.228278 / 1.541195 (0.687083) | 2.289943 / 1.468490 (0.821453) | 0.703146 / 4.584777 (-3.881631) | 3.414150 / 3.745712 (-0.331562) | 2.957730 / 5.269862 (-2.312132) | 1.531524 / 4.565676 (-3.034152) | 0.083449 / 0.424275 (-0.340826) | 0.012684 / 0.007607 (0.005077) | 0.587622 / 0.226044 (0.361578) | 5.888791 / 2.268929 (3.619863) | 2.884200 / 55.444624 (-52.560424) | 2.543739 / 6.876477 (-4.332737) | 2.596245 / 2.142072 (0.454173) | 0.813070 / 4.805227 (-3.992157) | 0.152706 / 6.500664 (-6.347958) | 0.069257 / 0.075469 (-0.006212) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.302945 / 1.841788 (-0.538842) | 14.484051 / 8.074308 (6.409743) | 14.216143 / 10.191392 (4.024751) | 0.154537 / 0.680424 (-0.525886) | 0.016909 / 0.534201 (-0.517292) | 0.389433 / 0.579283 (-0.189850) | 0.393280 / 0.434364 (-0.041084) | 0.446884 / 0.540337 (-0.093453) | 0.534394 / 1.386936 (-0.852542) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#2bcdeb952c57c5f22643061d49d16014a7b6426a \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008822 / 0.011353 (-0.002530) | 0.004826 / 0.011008 (-0.006182) | 0.102710 / 0.038508 (0.064202) | 0.030353 / 0.023109 (0.007244) | 0.297224 / 0.275898 (0.021326) | 0.371861 / 0.323480 (0.048381) | 0.007266 / 0.007986 (-0.000720) | 0.003632 / 0.004328 (-0.000696) | 0.079960 / 0.004250 (0.075710) | 0.036908 / 0.037052 (-0.000144) | 0.309582 / 0.258489 (0.051093) | 0.350108 / 0.293841 (0.056267) | 0.034280 / 0.128546 (-0.094266) | 0.011739 / 0.075646 (-0.063907) | 0.323217 / 0.419271 (-0.096054) | 0.043491 / 0.043533 (-0.000042) | 0.298454 / 0.255139 (0.043315) | 0.326735 / 0.283200 (0.043535) | 0.093955 / 0.141683 (-0.047728) | 1.494313 / 1.452155 (0.042159) | 1.562104 / 1.492716 (0.069388) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.182796 / 0.018006 (0.164790) | 0.420133 / 0.000490 (0.419643) | 0.002537 / 0.000200 (0.002337) | 0.000070 / 0.000054 (0.000015) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.023143 / 0.037411 (-0.014269) | 0.098560 / 0.014526 (0.084034) | 0.105060 / 0.176557 (-0.071496) | 0.140269 / 0.737135 (-0.596866) | 0.109120 / 0.296338 (-0.187219) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.419907 / 0.215209 (0.204698) | 4.196179 / 2.077655 (2.118524) | 1.887663 / 1.504120 (0.383543) | 1.686232 / 1.541195 (0.145037) | 1.741741 / 1.468490 (0.273251) | 0.696222 / 4.584777 (-3.888555) | 3.400250 / 3.745712 (-0.345462) | 1.875058 / 5.269862 (-3.394803) | 1.159466 / 4.565676 (-3.406211) | 0.082520 / 0.424275 (-0.341755) | 0.012408 / 0.007607 (0.004801) | 0.525212 / 0.226044 (0.299168) | 5.283691 / 2.268929 (3.014762) | 2.314487 / 55.444624 (-53.130138) | 1.966212 / 6.876477 (-4.910265) | 2.023458 / 2.142072 (-0.118615) | 0.808896 / 4.805227 (-3.996331) | 0.148973 / 6.500664 (-6.351691) | 0.065378 / 0.075469 (-0.010091) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.223833 / 1.841788 (-0.617955) | 14.053651 / 8.074308 (5.979343) | 14.072165 / 10.191392 (3.880773) | 0.156006 / 0.680424 (-0.524418) | 0.028665 / 0.534201 (-0.505536) | 0.392099 / 0.579283 (-0.187184) | 0.401460 / 0.434364 (-0.032904) | 0.462184 / 0.540337 (-0.078153) | 0.540459 / 1.386936 (-0.846477) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.006907 / 0.011353 (-0.004446) | 0.004585 / 0.011008 (-0.006423) | 0.099027 / 0.038508 (0.060519) | 0.028317 / 0.023109 (0.005208) | 0.421068 / 0.275898 (0.145170) | 0.450712 / 0.323480 (0.127233) | 0.005229 / 0.007986 (-0.002756) | 0.004873 / 0.004328 (0.000545) | 0.077374 / 0.004250 (0.073124) | 0.042530 / 0.037052 (0.005477) | 0.417392 / 0.258489 (0.158903) | 0.462605 / 0.293841 (0.168764) | 0.032195 / 0.128546 (-0.096351) | 0.011777 / 0.075646 (-0.063870) | 0.321927 / 0.419271 (-0.097344) | 0.041999 / 0.043533 (-0.001533) | 0.419402 / 0.255139 (0.164263) | 0.437179 / 0.283200 (0.153979) | 0.089549 / 0.141683 (-0.052134) | 1.469525 / 1.452155 (0.017370) | 1.586407 / 1.492716 (0.093691) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.209533 / 0.018006 (0.191526) | 0.413886 / 0.000490 (0.413396) | 0.003357 / 0.000200 (0.003157) | 0.000121 / 0.000054 (0.000067) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.026133 / 0.037411 (-0.011278) | 0.103128 / 0.014526 (0.088602) | 0.110604 / 0.176557 (-0.065952) | 0.153055 / 0.737135 (-0.584080) | 0.112257 / 0.296338 (-0.184081) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.471281 / 0.215209 (0.256072) | 4.708361 / 2.077655 (2.630706) | 2.572681 / 1.504120 (1.068561) | 2.370536 / 1.541195 (0.829341) | 2.456010 / 1.468490 (0.987520) | 0.694173 / 4.584777 (-3.890603) | 3.434511 / 3.745712 (-0.311201) | 1.877169 / 5.269862 (-3.392693) | 1.158387 / 4.565676 (-3.407289) | 0.081849 / 0.424275 (-0.342426) | 0.012176 / 0.007607 (0.004569) | 0.581736 / 0.226044 (0.355692) | 5.803173 / 2.268929 (3.534245) | 3.040003 / 55.444624 (-52.404621) | 2.704698 / 6.876477 (-4.171779) | 2.760138 / 2.142072 (0.618065) | 0.802557 / 4.805227 (-4.002671) | 0.151397 / 6.500664 (-6.349268) | 0.068308 / 0.075469 (-0.007161) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.304062 / 1.841788 (-0.537725) | 14.364809 / 8.074308 (6.290501) | 14.192131 / 10.191392 (4.000739) | 0.150025 / 0.680424 (-0.530399) | 0.017020 / 0.534201 (-0.517181) | 0.389235 / 0.579283 (-0.190048) | 0.387557 / 0.434364 (-0.046807) | 0.454636 / 0.540337 (-0.085702) | 0.558182 / 1.386936 (-0.828754) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#663e5eddca188abbb37e2f803846f02fe4ca0d9b \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008519 / 0.011353 (-0.002834) | 0.004538 / 0.011008 (-0.006470) | 0.102066 / 0.038508 (0.063558) | 0.029700 / 0.023109 (0.006591) | 0.304573 / 0.275898 (0.028675) | 0.366232 / 0.323480 (0.042752) | 0.007154 / 0.007986 (-0.000832) | 0.003497 / 0.004328 (-0.000831) | 0.079119 / 0.004250 (0.074868) | 0.036088 / 0.037052 (-0.000964) | 0.311076 / 0.258489 (0.052587) | 0.352205 / 0.293841 (0.058364) | 0.033706 / 0.128546 (-0.094840) | 0.011657 / 0.075646 (-0.063990) | 0.324024 / 0.419271 (-0.095247) | 0.040777 / 0.043533 (-0.002756) | 0.302661 / 0.255139 (0.047522) | 0.329091 / 0.283200 (0.045891) | 0.086774 / 0.141683 (-0.054909) | 1.485874 / 1.452155 (0.033720) | 1.535726 / 1.492716 (0.043009) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.194284 / 0.018006 (0.176277) | 0.412875 / 0.000490 (0.412385) | 0.003348 / 0.000200 (0.003148) | 0.000074 / 0.000054 (0.000020) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.022432 / 0.037411 (-0.014979) | 0.095008 / 0.014526 (0.080482) | 0.103268 / 0.176557 (-0.073288) | 0.140121 / 0.737135 (-0.597014) | 0.106619 / 0.296338 (-0.189719) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.414786 / 0.215209 (0.199577) | 4.146345 / 2.077655 (2.068690) | 1.873703 / 1.504120 (0.369583) | 1.673498 / 1.541195 (0.132303) | 1.716993 / 1.468490 (0.248502) | 0.692098 / 4.584777 (-3.892679) | 3.380991 / 3.745712 (-0.364721) | 1.846811 / 5.269862 (-3.423050) | 1.159617 / 4.565676 (-3.406059) | 0.081867 / 0.424275 (-0.342408) | 0.012371 / 0.007607 (0.004764) | 0.526228 / 0.226044 (0.300184) | 5.273139 / 2.268929 (3.004211) | 2.327147 / 55.444624 (-53.117477) | 1.968366 / 6.876477 (-4.908111) | 2.018053 / 2.142072 (-0.124019) | 0.816098 / 4.805227 (-3.989130) | 0.149438 / 6.500664 (-6.351226) | 0.065000 / 0.075469 (-0.010469) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.244408 / 1.841788 (-0.597380) | 13.774354 / 8.074308 (5.700046) | 14.178923 / 10.191392 (3.987531) | 0.150032 / 0.680424 (-0.530392) | 0.029736 / 0.534201 (-0.504465) | 0.399134 / 0.579283 (-0.180149) | 0.404214 / 0.434364 (-0.030150) | 0.462096 / 0.540337 (-0.078242) | 0.542256 / 1.386936 (-0.844680) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.006776 / 0.011353 (-0.004577) | 0.004586 / 0.011008 (-0.006422) | 0.097658 / 0.038508 (0.059150) | 0.027627 / 0.023109 (0.004517) | 0.423794 / 0.275898 (0.147896) | 0.447443 / 0.323480 (0.123963) | 0.005099 / 0.007986 (-0.002886) | 0.004846 / 0.004328 (0.000517) | 0.075135 / 0.004250 (0.070884) | 0.038068 / 0.037052 (0.001016) | 0.420999 / 0.258489 (0.162510) | 0.460368 / 0.293841 (0.166527) | 0.032107 / 0.128546 (-0.096439) | 0.011775 / 0.075646 (-0.063871) | 0.323854 / 0.419271 (-0.095418) | 0.045538 / 0.043533 (0.002005) | 0.420949 / 0.255139 (0.165810) | 0.441906 / 0.283200 (0.158706) | 0.091955 / 0.141683 (-0.049728) | 1.523736 / 1.452155 (0.071581) | 1.587865 / 1.492716 (0.095148) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.263297 / 0.018006 (0.245290) | 0.416170 / 0.000490 (0.415680) | 0.023161 / 0.000200 (0.022961) | 0.000243 / 0.000054 (0.000188) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.024000 / 0.037411 (-0.013412) | 0.097787 / 0.014526 (0.083262) | 0.106884 / 0.176557 (-0.069672) | 0.140861 / 0.737135 (-0.596274) | 0.108228 / 0.296338 (-0.188111) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.477222 / 0.215209 (0.262013) | 4.774729 / 2.077655 (2.697074) | 2.451575 / 1.504120 (0.947455) | 2.251255 / 1.541195 (0.710060) | 2.281154 / 1.468490 (0.812664) | 0.699394 / 4.584777 (-3.885383) | 3.421575 / 3.745712 (-0.324137) | 2.704713 / 5.269862 (-2.565148) | 1.508464 / 4.565676 (-3.057212) | 0.082199 / 0.424275 (-0.342076) | 0.012586 / 0.007607 (0.004979) | 0.588783 / 0.226044 (0.362739) | 5.878434 / 2.268929 (3.609505) | 2.927422 / 55.444624 (-52.517202) | 2.574357 / 6.876477 (-4.302120) | 2.603626 / 2.142072 (0.461554) | 0.804706 / 4.805227 (-4.000521) | 0.152919 / 6.500664 (-6.347745) | 0.069316 / 0.075469 (-0.006153) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.280025 / 1.841788 (-0.561763) | 13.968407 / 8.074308 (5.894099) | 13.874506 / 10.191392 (3.683114) | 0.154711 / 0.680424 (-0.525713) | 0.016827 / 0.534201 (-0.517374) | 0.377775 / 0.579283 (-0.201508) | 0.393035 / 0.434364 (-0.041329) | 0.439405 / 0.540337 (-0.100932) | 0.528135 / 1.386936 (-0.858801) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#00b27a59b8af9075967b800e3b0f1de8616aa0ce \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.009035 / 0.011353 (-0.002318) | 0.004518 / 0.011008 (-0.006490) | 0.102077 / 0.038508 (0.063569) | 0.030169 / 0.023109 (0.007060) | 0.297713 / 0.275898 (0.021815) | 0.364976 / 0.323480 (0.041496) | 0.007079 / 0.007986 (-0.000906) | 0.003438 / 0.004328 (-0.000890) | 0.079667 / 0.004250 (0.075416) | 0.035890 / 0.037052 (-0.001162) | 0.306065 / 0.258489 (0.047576) | 0.352133 / 0.293841 (0.058292) | 0.033800 / 0.128546 (-0.094746) | 0.011613 / 0.075646 (-0.064034) | 0.322917 / 0.419271 (-0.096354) | 0.040973 / 0.043533 (-0.002560) | 0.300896 / 0.255139 (0.045757) | 0.331540 / 0.283200 (0.048341) | 0.089579 / 0.141683 (-0.052103) | 1.466755 / 1.452155 (0.014600) | 1.522120 / 1.492716 (0.029404) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.193172 / 0.018006 (0.175166) | 0.408878 / 0.000490 (0.408389) | 0.001586 / 0.000200 (0.001386) | 0.000071 / 0.000054 (0.000017) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.023496 / 0.037411 (-0.013915) | 0.098046 / 0.014526 (0.083520) | 0.104599 / 0.176557 (-0.071957) | 0.139054 / 0.737135 (-0.598081) | 0.111163 / 0.296338 (-0.185175) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.417374 / 0.215209 (0.202165) | 4.145808 / 2.077655 (2.068153) | 1.847101 / 1.504120 (0.342981) | 1.637207 / 1.541195 (0.096012) | 1.676906 / 1.468490 (0.208416) | 0.689851 / 4.584777 (-3.894926) | 3.402099 / 3.745712 (-0.343614) | 1.896808 / 5.269862 (-3.373054) | 1.257876 / 4.565676 (-3.307801) | 0.081744 / 0.424275 (-0.342531) | 0.012206 / 0.007607 (0.004599) | 0.524830 / 0.226044 (0.298786) | 5.251344 / 2.268929 (2.982416) | 2.277907 / 55.444624 (-53.166717) | 1.933985 / 6.876477 (-4.942491) | 2.038500 / 2.142072 (-0.103573) | 0.808696 / 4.805227 (-3.996532) | 0.149488 / 6.500664 (-6.351176) | 0.065323 / 0.075469 (-0.010146) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.204294 / 1.841788 (-0.637493) | 13.696526 / 8.074308 (5.622218) | 13.947195 / 10.191392 (3.755802) | 0.136812 / 0.680424 (-0.543611) | 0.028625 / 0.534201 (-0.505576) | 0.397662 / 0.579283 (-0.181621) | 0.403423 / 0.434364 (-0.030941) | 0.465288 / 0.540337 (-0.075049) | 0.551919 / 1.386936 (-0.835017) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.006467 / 0.011353 (-0.004886) | 0.004562 / 0.011008 (-0.006447) | 0.097514 / 0.038508 (0.059006) | 0.027471 / 0.023109 (0.004362) | 0.425504 / 0.275898 (0.149606) | 0.458856 / 0.323480 (0.135376) | 0.004816 / 0.007986 (-0.003169) | 0.003264 / 0.004328 (-0.001065) | 0.074947 / 0.004250 (0.070697) | 0.037147 / 0.037052 (0.000095) | 0.429513 / 0.258489 (0.171024) | 0.463971 / 0.293841 (0.170130) | 0.031638 / 0.128546 (-0.096908) | 0.011545 / 0.075646 (-0.064101) | 0.320261 / 0.419271 (-0.099010) | 0.041570 / 0.043533 (-0.001963) | 0.424809 / 0.255139 (0.169670) | 0.447158 / 0.283200 (0.163959) | 0.088418 / 0.141683 (-0.053265) | 1.492242 / 1.452155 (0.040087) | 1.545523 / 1.492716 (0.052807) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.217865 / 0.018006 (0.199859) | 0.399925 / 0.000490 (0.399436) | 0.004853 / 0.000200 (0.004653) | 0.000073 / 0.000054 (0.000019) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.024275 / 0.037411 (-0.013137) | 0.098249 / 0.014526 (0.083723) | 0.107110 / 0.176557 (-0.069446) | 0.143870 / 0.737135 (-0.593265) | 0.108796 / 0.296338 (-0.187542) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.470856 / 0.215209 (0.255647) | 4.687921 / 2.077655 (2.610266) | 2.448631 / 1.504120 (0.944511) | 2.247748 / 1.541195 (0.706553) | 2.287713 / 1.468490 (0.819223) | 0.687534 / 4.584777 (-3.897243) | 3.421099 / 3.745712 (-0.324613) | 2.977280 / 5.269862 (-2.292582) | 1.274837 / 4.565676 (-3.290839) | 0.081611 / 0.424275 (-0.342664) | 0.012603 / 0.007607 (0.004996) | 0.574600 / 0.226044 (0.348556) | 5.802826 / 2.268929 (3.533898) | 2.913178 / 55.444624 (-52.531446) | 2.589486 / 6.876477 (-4.286991) | 2.630004 / 2.142072 (0.487932) | 0.790087 / 4.805227 (-4.015140) | 0.150019 / 6.500664 (-6.350645) | 0.067346 / 0.075469 (-0.008123) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.266521 / 1.841788 (-0.575267) | 13.818770 / 8.074308 (5.744462) | 13.872277 / 10.191392 (3.680885) | 0.147375 / 0.680424 (-0.533049) | 0.016837 / 0.534201 (-0.517363) | 0.376421 / 0.579283 (-0.202862) | 0.400236 / 0.434364 (-0.034128) | 0.436623 / 0.540337 (-0.103714) | 0.527173 / 1.386936 (-0.859763) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#5f347cf8443aa35401ba6a4159600b92bc6a156b \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.009341 / 0.011353 (-0.002012) | 0.005188 / 0.011008 (-0.005820) | 0.101831 / 0.038508 (0.063323) | 0.035141 / 0.023109 (0.012032) | 0.299324 / 0.275898 (0.023426) | 0.334749 / 0.323480 (0.011269) | 0.007958 / 0.007986 (-0.000027) | 0.005482 / 0.004328 (0.001153) | 0.077070 / 0.004250 (0.072820) | 0.044733 / 0.037052 (0.007680) | 0.310398 / 0.258489 (0.051909) | 0.347925 / 0.293841 (0.054084) | 0.038141 / 0.128546 (-0.090405) | 0.012135 / 0.075646 (-0.063512) | 0.333799 / 0.419271 (-0.085472) | 0.048881 / 0.043533 (0.005348) | 0.301336 / 0.255139 (0.046197) | 0.314592 / 0.283200 (0.031393) | 0.103635 / 0.141683 (-0.038048) | 1.437321 / 1.452155 (-0.014833) | 1.598781 / 1.492716 (0.106065) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.248911 / 0.018006 (0.230905) | 0.528932 / 0.000490 (0.528442) | 0.002495 / 0.000200 (0.002295) | 0.000094 / 0.000054 (0.000040) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.027903 / 0.037411 (-0.009509) | 0.106716 / 0.014526 (0.092190) | 0.122650 / 0.176557 (-0.053907) | 0.162481 / 0.737135 (-0.574654) | 0.126402 / 0.296338 (-0.169937) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.352819 / 0.215209 (0.137610) | 3.522761 / 2.077655 (1.445106) | 1.576761 / 1.504120 (0.072641) | 1.411631 / 1.541195 (-0.129563) | 1.449689 / 1.468490 (-0.018801) | 0.608987 / 4.584777 (-3.975790) | 3.705121 / 3.745712 (-0.040592) | 2.085071 / 5.269862 (-3.184790) | 1.308653 / 4.565676 (-3.257024) | 0.083763 / 0.424275 (-0.340512) | 0.011957 / 0.007607 (0.004350) | 0.502182 / 0.226044 (0.276137) | 5.008829 / 2.268929 (2.739900) | 2.244687 / 55.444624 (-53.199937) | 1.891411 / 6.876477 (-4.985065) | 1.940789 / 2.142072 (-0.201284) | 0.825966 / 4.805227 (-3.979261) | 0.165267 / 6.500664 (-6.335397) | 0.063020 / 0.075469 (-0.012449) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.196707 / 1.841788 (-0.645081) | 14.236877 / 8.074308 (6.162569) | 14.872954 / 10.191392 (4.681562) | 0.168560 / 0.680424 (-0.511864) | 0.029038 / 0.534201 (-0.505163) | 0.440192 / 0.579283 (-0.139091) | 0.437021 / 0.434364 (0.002657) | 0.519612 / 0.540337 (-0.020725) | 0.612013 / 1.386936 (-0.774923) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.007170 / 0.011353 (-0.004183) | 0.005303 / 0.011008 (-0.005705) | 0.098503 / 0.038508 (0.059995) | 0.032573 / 0.023109 (0.009463) | 0.398203 / 0.275898 (0.122305) | 0.446075 / 0.323480 (0.122595) | 0.005712 / 0.007986 (-0.002274) | 0.004165 / 0.004328 (-0.000164) | 0.074273 / 0.004250 (0.070023) | 0.049587 / 0.037052 (0.012534) | 0.399458 / 0.258489 (0.140969) | 0.459167 / 0.293841 (0.165327) | 0.036063 / 0.128546 (-0.092483) | 0.012394 / 0.075646 (-0.063253) | 0.332559 / 0.419271 (-0.086713) | 0.048499 / 0.043533 (0.004967) | 0.404044 / 0.255139 (0.148905) | 0.410462 / 0.283200 (0.127262) | 0.104104 / 0.141683 (-0.037579) | 1.488141 / 1.452155 (0.035986) | 1.535517 / 1.492716 (0.042801) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.292976 / 0.018006 (0.274970) | 0.569139 / 0.000490 (0.568649) | 0.000553 / 0.000200 (0.000353) | 0.000063 / 0.000054 (0.000008) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.030144 / 0.037411 (-0.007267) | 0.098699 / 0.014526 (0.084173) | 0.114437 / 0.176557 (-0.062120) | 0.156657 / 0.737135 (-0.580478) | 0.117449 / 0.296338 (-0.178890) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.441921 / 0.215209 (0.226712) | 4.413090 / 2.077655 (2.335435) | 2.190458 / 1.504120 (0.686338) | 2.008919 / 1.541195 (0.467724) | 2.049657 / 1.468490 (0.581167) | 0.691751 / 4.584777 (-3.893026) | 3.767524 / 3.745712 (0.021812) | 3.395564 / 5.269862 (-1.874297) | 1.633480 / 4.565676 (-2.932196) | 0.084880 / 0.424275 (-0.339395) | 0.012133 / 0.007607 (0.004526) | 0.555372 / 0.226044 (0.329327) | 5.522820 / 2.268929 (3.253892) | 2.723331 / 55.444624 (-52.721293) | 2.337583 / 6.876477 (-4.538894) | 2.368746 / 2.142072 (0.226674) | 0.830127 / 4.805227 (-3.975100) | 0.166239 / 6.500664 (-6.334425) | 0.064279 / 0.075469 (-0.011190) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.123421 / 1.841788 (-0.718367) | 14.413392 / 8.074308 (6.339084) | 12.865143 / 10.191392 (2.673751) | 0.132198 / 0.680424 (-0.548226) | 0.016138 / 0.534201 (-0.518063) | 0.380760 / 0.579283 (-0.198523) | 0.387223 / 0.434364 (-0.047141) | 0.445574 / 0.540337 (-0.094764) | 0.535658 / 1.386936 (-0.851278) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#a89564d3d17b5960db2435662cb9c49f8ad7488a \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008316 / 0.011353 (-0.003037) | 0.004503 / 0.011008 (-0.006505) | 0.100565 / 0.038508 (0.062057) | 0.030388 / 0.023109 (0.007279) | 0.304417 / 0.275898 (0.028519) | 0.369655 / 0.323480 (0.046175) | 0.007796 / 0.007986 (-0.000190) | 0.003450 / 0.004328 (-0.000878) | 0.078694 / 0.004250 (0.074443) | 0.038068 / 0.037052 (0.001016) | 0.316353 / 0.258489 (0.057864) | 0.352344 / 0.293841 (0.058503) | 0.033271 / 0.128546 (-0.095276) | 0.011427 / 0.075646 (-0.064220) | 0.322367 / 0.419271 (-0.096904) | 0.041497 / 0.043533 (-0.002036) | 0.305876 / 0.255139 (0.050737) | 0.332279 / 0.283200 (0.049079) | 0.086719 / 0.141683 (-0.054964) | 1.488367 / 1.452155 (0.036212) | 1.528943 / 1.492716 (0.036227) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.171072 / 0.018006 (0.153066) | 0.421048 / 0.000490 (0.420558) | 0.003622 / 0.000200 (0.003422) | 0.000075 / 0.000054 (0.000020) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.022632 / 0.037411 (-0.014779) | 0.095304 / 0.014526 (0.080778) | 0.106254 / 0.176557 (-0.070302) | 0.138437 / 0.737135 (-0.598698) | 0.107258 / 0.296338 (-0.189080) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.423201 / 0.215209 (0.207992) | 4.208397 / 2.077655 (2.130742) | 1.899800 / 1.504120 (0.395680) | 1.682782 / 1.541195 (0.141587) | 1.708840 / 1.468490 (0.240350) | 0.694492 / 4.584777 (-3.890285) | 3.380369 / 3.745712 (-0.365344) | 1.851731 / 5.269862 (-3.418130) | 1.151615 / 4.565676 (-3.414061) | 0.082446 / 0.424275 (-0.341829) | 0.012483 / 0.007607 (0.004876) | 0.533688 / 0.226044 (0.307643) | 5.373434 / 2.268929 (3.104505) | 2.346403 / 55.444624 (-53.098221) | 1.978505 / 6.876477 (-4.897971) | 2.005875 / 2.142072 (-0.136198) | 0.820785 / 4.805227 (-3.984442) | 0.150728 / 6.500664 (-6.349936) | 0.065761 / 0.075469 (-0.009708) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.244550 / 1.841788 (-0.597237) | 13.219096 / 8.074308 (5.144788) | 13.960463 / 10.191392 (3.769071) | 0.135572 / 0.680424 (-0.544852) | 0.028746 / 0.534201 (-0.505455) | 0.393082 / 0.579283 (-0.186201) | 0.402852 / 0.434364 (-0.031512) | 0.461191 / 0.540337 (-0.079147) | 0.543500 / 1.386936 (-0.843436) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.006316 / 0.011353 (-0.005037) | 0.004394 / 0.011008 (-0.006615) | 0.096478 / 0.038508 (0.057970) | 0.026965 / 0.023109 (0.003855) | 0.340371 / 0.275898 (0.064473) | 0.368334 / 0.323480 (0.044854) | 0.004744 / 0.007986 (-0.003242) | 0.004652 / 0.004328 (0.000324) | 0.074479 / 0.004250 (0.070228) | 0.036358 / 0.037052 (-0.000694) | 0.342968 / 0.258489 (0.084479) | 0.383675 / 0.293841 (0.089834) | 0.031439 / 0.128546 (-0.097107) | 0.011529 / 0.075646 (-0.064117) | 0.319560 / 0.419271 (-0.099711) | 0.041370 / 0.043533 (-0.002163) | 0.342594 / 0.255139 (0.087455) | 0.363237 / 0.283200 (0.080038) | 0.087316 / 0.141683 (-0.054367) | 1.468690 / 1.452155 (0.016535) | 1.553974 / 1.492716 (0.061257) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.198366 / 0.018006 (0.180360) | 0.401581 / 0.000490 (0.401091) | 0.000400 / 0.000200 (0.000200) | 0.000059 / 0.000054 (0.000005) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.023150 / 0.037411 (-0.014261) | 0.097797 / 0.014526 (0.083271) | 0.106198 / 0.176557 (-0.070359) | 0.139599 / 0.737135 (-0.597536) | 0.108361 / 0.296338 (-0.187978) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.472962 / 0.215209 (0.257753) | 4.702688 / 2.077655 (2.625033) | 2.401002 / 1.504120 (0.896882) | 2.193857 / 1.541195 (0.652663) | 2.219188 / 1.468490 (0.750697) | 0.689993 / 4.584777 (-3.894784) | 3.369409 / 3.745712 (-0.376304) | 1.824801 / 5.269862 (-3.445061) | 1.150815 / 4.565676 (-3.414862) | 0.082197 / 0.424275 (-0.342078) | 0.012287 / 0.007607 (0.004679) | 0.581963 / 0.226044 (0.355918) | 5.786943 / 2.268929 (3.518015) | 2.871235 / 55.444624 (-52.573389) | 2.516009 / 6.876477 (-4.360468) | 2.535669 / 2.142072 (0.393597) | 0.804733 / 4.805227 (-4.000494) | 0.150545 / 6.500664 (-6.350119) | 0.066964 / 0.075469 (-0.008505) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.285431 / 1.841788 (-0.556356) | 14.097108 / 8.074308 (6.022800) | 13.821497 / 10.191392 (3.630105) | 0.141922 / 0.680424 (-0.538502) | 0.016964 / 0.534201 (-0.517237) | 0.374784 / 0.579283 (-0.204500) | 0.381034 / 0.434364 (-0.053330) | 0.435487 / 0.540337 (-0.104850) | 0.521894 / 1.386936 (-0.865042) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#462000c2b12a11f1fc26853e842d3f6e40287737 \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.009486 / 0.011353 (-0.001867) | 0.005363 / 0.011008 (-0.005645) | 0.101008 / 0.038508 (0.062500) | 0.036355 / 0.023109 (0.013246) | 0.290575 / 0.275898 (0.014677) | 0.391634 / 0.323480 (0.068154) | 0.009085 / 0.007986 (0.001099) | 0.005780 / 0.004328 (0.001451) | 0.077848 / 0.004250 (0.073598) | 0.049062 / 0.037052 (0.012009) | 0.310900 / 0.258489 (0.052411) | 0.358224 / 0.293841 (0.064383) | 0.038838 / 0.128546 (-0.089708) | 0.012244 / 0.075646 (-0.063402) | 0.333701 / 0.419271 (-0.085570) | 0.048021 / 0.043533 (0.004488) | 0.289584 / 0.255139 (0.034445) | 0.317556 / 0.283200 (0.034356) | 0.109807 / 0.141683 (-0.031876) | 1.465966 / 1.452155 (0.013811) | 1.526341 / 1.492716 (0.033625) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.246221 / 0.018006 (0.228215) | 0.580659 / 0.000490 (0.580169) | 0.000627 / 0.000200 (0.000427) | 0.000089 / 0.000054 (0.000035) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.028352 / 0.037411 (-0.009059) | 0.110569 / 0.014526 (0.096043) | 0.126456 / 0.176557 (-0.050100) | 0.163633 / 0.737135 (-0.573503) | 0.128252 / 0.296338 (-0.168087) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.397271 / 0.215209 (0.182062) | 3.975336 / 2.077655 (1.897682) | 1.786957 / 1.504120 (0.282837) | 1.598468 / 1.541195 (0.057273) | 1.645299 / 1.468490 (0.176809) | 0.686221 / 4.584777 (-3.898556) | 3.753184 / 3.745712 (0.007472) | 2.089505 / 5.269862 (-3.180356) | 1.325799 / 4.565676 (-3.239878) | 0.084608 / 0.424275 (-0.339667) | 0.012343 / 0.007607 (0.004736) | 0.509951 / 0.226044 (0.283907) | 5.092102 / 2.268929 (2.823174) | 2.297551 / 55.444624 (-53.147073) | 1.938177 / 6.876477 (-4.938300) | 2.012448 / 2.142072 (-0.129625) | 0.835206 / 4.805227 (-3.970021) | 0.166373 / 6.500664 (-6.334291) | 0.063996 / 0.075469 (-0.011473) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.212936 / 1.841788 (-0.628851) | 15.067370 / 8.074308 (6.993062) | 14.165214 / 10.191392 (3.973822) | 0.157041 / 0.680424 (-0.523383) | 0.029612 / 0.534201 (-0.504589) | 0.440006 / 0.579283 (-0.139277) | 0.439165 / 0.434364 (0.004801) | 0.524970 / 0.540337 (-0.015368) | 0.608305 / 1.386936 (-0.778631) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.007433 / 0.011353 (-0.003920) | 0.005310 / 0.011008 (-0.005698) | 0.097194 / 0.038508 (0.058686) | 0.033265 / 0.023109 (0.010156) | 0.369908 / 0.275898 (0.094010) | 0.411508 / 0.323480 (0.088028) | 0.006000 / 0.007986 (-0.001986) | 0.005647 / 0.004328 (0.001319) | 0.075597 / 0.004250 (0.071347) | 0.051951 / 0.037052 (0.014899) | 0.378469 / 0.258489 (0.119980) | 0.424849 / 0.293841 (0.131008) | 0.036700 / 0.128546 (-0.091846) | 0.012535 / 0.075646 (-0.063111) | 0.333197 / 0.419271 (-0.086074) | 0.049046 / 0.043533 (0.005513) | 0.381845 / 0.255139 (0.126706) | 0.397846 / 0.283200 (0.114646) | 0.109152 / 0.141683 (-0.032531) | 1.432407 / 1.452155 (-0.019748) | 1.555509 / 1.492716 (0.062793) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.265433 / 0.018006 (0.247427) | 0.559590 / 0.000490 (0.559100) | 0.000492 / 0.000200 (0.000292) | 0.000060 / 0.000054 (0.000005) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.029748 / 0.037411 (-0.007663) | 0.110490 / 0.014526 (0.095964) | 0.124125 / 0.176557 (-0.052431) | 0.160089 / 0.737135 (-0.577046) | 0.128755 / 0.296338 (-0.167583) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.443976 / 0.215209 (0.228767) | 4.416960 / 2.077655 (2.339305) | 2.239408 / 1.504120 (0.735288) | 2.055341 / 1.541195 (0.514147) | 2.093479 / 1.468490 (0.624988) | 0.688846 / 4.584777 (-3.895930) | 3.797526 / 3.745712 (0.051814) | 3.578137 / 5.269862 (-1.691725) | 2.015073 / 4.565676 (-2.550603) | 0.084126 / 0.424275 (-0.340149) | 0.012581 / 0.007607 (0.004974) | 0.549774 / 0.226044 (0.323730) | 5.492185 / 2.268929 (3.223256) | 2.739851 / 55.444624 (-52.704773) | 2.371091 / 6.876477 (-4.505386) | 2.400178 / 2.142072 (0.258105) | 0.831227 / 4.805227 (-3.974001) | 0.166156 / 6.500664 (-6.334508) | 0.063901 / 0.075469 (-0.011568) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.236127 / 1.841788 (-0.605660) | 15.236884 / 8.074308 (7.162576) | 14.434351 / 10.191392 (4.242959) | 0.163725 / 0.680424 (-0.516699) | 0.018009 / 0.534201 (-0.516192) | 0.430612 / 0.579283 (-0.148671) | 0.420426 / 0.434364 (-0.013938) | 0.497062 / 0.540337 (-0.043275) | 0.590924 / 1.386936 (-0.796012) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#63377dc53fc94f19bc2b0bbfb118a90d01a1d020 \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.010862 / 0.011353 (-0.000491) | 0.005741 / 0.011008 (-0.005267) | 0.111911 / 0.038508 (0.073403) | 0.042316 / 0.023109 (0.019207) | 0.347665 / 0.275898 (0.071767) | 0.377335 / 0.323480 (0.053855) | 0.009400 / 0.007986 (0.001414) | 0.006814 / 0.004328 (0.002486) | 0.087194 / 0.004250 (0.082943) | 0.046878 / 0.037052 (0.009826) | 0.348920 / 0.258489 (0.090430) | 0.393347 / 0.293841 (0.099507) | 0.044212 / 0.128546 (-0.084334) | 0.013925 / 0.075646 (-0.061722) | 0.386076 / 0.419271 (-0.033195) | 0.054195 / 0.043533 (0.010662) | 0.358486 / 0.255139 (0.103347) | 0.360132 / 0.283200 (0.076932) | 0.109783 / 0.141683 (-0.031900) | 1.679875 / 1.452155 (0.227720) | 1.794379 / 1.492716 (0.301663) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.221927 / 0.018006 (0.203921) | 0.487352 / 0.000490 (0.486863) | 0.003494 / 0.000200 (0.003294) | 0.000091 / 0.000054 (0.000037) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.032201 / 0.037411 (-0.005210) | 0.125861 / 0.014526 (0.111335) | 0.133905 / 0.176557 (-0.042652) | 0.183319 / 0.737135 (-0.553817) | 0.142646 / 0.296338 (-0.153693) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.442720 / 0.215209 (0.227511) | 4.602619 / 2.077655 (2.524964) | 2.050214 / 1.504120 (0.546094) | 1.837968 / 1.541195 (0.296773) | 1.961199 / 1.468490 (0.492709) | 0.793426 / 4.584777 (-3.791351) | 4.472078 / 3.745712 (0.726366) | 2.364903 / 5.269862 (-2.904959) | 1.515076 / 4.565676 (-3.050600) | 0.103087 / 0.424275 (-0.321188) | 0.014676 / 0.007607 (0.007068) | 0.576887 / 0.226044 (0.350843) | 5.785525 / 2.268929 (3.516596) | 2.765231 / 55.444624 (-52.679393) | 2.365364 / 6.876477 (-4.511113) | 2.448335 / 2.142072 (0.306262) | 0.978726 / 4.805227 (-3.826501) | 0.191417 / 6.500664 (-6.309247) | 0.073295 / 0.075469 (-0.002174) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.378995 / 1.841788 (-0.462792) | 16.583655 / 8.074308 (8.509347) | 14.944731 / 10.191392 (4.753339) | 0.168916 / 0.680424 (-0.511508) | 0.035272 / 0.534201 (-0.498928) | 0.489729 / 0.579283 (-0.089554) | 0.496231 / 0.434364 (0.061867) | 0.576218 / 0.540337 (0.035880) | 0.673558 / 1.386936 (-0.713378) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008104 / 0.011353 (-0.003249) | 0.005179 / 0.011008 (-0.005829) | 0.103908 / 0.038508 (0.065400) | 0.034661 / 0.023109 (0.011552) | 0.398119 / 0.275898 (0.122221) | 0.411765 / 0.323480 (0.088286) | 0.006016 / 0.007986 (-0.001970) | 0.005637 / 0.004328 (0.001308) | 0.073662 / 0.004250 (0.069412) | 0.052411 / 0.037052 (0.015359) | 0.391826 / 0.258489 (0.133337) | 0.455217 / 0.293841 (0.161376) | 0.039924 / 0.128546 (-0.088622) | 0.013390 / 0.075646 (-0.062256) | 0.390319 / 0.419271 (-0.028953) | 0.054312 / 0.043533 (0.010779) | 0.395492 / 0.255139 (0.140353) | 0.446324 / 0.283200 (0.163124) | 0.116461 / 0.141683 (-0.025222) | 1.502163 / 1.452155 (0.050008) | 1.731541 / 1.492716 (0.238825) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.282612 / 0.018006 (0.264606) | 0.503170 / 0.000490 (0.502680) | 0.005307 / 0.000200 (0.005107) | 0.000100 / 0.000054 (0.000046) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.029071 / 0.037411 (-0.008340) | 0.123831 / 0.014526 (0.109306) | 0.133284 / 0.176557 (-0.043272) | 0.172029 / 0.737135 (-0.565106) | 0.140639 / 0.296338 (-0.155700) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.496812 / 0.215209 (0.281603) | 4.958915 / 2.077655 (2.881260) | 2.559188 / 1.504120 (1.055068) | 2.262434 / 1.541195 (0.721240) | 2.371126 / 1.468490 (0.902636) | 0.780150 / 4.584777 (-3.804627) | 4.417060 / 3.745712 (0.671348) | 2.401909 / 5.269862 (-2.867953) | 1.527943 / 4.565676 (-3.037733) | 0.100074 / 0.424275 (-0.324201) | 0.014853 / 0.007607 (0.007246) | 0.630192 / 0.226044 (0.404147) | 6.409685 / 2.268929 (4.140757) | 3.224718 / 55.444624 (-52.219906) | 2.795301 / 6.876477 (-4.081176) | 2.927205 / 2.142072 (0.785132) | 0.989537 / 4.805227 (-3.815690) | 0.199775 / 6.500664 (-6.300889) | 0.076725 / 0.075469 (0.001256) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.433504 / 1.841788 (-0.408284) | 17.117134 / 8.074308 (9.042825) | 16.606367 / 10.191392 (6.414975) | 0.165653 / 0.680424 (-0.514771) | 0.020818 / 0.534201 (-0.513383) | 0.496782 / 0.579283 (-0.082501) | 0.473895 / 0.434364 (0.039531) | 0.576796 / 0.540337 (0.036459) | 0.703272 / 1.386936 (-0.683664) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#6627fb6f2639ac3b1435b3386545612db038a42e \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.012501 / 0.011353 (0.001148) | 0.006437 / 0.011008 (-0.004571) | 0.129387 / 0.038508 (0.090878) | 0.035847 / 0.023109 (0.012737) | 0.339243 / 0.275898 (0.063345) | 0.423274 / 0.323480 (0.099794) | 0.008489 / 0.007986 (0.000503) | 0.004596 / 0.004328 (0.000268) | 0.103322 / 0.004250 (0.099071) | 0.043570 / 0.037052 (0.006517) | 0.357004 / 0.258489 (0.098515) | 0.426511 / 0.293841 (0.132670) | 0.062923 / 0.128546 (-0.065623) | 0.021168 / 0.075646 (-0.054478) | 0.387485 / 0.419271 (-0.031787) | 0.059745 / 0.043533 (0.016213) | 0.341101 / 0.255139 (0.085962) | 0.365530 / 0.283200 (0.082331) | 0.102110 / 0.141683 (-0.039573) | 1.729408 / 1.452155 (0.277253) | 1.759510 / 1.492716 (0.266794) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.187065 / 0.018006 (0.169059) | 0.499685 / 0.000490 (0.499196) | 0.004677 / 0.000200 (0.004478) | 0.000120 / 0.000054 (0.000065) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.025827 / 0.037411 (-0.011584) | 0.113780 / 0.014526 (0.099255) | 0.146060 / 0.176557 (-0.030496) | 0.158169 / 0.737135 (-0.578966) | 0.136133 / 0.296338 (-0.160206) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.608421 / 0.215209 (0.393211) | 5.907395 / 2.077655 (3.829741) | 2.193140 / 1.504120 (0.689021) | 1.870315 / 1.541195 (0.329120) | 1.885660 / 1.468490 (0.417170) | 1.227637 / 4.584777 (-3.357140) | 5.319242 / 3.745712 (1.573530) | 2.991595 / 5.269862 (-2.278267) | 2.043906 / 4.565676 (-2.521771) | 0.151829 / 0.424275 (-0.272447) | 0.018974 / 0.007607 (0.011367) | 0.778035 / 0.226044 (0.551991) | 7.705796 / 2.268929 (5.436868) | 2.990156 / 55.444624 (-52.454468) | 2.372643 / 6.876477 (-4.503834) | 2.240847 / 2.142072 (0.098775) | 1.407209 / 4.805227 (-3.398018) | 0.242336 / 6.500664 (-6.258328) | 0.069847 / 0.075469 (-0.005622) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.445817 / 1.841788 (-0.395970) | 16.059632 / 8.074308 (7.985324) | 18.541971 / 10.191392 (8.350579) | 0.237830 / 0.680424 (-0.442594) | 0.041060 / 0.534201 (-0.493141) | 0.496765 / 0.579283 (-0.082518) | 0.609666 / 0.434364 (0.175302) | 0.584614 / 0.540337 (0.044277) | 0.680858 / 1.386936 (-0.706078) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.009037 / 0.011353 (-0.002315) | 0.005961 / 0.011008 (-0.005047) | 0.127204 / 0.038508 (0.088696) | 0.030664 / 0.023109 (0.007555) | 0.417968 / 0.275898 (0.142070) | 0.515316 / 0.323480 (0.191836) | 0.006549 / 0.007986 (-0.001436) | 0.004456 / 0.004328 (0.000128) | 0.083715 / 0.004250 (0.079464) | 0.043701 / 0.037052 (0.006648) | 0.521153 / 0.258489 (0.262664) | 0.565456 / 0.293841 (0.271615) | 0.055298 / 0.128546 (-0.073248) | 0.018103 / 0.075646 (-0.057544) | 0.403990 / 0.419271 (-0.015282) | 0.060162 / 0.043533 (0.016629) | 0.486383 / 0.255139 (0.231244) | 0.470342 / 0.283200 (0.187142) | 0.102269 / 0.141683 (-0.039414) | 1.643241 / 1.452155 (0.191086) | 1.763850 / 1.492716 (0.271133) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.185602 / 0.018006 (0.167596) | 0.489163 / 0.000490 (0.488674) | 0.000426 / 0.000200 (0.000226) | 0.000086 / 0.000054 (0.000031) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.026689 / 0.037411 (-0.010722) | 0.111520 / 0.014526 (0.096994) | 0.119838 / 0.176557 (-0.056719) | 0.153698 / 0.737135 (-0.583437) | 0.130969 / 0.296338 (-0.165370) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.616170 / 0.215209 (0.400961) | 6.219702 / 2.077655 (4.142048) | 2.533554 / 1.504120 (1.029434) | 2.256009 / 1.541195 (0.714815) | 2.217617 / 1.468490 (0.749127) | 1.156920 / 4.584777 (-3.427857) | 5.175759 / 3.745712 (1.430046) | 2.848419 / 5.269862 (-2.421442) | 1.943864 / 4.565676 (-2.621813) | 0.138342 / 0.424275 (-0.285933) | 0.013140 / 0.007607 (0.005533) | 0.782105 / 0.226044 (0.556060) | 7.602003 / 2.268929 (5.333075) | 3.629577 / 55.444624 (-51.815047) | 2.713849 / 6.876477 (-4.162628) | 2.663888 / 2.142072 (0.521816) | 1.418381 / 4.805227 (-3.386847) | 0.250649 / 6.500664 (-6.250015) | 0.073564 / 0.075469 (-0.001905) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.483739 / 1.841788 (-0.358049) | 16.386204 / 8.074308 (8.311896) | 20.685262 / 10.191392 (10.493870) | 0.237084 / 0.680424 (-0.443340) | 0.039097 / 0.534201 (-0.495104) | 0.525399 / 0.579283 (-0.053884) | 0.587541 / 0.434364 (0.153177) | 0.566605 / 0.540337 (0.026268) | 0.677384 / 1.386936 (-0.709552) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#b3b67d42733dabb15ce4997c8324f8e047ce12bd \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.014050 / 0.011353 (0.002697) | 0.005981 / 0.011008 (-0.005028) | 0.126307 / 0.038508 (0.087799) | 0.035400 / 0.023109 (0.012290) | 0.387821 / 0.275898 (0.111923) | 0.462785 / 0.323480 (0.139305) | 0.009427 / 0.007986 (0.001441) | 0.005081 / 0.004328 (0.000753) | 0.097273 / 0.004250 (0.093023) | 0.044699 / 0.037052 (0.007647) | 0.396025 / 0.258489 (0.137536) | 0.450137 / 0.293841 (0.156296) | 0.055660 / 0.128546 (-0.072886) | 0.022710 / 0.075646 (-0.052936) | 0.443784 / 0.419271 (0.024513) | 0.065756 / 0.043533 (0.022223) | 0.379350 / 0.255139 (0.124211) | 0.396783 / 0.283200 (0.113583) | 0.114088 / 0.141683 (-0.027594) | 1.856834 / 1.452155 (0.404679) | 1.839292 / 1.492716 (0.346576) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.206748 / 0.018006 (0.188742) | 0.517711 / 0.000490 (0.517222) | 0.008302 / 0.000200 (0.008102) | 0.000494 / 0.000054 (0.000440) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.033987 / 0.037411 (-0.003424) | 0.131067 / 0.014526 (0.116542) | 0.155539 / 0.176557 (-0.021018) | 0.188598 / 0.737135 (-0.548537) | 0.156000 / 0.296338 (-0.140338) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.641413 / 0.215209 (0.426204) | 6.156680 / 2.077655 (4.079025) | 2.428858 / 1.504120 (0.924738) | 2.086195 / 1.541195 (0.545000) | 2.109604 / 1.468490 (0.641114) | 1.209426 / 4.584777 (-3.375351) | 5.139398 / 3.745712 (1.393686) | 3.041337 / 5.269862 (-2.228524) | 2.294809 / 4.565676 (-2.270868) | 0.142206 / 0.424275 (-0.282069) | 0.015167 / 0.007607 (0.007560) | 0.816269 / 0.226044 (0.590224) | 7.953931 / 2.268929 (5.685002) | 3.201793 / 55.444624 (-52.242832) | 2.448620 / 6.876477 (-4.427857) | 2.521670 / 2.142072 (0.379597) | 1.484094 / 4.805227 (-3.321133) | 0.255069 / 6.500664 (-6.245595) | 0.076031 / 0.075469 (0.000561) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.590951 / 1.841788 (-0.250836) | 17.661353 / 8.074308 (9.587045) | 21.097837 / 10.191392 (10.906445) | 0.229265 / 0.680424 (-0.451159) | 0.042618 / 0.534201 (-0.491583) | 0.535942 / 0.579283 (-0.043342) | 0.590195 / 0.434364 (0.155831) | 0.623985 / 0.540337 (0.083648) | 0.742637 / 1.386936 (-0.644299) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.009264 / 0.011353 (-0.002088) | 0.008798 / 0.011008 (-0.002210) | 0.122208 / 0.038508 (0.083700) | 0.034835 / 0.023109 (0.011726) | 0.462618 / 0.275898 (0.186720) | 0.505632 / 0.323480 (0.182152) | 0.006320 / 0.007986 (-0.001665) | 0.005383 / 0.004328 (0.001054) | 0.091229 / 0.004250 (0.086979) | 0.045828 / 0.037052 (0.008775) | 0.477507 / 0.258489 (0.219018) | 0.539616 / 0.293841 (0.245775) | 0.061913 / 0.128546 (-0.066633) | 0.019390 / 0.075646 (-0.056257) | 0.420016 / 0.419271 (0.000745) | 0.065958 / 0.043533 (0.022425) | 0.468603 / 0.255139 (0.213464) | 0.486246 / 0.283200 (0.203046) | 0.107924 / 0.141683 (-0.033759) | 1.843614 / 1.452155 (0.391459) | 1.988159 / 1.492716 (0.495442) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.247043 / 0.018006 (0.229037) | 0.515580 / 0.000490 (0.515090) | 0.005630 / 0.000200 (0.005430) | 0.000115 / 0.000054 (0.000060) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.030674 / 0.037411 (-0.006737) | 0.130783 / 0.014526 (0.116258) | 0.147669 / 0.176557 (-0.028888) | 0.175656 / 0.737135 (-0.561479) | 0.138317 / 0.296338 (-0.158022) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.727119 / 0.215209 (0.511909) | 6.848208 / 2.077655 (4.770553) | 3.121418 / 1.504120 (1.617298) | 2.701799 / 1.541195 (1.160604) | 2.749179 / 1.468490 (1.280689) | 1.312058 / 4.584777 (-3.272719) | 5.400562 / 3.745712 (1.654850) | 3.058142 / 5.269862 (-2.211719) | 2.076361 / 4.565676 (-2.489316) | 0.142169 / 0.424275 (-0.282106) | 0.014340 / 0.007607 (0.006733) | 0.853534 / 0.226044 (0.627490) | 8.734484 / 2.268929 (6.465556) | 3.968130 / 55.444624 (-51.476495) | 3.118032 / 6.876477 (-3.758444) | 3.078757 / 2.142072 (0.936684) | 1.460694 / 4.805227 (-3.344533) | 0.261858 / 6.500664 (-6.238806) | 0.081089 / 0.075469 (0.005620) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.611473 / 1.841788 (-0.230315) | 17.660545 / 8.074308 (9.586237) | 20.526023 / 10.191392 (10.334631) | 0.223320 / 0.680424 (-0.457103) | 0.027939 / 0.534201 (-0.506261) | 0.542704 / 0.579283 (-0.036579) | 0.563826 / 0.434364 (0.129462) | 0.639936 / 0.540337 (0.099599) | 0.755974 / 1.386936 (-0.630962) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#942141e13ba2be853e2231d9edbfa38044e2632d \"CML watermark\")\n", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008776 / 0.011353 (-0.002577) | 0.004532 / 0.011008 (-0.006476) | 0.100373 / 0.038508 (0.061865) | 0.029706 / 0.023109 (0.006597) | 0.304374 / 0.275898 (0.028476) | 0.337223 / 0.323480 (0.013743) | 0.007021 / 0.007986 (-0.000965) | 0.003420 / 0.004328 (-0.000908) | 0.077754 / 0.004250 (0.073504) | 0.034411 / 0.037052 (-0.002642) | 0.302926 / 0.258489 (0.044437) | 0.342654 / 0.293841 (0.048813) | 0.034528 / 0.128546 (-0.094018) | 0.011926 / 0.075646 (-0.063721) | 0.322971 / 0.419271 (-0.096301) | 0.041384 / 0.043533 (-0.002149) | 0.306433 / 0.255139 (0.051294) | 0.332293 / 0.283200 (0.049093) | 0.084972 / 0.141683 (-0.056711) | 1.493426 / 1.452155 (0.041271) | 1.570446 / 1.492716 (0.077729) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.189090 / 0.018006 (0.171084) | 0.433904 / 0.000490 (0.433414) | 0.001323 / 0.000200 (0.001124) | 0.000073 / 0.000054 (0.000019) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.023531 / 0.037411 (-0.013880) | 0.097774 / 0.014526 (0.083248) | 0.106383 / 0.176557 (-0.070174) | 0.139158 / 0.737135 (-0.597977) | 0.109443 / 0.296338 (-0.186896) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.419078 / 0.215209 (0.203869) | 4.182657 / 2.077655 (2.105002) | 1.887276 / 1.504120 (0.383156) | 1.679542 / 1.541195 (0.138347) | 1.718035 / 1.468490 (0.249545) | 0.692628 / 4.584777 (-3.892149) | 3.361354 / 3.745712 (-0.384358) | 1.928583 / 5.269862 (-3.341278) | 1.317291 / 4.565676 (-3.248386) | 0.081799 / 0.424275 (-0.342476) | 0.012318 / 0.007607 (0.004711) | 0.525927 / 0.226044 (0.299883) | 5.285905 / 2.268929 (3.016977) | 2.317524 / 55.444624 (-53.127100) | 1.966478 / 6.876477 (-4.909998) | 2.054869 / 2.142072 (-0.087204) | 0.807579 / 4.805227 (-3.997649) | 0.149854 / 6.500664 (-6.350810) | 0.065285 / 0.075469 (-0.010184) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.180516 / 1.841788 (-0.661271) | 13.889734 / 8.074308 (5.815426) | 14.076163 / 10.191392 (3.884771) | 0.156276 / 0.680424 (-0.524148) | 0.029187 / 0.534201 (-0.505013) | 0.403859 / 0.579283 (-0.175424) | 0.404998 / 0.434364 (-0.029366) | 0.471467 / 0.540337 (-0.068871) | 0.564526 / 1.386936 (-0.822410) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.006739 / 0.011353 (-0.004614) | 0.004644 / 0.011008 (-0.006364) | 0.097326 / 0.038508 (0.058818) | 0.027728 / 0.023109 (0.004619) | 0.413537 / 0.275898 (0.137639) | 0.452012 / 0.323480 (0.128532) | 0.005346 / 0.007986 (-0.002639) | 0.003338 / 0.004328 (-0.000991) | 0.075670 / 0.004250 (0.071420) | 0.038825 / 0.037052 (0.001772) | 0.415612 / 0.258489 (0.157123) | 0.454680 / 0.293841 (0.160839) | 0.031866 / 0.128546 (-0.096680) | 0.011616 / 0.075646 (-0.064031) | 0.319527 / 0.419271 (-0.099745) | 0.041283 / 0.043533 (-0.002250) | 0.412046 / 0.255139 (0.156907) | 0.435244 / 0.283200 (0.152044) | 0.088400 / 0.141683 (-0.053283) | 1.478125 / 1.452155 (0.025970) | 1.553677 / 1.492716 (0.060960) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.229919 / 0.018006 (0.211913) | 0.415446 / 0.000490 (0.414956) | 0.000386 / 0.000200 (0.000186) | 0.000058 / 0.000054 (0.000003) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.024365 / 0.037411 (-0.013046) | 0.098225 / 0.014526 (0.083699) | 0.106674 / 0.176557 (-0.069883) | 0.144755 / 0.737135 (-0.592380) | 0.109221 / 0.296338 (-0.187117) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.457665 / 0.215209 (0.242456) | 4.597849 / 2.077655 (2.520195) | 2.171275 / 1.504120 (0.667155) | 1.945547 / 1.541195 (0.404352) | 2.014043 / 1.468490 (0.545553) | 0.699732 / 4.584777 (-3.885045) | 3.420711 / 3.745712 (-0.325001) | 3.298702 / 5.269862 (-1.971159) | 1.390324 / 4.565676 (-3.175353) | 0.082668 / 0.424275 (-0.341607) | 0.012556 / 0.007607 (0.004949) | 0.550406 / 0.226044 (0.324361) | 5.501060 / 2.268929 (3.232132) | 2.659841 / 55.444624 (-52.784783) | 2.243443 / 6.876477 (-4.633034) | 2.266006 / 2.142072 (0.123934) | 0.806295 / 4.805227 (-3.998933) | 0.151399 / 6.500664 (-6.349265) | 0.067048 / 0.075469 (-0.008421) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.291404 / 1.841788 (-0.550384) | 14.164728 / 8.074308 (6.090419) | 13.980219 / 10.191392 (3.788827) | 0.140599 / 0.680424 (-0.539824) | 0.016880 / 0.534201 (-0.517321) | 0.379073 / 0.579283 (-0.200210) | 0.385770 / 0.434364 (-0.048594) | 0.442516 / 0.540337 (-0.097822) | 0.533569 / 1.386936 (-0.853367) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#29fa15df972353f51fc434cf8eceb574b60a415f \"CML watermark\")\n", "Tests seem to be failing for unrelated reasons.", "Tests are failing because of a bug on the Hub side - this is being fixed :)\r\n\r\nlmk once the TF documentation page is updated and we can merge !", "@lhoestq Docs updated!" ]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/datasets/pr_5376). All of your documentation changes will be reflected on that endpoint." ]

[ "_The documentation is not available anymore as the PR was closed or merged._" ]

[ "The data files are hosted on HF at https://huggingface.co/datasets/allenai/c4/tree/main\r\n\r\nYou have 200 runs streaming the same files in parallel. So this is probably a Hub limitation. Maybe rate limiting ? cc @julien-c \r\n\r\nMaybe you can also try to reduce the number of HTTP requests by increasing the block size of each request. This can be done by increasing `DEFAULT_BLOCK_SIZE` in `fsspec.implementations.http`. Default is `5 * 2**20` (5MiB)\r\n\r\nAnyway maybe it's just better to save the dataset locally in that case ?", "you don't get an HTTP error code or something in your stack trace? Kinda hard to debug with this info", "You could try to re-run using this `datasets` branch: [raise-err-when-disconnect](https://github.com/huggingface/datasets/compare/raise-err-when-disconnect?expand=1)\r\nIt should raise the fsspec error", "The weird thing is that I already have it saved locally & it seems to indeed be using the cached one 🧐 ; I'm also using offline mode, so I don't think it has something to do with the Hub.\r\n```\r\nWARNING:datasets.load:Using the latest cached version of the module from /users/muennighoff/.cache/huggingface/modules/datasets_modules/datasets/c4/df532b158939272d032cc63ef19cd5b83e9b4d00c922b833e4cb18b2e9869b01 (last modified on Mon Dec 12 10:45:02 2022) since it couldn't be found locally at c4.\r\n```\r\n\r\n", "No, you passed `streaming=True` so it streams the data from the Hub.\r\nThis message just shows that you use the cached version of the `c4` **module**, aka the python script that is run to generate the examples from the raw data files.\r\n\r\nMaybe the offline mode should also disable `fsspec`/`aiohttp` HTTP calls in `datasets` and not just the `requests` ones.", "> This message just shows that you use the cached version of the c4 module\r\n\r\nAh my bad you're right about the module, but it's also using the downloaded & cached c4 dataset. There's no internet during the runs so it wouldn't work otherwise", "You don't have internet, therefore you get an error while trying to stream ;)" ]

[ "_The documentation is not available anymore as the PR was closed or merged._" ]

[ "_The documentation is not available anymore as the PR was closed or merged._", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.009517 / 0.011353 (-0.001835) | 0.005210 / 0.011008 (-0.005798) | 0.098916 / 0.038508 (0.060408) | 0.036123 / 0.023109 (0.013014) | 0.301564 / 0.275898 (0.025666) | 0.358086 / 0.323480 (0.034606) | 0.008159 / 0.007986 (0.000174) | 0.004122 / 0.004328 (-0.000206) | 0.075899 / 0.004250 (0.071648) | 0.046082 / 0.037052 (0.009030) | 0.302871 / 0.258489 (0.044382) | 0.351162 / 0.293841 (0.057321) | 0.038215 / 0.128546 (-0.090331) | 0.012026 / 0.075646 (-0.063620) | 0.330988 / 0.419271 (-0.088284) | 0.048351 / 0.043533 (0.004818) | 0.291840 / 0.255139 (0.036701) | 0.320387 / 0.283200 (0.037187) | 0.105018 / 0.141683 (-0.036665) | 1.447158 / 1.452155 (-0.004997) | 1.491205 / 1.492716 (-0.001511) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.250870 / 0.018006 (0.232863) | 0.562974 / 0.000490 (0.562484) | 0.001789 / 0.000200 (0.001589) | 0.000252 / 0.000054 (0.000197) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.028208 / 0.037411 (-0.009203) | 0.110897 / 0.014526 (0.096371) | 0.120394 / 0.176557 (-0.056163) | 0.164980 / 0.737135 (-0.572156) | 0.126283 / 0.296338 (-0.170056) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.397922 / 0.215209 (0.182713) | 3.969233 / 2.077655 (1.891578) | 1.766422 / 1.504120 (0.262302) | 1.577503 / 1.541195 (0.036308) | 1.672344 / 1.468490 (0.203854) | 0.695708 / 4.584777 (-3.889069) | 3.770763 / 3.745712 (0.025051) | 3.369592 / 5.269862 (-1.900269) | 1.851122 / 4.565676 (-2.714554) | 0.084063 / 0.424275 (-0.340212) | 0.012156 / 0.007607 (0.004549) | 0.534639 / 0.226044 (0.308594) | 5.021955 / 2.268929 (2.753027) | 2.215438 / 55.444624 (-53.229186) | 1.890459 / 6.876477 (-4.986018) | 2.071361 / 2.142072 (-0.070712) | 0.834623 / 4.805227 (-3.970604) | 0.165588 / 6.500664 (-6.335076) | 0.064336 / 0.075469 (-0.011133) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.205651 / 1.841788 (-0.636136) | 14.916871 / 8.074308 (6.842563) | 14.559495 / 10.191392 (4.368103) | 0.166889 / 0.680424 (-0.513535) | 0.028645 / 0.534201 (-0.505556) | 0.433634 / 0.579283 (-0.145649) | 0.429849 / 0.434364 (-0.004515) | 0.508617 / 0.540337 (-0.031720) | 0.595261 / 1.386936 (-0.791675) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.007696 / 0.011353 (-0.003657) | 0.005434 / 0.011008 (-0.005574) | 0.099234 / 0.038508 (0.060725) | 0.033904 / 0.023109 (0.010795) | 0.379181 / 0.275898 (0.103283) | 0.401858 / 0.323480 (0.078379) | 0.006257 / 0.007986 (-0.001729) | 0.004406 / 0.004328 (0.000077) | 0.073174 / 0.004250 (0.068923) | 0.056033 / 0.037052 (0.018981) | 0.379375 / 0.258489 (0.120886) | 0.425928 / 0.293841 (0.132087) | 0.037476 / 0.128546 (-0.091071) | 0.012520 / 0.075646 (-0.063127) | 0.364975 / 0.419271 (-0.054297) | 0.049341 / 0.043533 (0.005808) | 0.370519 / 0.255139 (0.115380) | 0.390585 / 0.283200 (0.107385) | 0.113339 / 0.141683 (-0.028344) | 1.460575 / 1.452155 (0.008421) | 1.564951 / 1.492716 (0.072235) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.246217 / 0.018006 (0.228210) | 0.554358 / 0.000490 (0.553869) | 0.000451 / 0.000200 (0.000251) | 0.000059 / 0.000054 (0.000004) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.029557 / 0.037411 (-0.007855) | 0.110472 / 0.014526 (0.095946) | 0.122652 / 0.176557 (-0.053904) | 0.159396 / 0.737135 (-0.577739) | 0.128852 / 0.296338 (-0.167486) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.447927 / 0.215209 (0.232718) | 4.448292 / 2.077655 (2.370637) | 2.228874 / 1.504120 (0.724754) | 2.030231 / 1.541195 (0.489036) | 2.116417 / 1.468490 (0.647927) | 0.702713 / 4.584777 (-3.882064) | 3.774063 / 3.745712 (0.028351) | 3.521662 / 5.269862 (-1.748200) | 1.476700 / 4.565676 (-3.088976) | 0.084921 / 0.424275 (-0.339354) | 0.012862 / 0.007607 (0.005255) | 0.559142 / 0.226044 (0.333098) | 5.512233 / 2.268929 (3.243305) | 2.750024 / 55.444624 (-52.694600) | 2.388845 / 6.876477 (-4.487632) | 2.541786 / 2.142072 (0.399714) | 0.842256 / 4.805227 (-3.962971) | 0.168088 / 6.500664 (-6.332576) | 0.064211 / 0.075469 (-0.011258) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.239001 / 1.841788 (-0.602787) | 15.286345 / 8.074308 (7.212036) | 13.883981 / 10.191392 (3.692589) | 0.186212 / 0.680424 (-0.494212) | 0.018305 / 0.534201 (-0.515896) | 0.420459 / 0.579283 (-0.158824) | 0.421039 / 0.434364 (-0.013325) | 0.487348 / 0.540337 (-0.052989) | 0.587730 / 1.386936 (-0.799206) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png \"CML watermark\")\n" ]

[ "Ccing @nazneenrajani @lvwerra @osanseviero " ]

[ "_The documentation is not available anymore as the PR was closed or merged._", "Alright all the tests are passing - this is ready for review", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.015146 / 0.011353 (0.003793) | 0.006683 / 0.011008 (-0.004326) | 0.125994 / 0.038508 (0.087486) | 0.041345 / 0.023109 (0.018235) | 0.378609 / 0.275898 (0.102711) | 0.483139 / 0.323480 (0.159659) | 0.009669 / 0.007986 (0.001684) | 0.005143 / 0.004328 (0.000814) | 0.092015 / 0.004250 (0.087765) | 0.052728 / 0.037052 (0.015676) | 0.397166 / 0.258489 (0.138677) | 0.465820 / 0.293841 (0.171979) | 0.051025 / 0.128546 (-0.077521) | 0.018451 / 0.075646 (-0.057196) | 0.397311 / 0.419271 (-0.021960) | 0.054842 / 0.043533 (0.011309) | 0.391203 / 0.255139 (0.136064) | 0.412743 / 0.283200 (0.129543) | 0.111356 / 0.141683 (-0.030327) | 1.697526 / 1.452155 (0.245372) | 1.795017 / 1.492716 (0.302301) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.253737 / 0.018006 (0.235731) | 0.583071 / 0.000490 (0.582581) | 0.005958 / 0.000200 (0.005758) | 0.000110 / 0.000054 (0.000056) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.030397 / 0.037411 (-0.007014) | 0.112242 / 0.014526 (0.097716) | 0.138807 / 0.176557 (-0.037749) | 0.209820 / 0.737135 (-0.527316) | 0.139530 / 0.296338 (-0.156808) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.574111 / 0.215209 (0.358902) | 5.623713 / 2.077655 (3.546058) | 2.416880 / 1.504120 (0.912760) | 1.951013 / 1.541195 (0.409819) | 2.124565 / 1.468490 (0.656075) | 1.268854 / 4.584777 (-3.315923) | 5.942368 / 3.745712 (2.196656) | 5.413814 / 5.269862 (0.143952) | 2.931638 / 4.565676 (-1.634038) | 0.135070 / 0.424275 (-0.289205) | 0.014290 / 0.007607 (0.006683) | 0.708384 / 0.226044 (0.482340) | 7.487994 / 2.268929 (5.219065) | 3.074210 / 55.444624 (-52.370414) | 2.380583 / 6.876477 (-4.495893) | 2.522298 / 2.142072 (0.380226) | 1.336741 / 4.805227 (-3.468486) | 0.236761 / 6.500664 (-6.263903) | 0.076592 / 0.075469 (0.001123) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.629415 / 1.841788 (-0.212373) | 19.000640 / 8.074308 (10.926332) | 21.474058 / 10.191392 (11.282666) | 0.231227 / 0.680424 (-0.449197) | 0.046213 / 0.534201 (-0.487988) | 0.565703 / 0.579283 (-0.013580) | 0.662956 / 0.434364 (0.228592) | 0.656475 / 0.540337 (0.116137) | 0.762534 / 1.386936 (-0.624402) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.010952 / 0.011353 (-0.000400) | 0.006259 / 0.011008 (-0.004749) | 0.132430 / 0.038508 (0.093922) | 0.037920 / 0.023109 (0.014811) | 0.483565 / 0.275898 (0.207667) | 0.528190 / 0.323480 (0.204710) | 0.008116 / 0.007986 (0.000130) | 0.006768 / 0.004328 (0.002440) | 0.100520 / 0.004250 (0.096270) | 0.055208 / 0.037052 (0.018155) | 0.484672 / 0.258489 (0.226183) | 0.556937 / 0.293841 (0.263096) | 0.057938 / 0.128546 (-0.070609) | 0.020821 / 0.075646 (-0.054826) | 0.430735 / 0.419271 (0.011464) | 0.066317 / 0.043533 (0.022785) | 0.496652 / 0.255139 (0.241513) | 0.502004 / 0.283200 (0.218804) | 0.125403 / 0.141683 (-0.016280) | 1.833396 / 1.452155 (0.381241) | 1.974517 / 1.492716 (0.481800) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.269198 / 0.018006 (0.251191) | 0.620314 / 0.000490 (0.619824) | 0.000535 / 0.000200 (0.000335) | 0.000083 / 0.000054 (0.000029) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.032373 / 0.037411 (-0.005039) | 0.130043 / 0.014526 (0.115517) | 0.146217 / 0.176557 (-0.030339) | 0.200187 / 0.737135 (-0.536948) | 0.152839 / 0.296338 (-0.143499) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.677478 / 0.215209 (0.462268) | 6.678856 / 2.077655 (4.601201) | 3.025870 / 1.504120 (1.521750) | 2.678196 / 1.541195 (1.137001) | 2.740640 / 1.468490 (1.272150) | 1.237163 / 4.584777 (-3.347614) | 5.752621 / 3.745712 (2.006908) | 3.170435 / 5.269862 (-2.099427) | 2.049174 / 4.565676 (-2.516502) | 0.147663 / 0.424275 (-0.276612) | 0.016107 / 0.007607 (0.008500) | 0.849666 / 0.226044 (0.623621) | 8.395212 / 2.268929 (6.126283) | 3.741120 / 55.444624 (-51.703505) | 3.102926 / 6.876477 (-3.773550) | 3.233655 / 2.142072 (1.091583) | 1.520349 / 4.805227 (-3.284878) | 0.267159 / 6.500664 (-6.233505) | 0.083646 / 0.075469 (0.008177) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.640458 / 1.841788 (-0.201330) | 19.043169 / 8.074308 (10.968861) | 22.786126 / 10.191392 (12.594734) | 0.218040 / 0.680424 (-0.462384) | 0.032948 / 0.534201 (-0.501253) | 0.569574 / 0.579283 (-0.009710) | 0.658746 / 0.434364 (0.224382) | 0.650501 / 0.540337 (0.110164) | 0.730588 / 1.386936 (-0.656348) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png \"CML watermark\")\n", "just added a note :)" ]

[ "_The documentation is not available anymore as the PR was closed or merged._" ]

[ "_The documentation is not available anymore as the PR was closed or merged._" ]

[ "_The documentation is not available anymore as the PR was closed or merged._" ]

[ "_The documentation is not available anymore as the PR was closed or merged._", "Hi, thanks for working on this! \r\n\r\nI agree that the current type-casting (always cast to `np.uint8` as Tensorflow Datasets does) is a bit too harsh. However, not all dtypes are supported in `Image.fromarray` (e.g. np.int64), so we need to treat these with special care (e.g. downcast to the closest supported dtype, maybe with warnings to let the user know what's happening).\r\n\r\nPS: To avoid the CI failures, we need to handle two more instances of the cast to `np.uint8` (both are in the `image.py` file).", "I've made some changes to the PR.\r\n\r\nNow the encoding procedure behaves as follows:\r\n* for multi-channel arrays: if their dtype is `int`/`uint`, cast to np.uint8 (the only supported dtype for multi-channel arrays), throw an error otherwise\r\n* if the array dtype is of valid kind (\"u\", \"i\", \"f\", ...):\r\n * don't do anything if Pillow natively supports it\r\n * otherwise, downcast until it becomes compatible with Pillow\r\n* raise an error if nothing from above is true", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.009537 / 0.011353 (-0.001816) | 0.004946 / 0.011008 (-0.006062) | 0.100552 / 0.038508 (0.062043) | 0.035119 / 0.023109 (0.012009) | 0.295989 / 0.275898 (0.020091) | 0.361326 / 0.323480 (0.037846) | 0.007608 / 0.007986 (-0.000378) | 0.004151 / 0.004328 (-0.000177) | 0.077301 / 0.004250 (0.073050) | 0.042921 / 0.037052 (0.005869) | 0.304804 / 0.258489 (0.046315) | 0.345934 / 0.293841 (0.052093) | 0.038987 / 0.128546 (-0.089559) | 0.012055 / 0.075646 (-0.063591) | 0.334035 / 0.419271 (-0.085236) | 0.052679 / 0.043533 (0.009146) | 0.291700 / 0.255139 (0.036561) | 0.335423 / 0.283200 (0.052223) | 0.107002 / 0.141683 (-0.034680) | 1.516780 / 1.452155 (0.064625) | 1.514137 / 1.492716 (0.021420) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.014719 / 0.018006 (-0.003287) | 0.545251 / 0.000490 (0.544761) | 0.004719 / 0.000200 (0.004519) | 0.000275 / 0.000054 (0.000220) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.026633 / 0.037411 (-0.010779) | 0.106911 / 0.014526 (0.092385) | 0.120258 / 0.176557 (-0.056299) | 0.156196 / 0.737135 (-0.580940) | 0.123132 / 0.296338 (-0.173207) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.398018 / 0.215209 (0.182809) | 3.973992 / 2.077655 (1.896337) | 1.776436 / 1.504120 (0.272316) | 1.579036 / 1.541195 (0.037841) | 1.643345 / 1.468490 (0.174855) | 0.692408 / 4.584777 (-3.892369) | 3.757243 / 3.745712 (0.011531) | 3.226212 / 5.269862 (-2.043649) | 1.797845 / 4.565676 (-2.767831) | 0.085878 / 0.424275 (-0.338398) | 0.012451 / 0.007607 (0.004844) | 0.509755 / 0.226044 (0.283711) | 5.029035 / 2.268929 (2.760107) | 2.255507 / 55.444624 (-53.189117) | 1.892868 / 6.876477 (-4.983609) | 1.900017 / 2.142072 (-0.242055) | 0.853965 / 4.805227 (-3.951263) | 0.167268 / 6.500664 (-6.333396) | 0.062796 / 0.075469 (-0.012673) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.183361 / 1.841788 (-0.658427) | 15.103797 / 8.074308 (7.029489) | 14.112931 / 10.191392 (3.921539) | 0.167234 / 0.680424 (-0.513190) | 0.029487 / 0.534201 (-0.504713) | 0.444121 / 0.579283 (-0.135162) | 0.437821 / 0.434364 (0.003457) | 0.544900 / 0.540337 (0.004562) | 0.642142 / 1.386936 (-0.744794) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.007078 / 0.011353 (-0.004275) | 0.004983 / 0.011008 (-0.006026) | 0.097106 / 0.038508 (0.058598) | 0.033747 / 0.023109 (0.010637) | 0.382030 / 0.275898 (0.106132) | 0.410193 / 0.323480 (0.086713) | 0.006658 / 0.007986 (-0.001327) | 0.005358 / 0.004328 (0.001029) | 0.073878 / 0.004250 (0.069628) | 0.049292 / 0.037052 (0.012240) | 0.384053 / 0.258489 (0.125564) | 0.427826 / 0.293841 (0.133985) | 0.036780 / 0.128546 (-0.091766) | 0.012469 / 0.075646 (-0.063178) | 0.332989 / 0.419271 (-0.086283) | 0.059531 / 0.043533 (0.015998) | 0.378431 / 0.255139 (0.123292) | 0.402672 / 0.283200 (0.119473) | 0.110782 / 0.141683 (-0.030901) | 1.484570 / 1.452155 (0.032416) | 1.608081 / 1.492716 (0.115365) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.232356 / 0.018006 (0.214350) | 0.545648 / 0.000490 (0.545158) | 0.003113 / 0.000200 (0.002913) | 0.000089 / 0.000054 (0.000035) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.028138 / 0.037411 (-0.009273) | 0.110786 / 0.014526 (0.096260) | 0.123615 / 0.176557 (-0.052941) | 0.165773 / 0.737135 (-0.571362) | 0.126401 / 0.296338 (-0.169937) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.440518 / 0.215209 (0.225309) | 4.393821 / 2.077655 (2.316166) | 2.295479 / 1.504120 (0.791359) | 2.116679 / 1.541195 (0.575485) | 2.215561 / 1.468490 (0.747071) | 0.722343 / 4.584777 (-3.862434) | 3.783360 / 3.745712 (0.037647) | 3.302242 / 5.269862 (-1.967620) | 1.681535 / 4.565676 (-2.884142) | 0.085738 / 0.424275 (-0.338537) | 0.012373 / 0.007607 (0.004766) | 0.540499 / 0.226044 (0.314455) | 5.384915 / 2.268929 (3.115986) | 2.766346 / 55.444624 (-52.678279) | 2.451994 / 6.876477 (-4.424483) | 2.505720 / 2.142072 (0.363647) | 0.833006 / 4.805227 (-3.972221) | 0.168206 / 6.500664 (-6.332458) | 0.064971 / 0.075469 (-0.010498) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.253499 / 1.841788 (-0.588289) | 15.381840 / 8.074308 (7.307532) | 13.519493 / 10.191392 (3.328101) | 0.165559 / 0.680424 (-0.514865) | 0.017682 / 0.534201 (-0.516519) | 0.422248 / 0.579283 (-0.157035) | 0.422750 / 0.434364 (-0.011614) | 0.524546 / 0.540337 (-0.015792) | 0.626956 / 1.386936 (-0.759980) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#d9a8d8af0961c473103516dd018e2d34d23cea02 \"CML watermark\")\n" ]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/datasets/pr_5364). All of your documentation changes will be reflected on that endpoint.", "Deleting `BeamPipeline` and `upload_local_to_remote` would break the existing Beam scripts, so I reverted this change.\r\n\r\nFrom what I understand, we need these components in our scripts for the pattern:\r\n```python\r\nif not pipeline.is_local():\r\n dl_manager.ship_files_with_pipeline()\r\n```\r\n\r\nI plan to address this in a subsequent PR by (implicitly) downloading the files directly to the remote storage of the non-local runners.", "I got `AttributeError: 'Pipeline' object has no attribute 'is_local'` when running\r\n```python\r\nload_dataset(\"wikipedia\", language=\"af\", date=\"20230101\", beam_runner=\"DirectRunner\")\r\n```\r\n```python\r\n~/.cache/huggingface/modules/datasets_modules/datasets/wikipedia/aa542ed919df55cc5d3347f42dd4521d05ca68751f50dbc32bae2a7f1e167559/wikipedia.py in _split_generators(self, dl_manager, pipeline)\r\n 965 # Use dictionary since testing mock always returns the same result.\r\n 966 downloaded_files = dl_manager.download({\"xml\": xml_urls})\r\n--> 967 if not pipeline.is_local():\r\n 968 downloaded_files = dl_manager.ship_files_with_pipeline(downloaded_files, pipeline)\r\n 969 \r\n\r\nAttributeError: 'Pipeline' object has no attribute 'is_local'\r\n```", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.010649 / 0.011353 (-0.000704) | 0.006116 / 0.011008 (-0.004892) | 0.115568 / 0.038508 (0.077060) | 0.041704 / 0.023109 (0.018595) | 0.360459 / 0.275898 (0.084561) | 0.425679 / 0.323480 (0.102200) | 0.008992 / 0.007986 (0.001006) | 0.006321 / 0.004328 (0.001993) | 0.090223 / 0.004250 (0.085973) | 0.049877 / 0.037052 (0.012824) | 0.382447 / 0.258489 (0.123958) | 0.406567 / 0.293841 (0.112726) | 0.045138 / 0.128546 (-0.083409) | 0.014203 / 0.075646 (-0.061444) | 0.388897 / 0.419271 (-0.030375) | 0.057176 / 0.043533 (0.013644) | 0.358729 / 0.255139 (0.103590) | 0.386086 / 0.283200 (0.102887) | 0.119221 / 0.141683 (-0.022462) | 1.731574 / 1.452155 (0.279419) | 1.744103 / 1.492716 (0.251386) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.230380 / 0.018006 (0.212373) | 0.493690 / 0.000490 (0.493201) | 0.005150 / 0.000200 (0.004950) | 0.000097 / 0.000054 (0.000042) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.030771 / 0.037411 (-0.006641) | 0.123196 / 0.014526 (0.108671) | 0.134097 / 0.176557 (-0.042459) | 0.190442 / 0.737135 (-0.546693) | 0.138416 / 0.296338 (-0.157923) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.469763 / 0.215209 (0.254554) | 4.682847 / 2.077655 (2.605192) | 2.076717 / 1.504120 (0.572597) | 1.843721 / 1.541195 (0.302527) | 1.923486 / 1.468490 (0.454996) | 0.817680 / 4.584777 (-3.767097) | 4.482409 / 3.745712 (0.736697) | 3.898695 / 5.269862 (-1.371167) | 2.078291 / 4.565676 (-2.487386) | 0.100285 / 0.424275 (-0.323990) | 0.014761 / 0.007607 (0.007154) | 0.611261 / 0.226044 (0.385217) | 5.926919 / 2.268929 (3.657990) | 2.685080 / 55.444624 (-52.759544) | 2.232179 / 6.876477 (-4.644298) | 2.305576 / 2.142072 (0.163504) | 0.993729 / 4.805227 (-3.811498) | 0.194491 / 6.500664 (-6.306173) | 0.074176 / 0.075469 (-0.001293) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.388592 / 1.841788 (-0.453196) | 17.146945 / 8.074308 (9.072636) | 15.989570 / 10.191392 (5.798178) | 0.200147 / 0.680424 (-0.480277) | 0.034009 / 0.534201 (-0.500192) | 0.517531 / 0.579283 (-0.061753) | 0.533966 / 0.434364 (0.099602) | 0.637024 / 0.540337 (0.096687) | 0.749166 / 1.386936 (-0.637770) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008240 / 0.011353 (-0.003113) | 0.006139 / 0.011008 (-0.004869) | 0.112258 / 0.038508 (0.073750) | 0.039001 / 0.023109 (0.015891) | 0.449467 / 0.275898 (0.173569) | 0.483422 / 0.323480 (0.159942) | 0.006176 / 0.007986 (-0.001810) | 0.006340 / 0.004328 (0.002012) | 0.083105 / 0.004250 (0.078855) | 0.047002 / 0.037052 (0.009950) | 0.458564 / 0.258489 (0.200075) | 0.513704 / 0.293841 (0.219863) | 0.041359 / 0.128546 (-0.087188) | 0.014515 / 0.075646 (-0.061131) | 0.392599 / 0.419271 (-0.026673) | 0.055222 / 0.043533 (0.011690) | 0.446956 / 0.255139 (0.191817) | 0.469194 / 0.283200 (0.185994) | 0.118212 / 0.141683 (-0.023471) | 1.682647 / 1.452155 (0.230492) | 1.780076 / 1.492716 (0.287360) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.259124 / 0.018006 (0.241117) | 0.507559 / 0.000490 (0.507069) | 0.001080 / 0.000200 (0.000880) | 0.000081 / 0.000054 (0.000026) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.031969 / 0.037411 (-0.005442) | 0.126997 / 0.014526 (0.112471) | 0.139593 / 0.176557 (-0.036963) | 0.182735 / 0.737135 (-0.554400) | 0.145871 / 0.296338 (-0.150468) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.530894 / 0.215209 (0.315685) | 5.284979 / 2.077655 (3.207324) | 2.592886 / 1.504120 (1.088766) | 2.407202 / 1.541195 (0.866007) | 2.434079 / 1.468490 (0.965589) | 0.829382 / 4.584777 (-3.755395) | 4.481710 / 3.745712 (0.735998) | 3.912280 / 5.269862 (-1.357581) | 1.962291 / 4.565676 (-2.603386) | 0.101840 / 0.424275 (-0.322435) | 0.014528 / 0.007607 (0.006921) | 0.639956 / 0.226044 (0.413911) | 6.414685 / 2.268929 (4.145756) | 3.240290 / 55.444624 (-52.204334) | 2.795208 / 6.876477 (-4.081269) | 2.912122 / 2.142072 (0.770050) | 0.992188 / 4.805227 (-3.813039) | 0.200701 / 6.500664 (-6.299964) | 0.074235 / 0.075469 (-0.001234) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.455075 / 1.841788 (-0.386712) | 17.186669 / 8.074308 (9.112361) | 15.404357 / 10.191392 (5.212965) | 0.168267 / 0.680424 (-0.512157) | 0.020774 / 0.534201 (-0.513427) | 0.502603 / 0.579283 (-0.076680) | 0.506500 / 0.434364 (0.072136) | 0.624245 / 0.540337 (0.083907) | 0.735529 / 1.386936 (-0.651407) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png \"CML watermark\")\n" ]

[]

[ "Thanks for reporting, @shaoyuta.\r\n\r\nWe have checked and yes, apparently there is an issue with the server hosting the data of the \"enron_emails\" subset of \"the_pile\" dataset: http://eaidata.bmk.sh/data/enron_emails.jsonl.zst\r\nIt seems to be down: The connection has timed out.\r\n\r\nPlease note that at the Hugging Face Hub, we are not hosting their data for this dataset, but only a script that downloads the data from their servers. We are updating the data URL to one in another server.\r\n\r\nIn the meantime, please note that you can train your model in the entire \"the_pile\" dataset, by passing the \"all\" config (instead of the \"enron_emails\" one).", "We have transferred this issue to the corresponding dataset Community tab: https://huggingface.co/datasets/the_pile/discussions/2\r\n\r\nPlease, follow the updates there." ]

[ "You can try something like this ?\r\n```python\r\ndef mapper_function(batch):\r\n return {\"concatenated_audio\": [np.concatenate([audio[\"array\"] for audio in batch[\"audio\"]])]}\r\n\r\ndataset = dataset.map(\r\n mapper_function,\r\n batched=True,\r\n batch_size=3,\r\n remove_columns=list(dataset.features),\r\n)\r\n```", "Thanks for the snippet!\r\n\r\nOne more question. I wonder why those two mappers are working so different that one taking 4 sec while other taking over 1 min :\r\n\r\n```python\r\n%%time\r\ndef mapper_function1(batch):\r\n # list_audio\r\n return {\r\n \"audio\": [\r\n {\r\n \"array\": np.concatenate([audio[\"array\"] for audio in batch[\"audio\"]]),\r\n \"sampling_rate\": 16_000,\r\n }\r\n ]\r\n }\r\n\r\ndataset.map(\r\n mapper_function1,\r\n batched=True,\r\n batch_size=3,\r\n remove_columns=list(dataset.features),\r\n)\r\n\r\n# 100%\r\n# 135/135 [01:13<00:00, 1.93ba/s]\r\n# CPU times: user 1min 10s, sys: 3.21 s, total: 1min 13s\r\n# Wall time: 1min 13s\r\n# Dataset({\r\n# features: ['audio'],\r\n# num_rows: 135\r\n# })\r\n\r\n# --------------------------------\r\n%%time\r\ndef mapper_function2(batch):\r\n # list_audio\r\n return {\"audio\": [np.concatenate([audio[\"array\"] for audio in batch[\"audio\"]])]}\r\n\r\ndataset.map(\r\n mapper_function2,\r\n batched=True,\r\n batch_size=3,\r\n remove_columns=list(dataset.features),\r\n)\r\n\r\n# 100%\r\n# 135/135 [00:03<00:00, 40.69ba/s]\r\n# CPU times: user 1.88 s, sys: 1.48 s, total: 3.36 s\r\n# Wall time: 4.8 s\r\n# Dataset({\r\n# features: ['audio'],\r\n# num_rows: 135\r\n# })\r\n```\r\n", "In the first one you get a dataset with an Audio type, and in the second one you get a dataset with a sequence of floats type.\r\n\r\nThe Audio type encodes the data as WAV to save disk space, so it takes more time to create.\r\nThe Audio type is automatically inferred because you modify the column \"audio\" which was already an Audio type. If you name it to something else, type inference will use a type struct with array and sampling rate fields." ]

[ "If you use huggingface trainer, you will find the trainer has wrapped a `IterableDatasetShard` to avoid duplication.\r\nSee:\r\nhttps://github.com/huggingface/transformers/blob/dfd818420dcbad68e05a502495cf666d338b2bfb/src/transformers/trainer.py#L835\r\n", "If you want to support it by datasets natively, maybe we also need to change the code in `transformers` ?", "Opened https://github.com/huggingface/transformers/issues/20770 to discuss this :)", "Maybe something like this then ?\r\n```python\r\nfrom datasets.distributed import split_dataset_by_node\r\nds = split_dataset_by_node(ds, rank=rank, world_size=world_size)\r\n```\r\n\r\nFor map-style datasets the implementation is trivial (it can simply use `.shard()`).\r\n\r\nFor iterable datasets we would need to implement a new ExamplesIterable that would only iterate on a subset of the (possibly shuffled and re-shuffled after each epoch) list of shards, based on the rank and world size.", "My plan is to skip examples by default to not end up with duplicates.\r\n\r\nAnd if a dataset has a number of shards that is a factor of the world size, then I'd make it more optimized by distributing the shards evenly across nodes instead.", "Opened a PR here: https://github.com/huggingface/datasets/pull/5369\r\n\r\nfeel free to play with it and share your feedbacks :)" ]

[ "_The documentation is not available anymore as the PR was closed or merged._", "Thanks for your proposed fix, @freddyheppell.\r\n\r\nCurrently the CI fails because in a test we pass a `tuple` instead of a `list`. I would say we should accept `tuple` as a valid input type as well...\r\n\r\nWhat about checking for `Sequence` instead?", "Fixed that @albertvillanova, can you approve CI again please? Had some issues related to Pytorch .so files when running tests on my M1 mac, so wasn't able to test locally first. Have got them working on my desktop now though." ]

[ "_The documentation is not available anymore as the PR was closed or merged._", "@mariosasko Sorry, I didn't check tests/style after doing a merge from the Git UI last week. Thx for fixing. \r\n\r\nFYI I'm getting \"Only those with [write access](https://docs.github.com/articles/what-are-the-different-access-permissions) to this repository can merge pull requests.\" so it seems somebody else needs to merge this.", "<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008816 / 0.011353 (-0.002536) | 0.004691 / 0.011008 (-0.006317) | 0.100039 / 0.038508 (0.061531) | 0.035422 / 0.023109 (0.012313) | 0.312600 / 0.275898 (0.036702) | 0.378684 / 0.323480 (0.055204) | 0.007593 / 0.007986 (-0.000392) | 0.005183 / 0.004328 (0.000855) | 0.078040 / 0.004250 (0.073790) | 0.041845 / 0.037052 (0.004793) | 0.325251 / 0.258489 (0.066762) | 0.363459 / 0.293841 (0.069618) | 0.038006 / 0.128546 (-0.090540) | 0.011911 / 0.075646 (-0.063735) | 0.335020 / 0.419271 (-0.084251) | 0.048765 / 0.043533 (0.005233) | 0.305913 / 0.255139 (0.050774) | 0.337620 / 0.283200 (0.054420) | 0.101867 / 0.141683 (-0.039816) | 1.450091 / 1.452155 (-0.002064) | 1.437303 / 1.492716 (-0.055413) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.225650 / 0.018006 (0.207644) | 0.492480 / 0.000490 (0.491990) | 0.002857 / 0.000200 (0.002658) | 0.000075 / 0.000054 (0.000020) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.026231 / 0.037411 (-0.011180) | 0.105479 / 0.014526 (0.090953) | 0.118438 / 0.176557 (-0.058119) | 0.167313 / 0.737135 (-0.569822) | 0.119416 / 0.296338 (-0.176923) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.396233 / 0.215209 (0.181024) | 3.943325 / 2.077655 (1.865671) | 1.778864 / 1.504120 (0.274744) | 1.587957 / 1.541195 (0.046763) | 1.615404 / 1.468490 (0.146914) | 0.709427 / 4.584777 (-3.875350) | 3.823310 / 3.745712 (0.077598) | 3.461376 / 5.269862 (-1.808486) | 1.888330 / 4.565676 (-2.677346) | 0.086910 / 0.424275 (-0.337365) | 0.012215 / 0.007607 (0.004608) | 0.504877 / 0.226044 (0.278833) | 5.051513 / 2.268929 (2.782584) | 2.249389 / 55.444624 (-53.195235) | 1.890949 / 6.876477 (-4.985528) | 2.015584 / 2.142072 (-0.126489) | 0.862313 / 4.805227 (-3.942914) | 0.166295 / 6.500664 (-6.334369) | 0.061131 / 0.075469 (-0.014338) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.201804 / 1.841788 (-0.639984) | 14.589425 / 8.074308 (6.515117) | 13.855522 / 10.191392 (3.664130) | 0.193406 / 0.680424 (-0.487018) | 0.028614 / 0.534201 (-0.505587) | 0.439857 / 0.579283 (-0.139426) | 0.443330 / 0.434364 (0.008966) | 0.514078 / 0.540337 (-0.026259) | 0.608245 / 1.386936 (-0.778691) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.007087 / 0.011353 (-0.004265) | 0.005024 / 0.011008 (-0.005985) | 0.096852 / 0.038508 (0.058344) | 0.032870 / 0.023109 (0.009761) | 0.397790 / 0.275898 (0.121892) | 0.420717 / 0.323480 (0.097237) | 0.005552 / 0.007986 (-0.002434) | 0.003742 / 0.004328 (-0.000586) | 0.074788 / 0.004250 (0.070537) | 0.048030 / 0.037052 (0.010977) | 0.398520 / 0.258489 (0.140031) | 0.460919 / 0.293841 (0.167078) | 0.037652 / 0.128546 (-0.090894) | 0.012249 / 0.075646 (-0.063397) | 0.333077 / 0.419271 (-0.086194) | 0.052364 / 0.043533 (0.008831) | 0.394358 / 0.255139 (0.139219) | 0.414193 / 0.283200 (0.130994) | 0.103569 / 0.141683 (-0.038114) | 1.499208 / 1.452155 (0.047053) | 1.619481 / 1.492716 (0.126764) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.229476 / 0.018006 (0.211470) | 0.448670 / 0.000490 (0.448180) | 0.000399 / 0.000200 (0.000199) | 0.000056 / 0.000054 (0.000001) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.027550 / 0.037411 (-0.009862) | 0.109180 / 0.014526 (0.094654) | 0.118372 / 0.176557 (-0.058185) | 0.153136 / 0.737135 (-0.583999) | 0.122689 / 0.296338 (-0.173650) |\n\n### Benchmark: benchmark_iterating.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.445163 / 0.215209 (0.229954) | 4.426350 / 2.077655 (2.348695) | 2.194902 / 1.504120 (0.690782) | 2.019049 / 1.541195 (0.477854) | 2.032795 / 1.468490 (0.564305) | 0.700752 / 4.584777 (-3.884025) | 3.797616 / 3.745712 (0.051903) | 2.046414 / 5.269862 (-3.223447) | 1.345037 / 4.565676 (-3.220639) | 0.085389 / 0.424275 (-0.338886) | 0.012824 / 0.007607 (0.005217) | 0.553875 / 0.226044 (0.327831) | 5.550252 / 2.268929 (3.281323) | 2.702822 / 55.444624 (-52.741803) | 2.346257 / 6.876477 (-4.530220) | 2.410772 / 2.142072 (0.268699) | 0.848271 / 4.805227 (-3.956957) | 0.170787 / 6.500664 (-6.329877) | 0.064344 / 0.075469 (-0.011125) |\n\n### Benchmark: benchmark_map_filter.json\n\n| 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 |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.266222 / 1.841788 (-0.575566) | 14.501194 / 8.074308 (6.426886) | 13.413678 / 10.191392 (3.222286) | 0.589048 / 0.680424 (-0.091375) | 0.018246 / 0.534201 (-0.515955) | 0.425221 / 0.579283 (-0.154062) | 0.425900 / 0.434364 (-0.008464) | 0.494023 / 0.540337 (-0.046314) | 0.604324 / 1.386936 (-0.782612) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png \"CML watermark\")\n" ]

[ "_The documentation is not available anymore as the PR was closed or merged._", "Need some more time to fix the tests, especially with pickle", "> And I actually don't quite understand the idea - what's the motivation behind making only IterableDataset compatible with torch DataLoader without setting the format explicitly?\r\n\r\nSetting the format to pytorch = set the output types of the dataset to be pytorch tensors. However sometimes your dataset is not made of tensors but you still want to be able to use a pytorch DataLoader", "A bit more context. \r\n\r\nThe arrow-backed `Dataset` supports `DataLoader(ds)` (even if the format is not \"torch\"), and we want to be able to do the same with `IterableDataset` for consistency. However, this is when the PyTorch internals come into play - an iterable dataset needs to be an instance of `torch.utils.data.IterableDataset` due to [this](https://github.com/pytorch/pytorch/blob/abc54f93145830b502400faa92bec86e05422fbd/torch/utils/data/dataloader.py#L276) check (notice there is no check for the map-style version). Hence the explicit subclassing in this PR.", "Exactly :) Btw I just took your comments into account @polinaeterna , so feel free to review again", "@lhoestq just checking, does this change still preserve the fix to the \"data duplicate when setting num_works > 1 with streaming data\" issue from before?\r\n\r\nhttps://github.com/huggingface/datasets/issues/3423", "Yes :)" ]

[ "_The documentation is not available anymore as the PR was closed or merged._" ]

[ "_The documentation is not available anymore as the PR was closed or merged._" ]

[ "Hi! Linking a comment to provide more info on the issue: https://stackoverflow.com/a/39458225. This means we should replace all (most of) the occurrences of `List` with `Sequence` in function signatures.\r\n\r\n@tranhd95 Would you be interested in submitting a PR?", "Hi all! I tried to reproduce this issue and didn't work for me. Also in your example i noticed that the variables have different names: `list_of_filenames` and `list_of_files`, could this be related to that?\r\n```python\r\n#I found random data in parquet format:\r\n!wget \"https://github.com/Teradata/kylo/raw/master/samples/sample-data/parquet/userdata1.parquet\"\r\n!wget \"https://github.com/Teradata/kylo/raw/master/samples/sample-data/parquet/userdata2.parquet\"\r\n\r\n#Then i try reproduce\r\nlist_of_files = [\"userdata1.parquet\", \"userdata2.parquet\"]\r\nds = Dataset.from_parquet(list_of_files)\r\n```\r\n**My output:**\r\n```python\r\nWARNING:datasets.builder:Using custom data configuration default-e287d097dc54e046\r\nDownloading and preparing dataset parquet/default to /root/.cache/huggingface/datasets/parquet/default-e287d097dc54e046/0.0.0/2a3b91fbd88a2c90d1dbbb32b460cf621d31bd5b05b934492fdef7d8d6f236ec...\r\nDownloading data files: 100%\r\n1/1 [00:00<00:00, 40.38it/s]\r\nExtracting data files: 100%\r\n1/1 [00:00<00:00, 23.43it/s]\r\nDataset parquet downloaded and prepared to /root/.cache/huggingface/datasets/parquet/default-e287d097dc54e046/0.0.0/2a3b91fbd88a2c90d1dbbb32b460cf621d31bd5b05b934492fdef7d8d6f236ec. Subsequent calls will reuse this data.\r\n```\r\nP.S. This is my first experience with open source. So do not judge strictly if I do not understand something)", "@dantema There is indeed a typo in variable names. Nevertheless, I'm sorry if I was not clear but the output is from `mypy` type checker. You can run the code snippet without issues. The problem is with the type checking.", "However, I found out that the type annotation is actually misleading. The [`from_parquet`](https://github.com/huggingface/datasets/blob/5ef1ab1cc06c2b7a574bf2df454cd9fcb071ccb2/src/datasets/arrow_dataset.py#L1039) method should also accept list of [`PathLike`](https://github.com/huggingface/datasets/blob/main/src/datasets/utils/typing.py#L8) objects which includes [`os.PathLike`](https://docs.python.org/3/library/os.html#os.PathLike). But if I would ran the code snippet below, an exception is thrown.\r\n\r\n**Code**\r\n```py\r\nfrom pathlib import Path\r\n\r\nlist_of_filenames = [Path(\"foo.parquet\"), Path(\"bar.parquet\")]\r\nds = Dataset.from_parquet(list_of_filenames)\r\n```\r\n**Output**\r\n```py\r\n[/usr/local/lib/python3.8/dist-packages/datasets/arrow_dataset.py](https://localhost:8080/#) in from_parquet(path_or_paths, split, features, cache_dir, keep_in_memory, columns, **kwargs)\r\n 1071 from .io.parquet import ParquetDatasetReader\r\n 1072 \r\n-> 1073 return ParquetDatasetReader(\r\n 1074 path_or_paths,\r\n 1075 split=split,\r\n\r\n[/usr/local/lib/python3.8/dist-packages/datasets/io/parquet.py](https://localhost:8080/#) in __init__(self, path_or_paths, split, features, cache_dir, keep_in_memory, streaming, **kwargs)\r\n 35 path_or_paths = path_or_paths if isinstance(path_or_paths, dict) else {self.split: path_or_paths}\r\n 36 hash = _PACKAGED_DATASETS_MODULES[\"parquet\"][1]\r\n---> 37 self.builder = Parquet(\r\n 38 cache_dir=cache_dir,\r\n 39 data_files=path_or_paths,\r\n\r\n[/usr/local/lib/python3.8/dist-packages/datasets/builder.py](https://localhost:8080/#) in __init__(self, cache_dir, config_name, hash, base_path, info, features, use_auth_token, repo_id, data_files, data_dir, name, **config_kwargs)\r\n 298 \r\n 299 if data_files is not None and not isinstance(data_files, DataFilesDict):\r\n--> 300 data_files = DataFilesDict.from_local_or_remote(\r\n 301 sanitize_patterns(data_files), base_path=base_path, use_auth_token=use_auth_token\r\n 302 )\r\n\r\n[/usr/local/lib/python3.8/dist-packages/datasets/data_files.py](https://localhost:8080/#) in from_local_or_remote(cls, patterns, base_path, allowed_extensions, use_auth_token)\r\n 794 for key, patterns_for_key in patterns.items():\r\n 795 out[key] = (\r\n--> 796 DataFilesList.from_local_or_remote(\r\n 797 patterns_for_key,\r\n 798 base_path=base_path,\r\n\r\n[/usr/local/lib/python3.8/dist-packages/datasets/data_files.py](https://localhost:8080/#) in from_local_or_remote(cls, patterns, base_path, allowed_extensions, use_auth_token)\r\n 762 ) -> \"DataFilesList\":\r\n 763 base_path = base_path if base_path is not None else str(Path().resolve())\r\n--> 764 data_files = resolve_patterns_locally_or_by_urls(base_path, patterns, allowed_extensions)\r\n 765 origin_metadata = _get_origin_metadata_locally_or_by_urls(data_files, use_auth_token=use_auth_token)\r\n 766 return cls(data_files, origin_metadata)\r\n\r\n[/usr/local/lib/python3.8/dist-packages/datasets/data_files.py](https://localhost:8080/#) in resolve_patterns_locally_or_by_urls(base_path, patterns, allowed_extensions)\r\n 357 data_files = []\r\n 358 for pattern in patterns:\r\n--> 359 if is_remote_url(pattern):\r\n 360 data_files.append(Url(pattern))\r\n 361 else:\r\n\r\n[/usr/local/lib/python3.8/dist-packages/datasets/utils/file_utils.py](https://localhost:8080/#) in is_remote_url(url_or_filename)\r\n 62 \r\n 63 def is_remote_url(url_or_filename: str) -> bool:\r\n---> 64 parsed = urlparse(url_or_filename)\r\n 65 return parsed.scheme in (\"http\", \"https\", \"s3\", \"gs\", \"hdfs\", \"ftp\")\r\n 66 \r\n\r\n[/usr/lib/python3.8/urllib/parse.py](https://localhost:8080/#) in urlparse(url, scheme, allow_fragments)\r\n 373 Note that we don't break the components up in smaller bits\r\n 374 (e.g. netloc is a single string) and we don't expand % escapes.\"\"\"\r\n--> 375 url, scheme, _coerce_result = _coerce_args(url, scheme)\r\n 376 splitresult = urlsplit(url, scheme, allow_fragments)\r\n 377 scheme, netloc, url, query, fragment = splitresult\r\n\r\n[/usr/lib/python3.8/urllib/parse.py](https://localhost:8080/#) in _coerce_args(*args)\r\n 125 if str_input:\r\n 126 return args + (_noop,)\r\n--> 127 return _decode_args(args) + (_encode_result,)\r\n 128 \r\n 129 # Result objects are more helpful than simple tuples\r\n\r\n[/usr/lib/python3.8/urllib/parse.py](https://localhost:8080/#) in _decode_args(args, encoding, errors)\r\n 109 def _decode_args(args, encoding=_implicit_encoding,\r\n 110 errors=_implicit_errors):\r\n--> 111 return tuple(x.decode(encoding, errors) if x else '' for x in args)\r\n 112 \r\n 113 def _coerce_args(*args):\r\n\r\n[/usr/lib/python3.8/urllib/parse.py](https://localhost:8080/#) in <genexpr>(.0)\r\n 109 def _decode_args(args, encoding=_implicit_encoding,\r\n 110 errors=_implicit_errors):\r\n--> 111 return tuple(x.decode(encoding, errors) if x else '' for x in args)\r\n 112 \r\n 113 def _coerce_args(*args):\r\n\r\nAttributeError: 'PosixPath' object has no attribute 'decode'\r\n```\r\n\r\n@mariosasko Should I create a new issue? ", "@mariosasko I would like to take this issue up. ", "@avinashsai Hi, I've assigned you the issue.\r\n\r\n@tranhd95 Yes, feel free to report this in a new issue." ]

[ "Just seen https://github.com/huggingface/datasets/issues/5281" ]

[ "Hi @J-shel, thanks for reporting.\r\n\r\nI think the issue comes from your call to `load_dataset`. As first argument, you should pass:\r\n- either the name of your dataset (\"mrf\") if this is already published on the Hub\r\n- or the path to the loading script of your dataset (\"path/to/your/local/mrf.py\").", "Hi, following your suggestion, I changed my call to load_dataset. Below is the latest:\r\nreader = load_dataset('data/mrf.py',\"default\", input_size=1024, split=split, streaming=True, keep_in_memory=None)\r\nHowever, I still got the same error.\r\nI have one question that is if I only define input_size=2048 in BUILDER_CONFIGS, may I specify input_size=1024 when loading the dataset? Cause I found that I could only specify name=\"default\" since I only define name=\"default\" in BUILDER_CONFIGS." ]

[ "Hi! `DatasetBuilder` is a parent class for concrete builders: `GeneratorBasedBuilder`, `ArrowBasedBuilder` and `BeamBasedBuilder`. When writing a builder script, these classes are the ones you should inherit from. And since all of them implement `_prepare_split`, you only have to implement the three methods mentioned above.", "Thanks so much @mariosasko for the fast response! I've been referencing [this page in the docs](https://huggingface.co/docs/datasets/v2.4.0/en/about_dataset_load) because it it pretty comprehensive in terms of what we have to do and I figured since we subclass the `BuilderConfig` the same pattern would hold, but I've also seen the page with those sub-classed builders as well, so that fills in a knowledge gap for me.", "cc @stevhliu who may have some ideas on how to improve this part of the docs.", "one more question for my understanding @mariosasko. the requirement of a loading script has always seemed counterintuitive to me. if i have to provide a script with every dataset, what is the point of using `datasets` if we're doing all the work of loading it, I can just do that in my code and skip the datasets integration (this of course discounts other potential benefits around metadata management, etc., my example is just simplest use case though for the sake of discussion).\r\n\r\nso i figured I would implement my own `BuilderConfig` and `DatasetBuilder` to handle that portion of it and not have to make a script. i _thought_ this would result in `datasets` (via `download_and_prepare`) then making me something that I could load using `load_dataset` moving forward.\r\n\r\nConcretely, i envisioned this pattern being possible:\r\n\r\n ```\r\nclass MyBuilderConfig(BuilderConfig):\r\n def __init__(self, name=\"my_named_dataset\", ...):\r\n super().__init__(name, ...)\r\n\r\nclass MyDatasetBuilder(GeneratorBasedBuilder):\r\n BUILDER_CONFIG_CLASS = MyBuilderConfig\r\n ....\r\n\r\nmy_builder = MyDatasetBuilder(...)\r\n\r\n# this doesn't exactly work like I thought; I don't get a dataset back, but NoneType instead\r\n# though I can see it loading the files and it generates the cache, etc.\r\nmy_dataset = my_builder.download_and_prepare()\r\n\r\n# load the dataset in the future by referencing it by name and loading from the cached arrow version\r\nnew_instance_of_my_dataset = load_dataset(\"my_named_dataset\")\r\n```\r\n\r\nI've seen references to the `save_to_disk` method which might be the next step I need in order to load it by name, in which case, that makes sense, then i just need to debug why `download_and_prepare` isn't returning me a dataset, but I feel like I still have a larger conceptual knowledge gap on how to use the library correctly.\r\n\r\nThanks again in advance!", "> the requirement of a loading script has always seemed counterintuitive to me\r\n\r\nThis is a requirement only for datasets not stored in standard formats such as CSV, JSON, SQL, Parquet, ImageFolder, etc. \r\n\r\n> if i have to provide a script with every dataset, what is the point of using datasets if we're doing all the work of loading it, I can just do that in my code and skip the datasets integration (this of course discounts other potential benefits around metadata management, etc., my example is just simplest use case though for the sake of discussion)\r\n\r\nOur README/documentation lists the main features... \r\n\r\nOne of the main ones is that our library makes it easy to work with datasets larger than RAM (thanks to Arrow and the caching mechanism), and this is not trivial to implement.\r\n\r\nRegarding the step-by-step builder, this is the pattern:\r\n```python\r\nfrom datasets import load_dataset_builder\r\nbuilder = load_dataset_builder(\"path/to/script\") # or direct instantiation with MyDatasetBuilder(...)\r\nbuilder.download_and_prepare()\r\ndset = builder.as_dataset()\r\n```", "ok, that makes sense. thank you @mariosasko. I realized i'd never looked on the hub at any of the files associated with any datasets. just did that now and it appears that i'll need to have a script regardless _but_ that will just contain my custom config and builder classes, so without realizing it I was already making my script, I just need to wrap that in a file that sits alongside my data (I looked at Glue and realized I was already doing what I thought didn't make sense to have to do, lol).\r\n\r\n`download_and_prepare` isn't returning me a dataset though, but I'll look into that and open another issue if I can't figure it out.", "`download_and_prepare` downloads and prepares the arrow files. You need to call `as_dataset` on the builder to get the dataset.", "ok, I think I was assigning the output of `builder.download_and_prepare` but it's an inplace op, so that explains the `NoneType` i was getting back. Now I'm getting:\r\n\r\n```\r\nArrowInvalid Traceback (most recent call last)\r\n<ipython-input-7-3ed50fb87c70> in <module>\r\n----> 1 ds = dataset_builder.as_dataset()\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/builder.py in as_dataset(self, split, run_post_process, ignore_verifications, in_memory)\r\n 1020 \r\n 1021 # Create a dataset for each of the given splits\r\n-> 1022 datasets = map_nested(\r\n 1023 partial(\r\n 1024 self._build_single_dataset,\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/utils/py_utils.py in map_nested(function, data_struct, dict_only, map_list, map_tuple, map_numpy, num_proc, parallel_min_length, types, disable_tqdm, desc)\r\n 442 num_proc = 1\r\n 443 if num_proc <= 1 or len(iterable) < parallel_min_length:\r\n--> 444 mapped = [\r\n 445 _single_map_nested((function, obj, types, None, True, None))\r\n 446 for obj in logging.tqdm(iterable, disable=disable_tqdm, desc=desc)\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/utils/py_utils.py in <listcomp>(.0)\r\n 443 if num_proc <= 1 or len(iterable) < parallel_min_length:\r\n 444 mapped = [\r\n--> 445 _single_map_nested((function, obj, types, None, True, None))\r\n 446 for obj in logging.tqdm(iterable, disable=disable_tqdm, desc=desc)\r\n 447 ]\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/utils/py_utils.py in _single_map_nested(args)\r\n 344 # Singleton first to spare some computation\r\n 345 if not isinstance(data_struct, dict) and not isinstance(data_struct, types):\r\n--> 346 return function(data_struct)\r\n 347 \r\n 348 # Reduce logging to keep things readable in multiprocessing with tqdm\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/builder.py in _build_single_dataset(self, split, run_post_process, ignore_verifications, in_memory)\r\n 1051 \r\n 1052 # Build base dataset\r\n-> 1053 ds = self._as_dataset(\r\n 1054 split=split,\r\n 1055 in_memory=in_memory,\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/builder.py in _as_dataset(self, split, in_memory)\r\n 1120 \"\"\"\r\n 1121 cache_dir = self._fs._strip_protocol(self._output_dir)\r\n-> 1122 dataset_kwargs = ArrowReader(cache_dir, self.info).read(\r\n 1123 name=self.name,\r\n 1124 instructions=split,\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/arrow_reader.py in read(self, name, instructions, split_infos, in_memory)\r\n 236 msg = f'Instruction \"{instructions}\" corresponds to no data!'\r\n 237 raise ValueError(msg)\r\n--> 238 return self.read_files(files=files, original_instructions=instructions, in_memory=in_memory)\r\n 239 \r\n 240 def read_files(\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/arrow_reader.py in read_files(self, files, original_instructions, in_memory)\r\n 257 \"\"\"\r\n 258 # Prepend path to filename\r\n--> 259 pa_table = self._read_files(files, in_memory=in_memory)\r\n 260 # If original_instructions is not None, convert it to a human-readable NamedSplit\r\n 261 if original_instructions is not None:\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/arrow_reader.py in _read_files(self, files, in_memory)\r\n 192 f[\"filename\"] = os.path.join(self._path, f[\"filename\"])\r\n 193 for f_dict in files:\r\n--> 194 pa_table: Table = self._get_table_from_filename(f_dict, in_memory=in_memory)\r\n 195 pa_tables.append(pa_table)\r\n 196 pa_tables = [t for t in pa_tables if len(t) > 0]\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/arrow_reader.py in _get_table_from_filename(self, filename_skip_take, in_memory)\r\n 327 filename_skip_take[\"take\"] if \"take\" in filename_skip_take else None,\r\n 328 )\r\n--> 329 table = ArrowReader.read_table(filename, in_memory=in_memory)\r\n 330 if take == -1:\r\n 331 take = len(table) - skip\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/arrow_reader.py in read_table(filename, in_memory)\r\n 348 \"\"\"\r\n 349 table_cls = InMemoryTable if in_memory else MemoryMappedTable\r\n--> 350 return table_cls.from_file(filename)\r\n 351 \r\n 352 \r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/table.py in from_file(cls, filename, replays)\r\n 1034 @classmethod\r\n 1035 def from_file(cls, filename: str, replays=None):\r\n-> 1036 table = _memory_mapped_arrow_table_from_file(filename)\r\n 1037 table = cls._apply_replays(table, replays)\r\n 1038 return cls(table, filename, replays)\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/table.py in _memory_mapped_arrow_table_from_file(filename)\r\n 48 def _memory_mapped_arrow_table_from_file(filename: str) -> pa.Table:\r\n 49 memory_mapped_stream = pa.memory_map(filename)\r\n---> 50 opened_stream = pa.ipc.open_stream(memory_mapped_stream)\r\n 51 pa_table = opened_stream.read_all()\r\n 52 return pa_table\r\n\r\n/databricks/python/lib/python3.8/site-packages/pyarrow/ipc.py in open_stream(source)\r\n 152 reader : RecordBatchStreamReader\r\n 153 \"\"\"\r\n--> 154 return RecordBatchStreamReader(source)\r\n 155 \r\n 156 \r\n\r\n/databricks/python/lib/python3.8/site-packages/pyarrow/ipc.py in __init__(self, source)\r\n 43 \r\n 44 def __init__(self, source):\r\n---> 45 self._open(source)\r\n 46 \r\n 47 \r\n\r\n/databricks/python/lib/python3.8/site-packages/pyarrow/ipc.pxi in pyarrow.lib._RecordBatchStreamReader._open()\r\n\r\n/databricks/python/lib/python3.8/site-packages/pyarrow/error.pxi in pyarrow.lib.pyarrow_internal_check_status()\r\n\r\n/databricks/python/lib/python3.8/site-packages/pyarrow/error.pxi in pyarrow.lib.check_status()\r\n\r\nArrowInvalid: Tried reading schema message, was null or length 0\r\n```\r\n\r\n", "looks like my arrow files are all empty @mariosasko \r\n\r\n\r\n\r\n\r\ni also see the `incomplete_info.lock` file a level up too. seems like the data isn't being persisted to disk when I call `download_and_prepare`. is there something else i need to do before then, perhaps?", "quick update @mariosasko. i got it working! i had to downgrade to `datasets==2.4.0`. testing other versions now and will let you know the results.", "I've tested with every version of `datasets>2.4.0` and i get the same error with all of them." ]

[ "_The documentation is not available anymore as the PR was closed or merged._" ]

[ "_The documentation is not available anymore as the PR was closed or merged._" ]

[ "note, that `datasets` already did some of that umask fixing in the past and also at the hub - the recent work on the hub about the same: https://github.com/huggingface/huggingface_hub/pull/1220\r\n\r\nAlso I noticed that each file has a .json counterpart and the latter always has the correct perms:\r\n\r\n```\r\n-rw------- 1 uue59kq cnw 173M Dec 9 01:37 537596e64721e2ae3d98785b91d30fda0360c196a8224e29658ad629e7303a4d\r\n-rw-rw---- 1 uue59kq cnw 101 Dec 9 01:37 537596e64721e2ae3d98785b91d30fda0360c196a8224e29658ad629e7303a4d.json\r\n```\r\n\r\nso perhaps cheating is possible and syncing the perms between the 2 will do the trick." ]

[ "cc @polinaeterna", "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/datasets/pr_5347). All of your documentation changes will be reflected on that endpoint.", "Cool ! Feel free to add a comment in the code to explain that and we can merge :)", "I'm not sure if this is a good change since we plan to get rid of `torchaudio` in the next couple of months...", "What do you think @polinaeterna @patrickvonplaten ? Models are usually using float32 (e.g. Wev2vec2 in `transformers`) IIRC", "IMO we can safely assume that float32 is always good enough when using audio models in inference or training. Nevertheless there might be use cases for audio datasets in the future where float64 is needed. \r\n\r\n=> I would by default always cast to float32, but then possible allow the user to cast to float64 ", "> I'm not sure if this is a good change since we plan to get rid of torchaudio in the next couple of months...\r\n\r\n@mariosasko I agree but who knows how long we will have to wait until we are really able to do so (https://github.com/bastibe/libsndfile-binaries/pull/17 is a draft. so as @patrickvonplaten is okay with float32, I'd merge.\r\n\r\n\r\n", "@polinaeterna Can you comment on the linked PR to see why it's still a draft? Maybe we can help somehow to get this merged finally.\r\n\r\nI think it's weird to align `soundfile` with `torchaudio` when the latter is only used for MP3 (and prob for not much longer). " ]

[ "As the survey is finished, can we close this issue, @LysandreJik ?", "Yes! I'll post a public summary on the forums shortly." ]

[ "After some more investigation, this is due to [this line of code](https://github.com/huggingface/datasets/blob/main/src/datasets/features/audio.py#L279). The function `sf.read(file)` should be updated to `sf.read(file, dtype=\"float32\")`\r\n\r\nIndeed, the default value in soundfile is `float64` ([see here](https://pysoundfile.readthedocs.io/en/latest/#soundfile.read)). \r\n", "@qmeeus I agree, decoding of different audio formats should return the same dtypes indeed!\r\n\r\nBut note that here you are concatenating datasets with different sampling rates: 48000 for CommonVoice and 16000 for Voxpopuli. So you should cast them to the same sampling rate value before interleaving, for example:\r\n```\r\ncv = cv.cast_column(\"audio\", Audio(sampling_rate=16000))\r\n```\r\notherwise you would get the same error because features of the same column (\"audio\") are not the same.\r\n\r\nAlso, the error you get is unexpected. Could you please confirm that you use the latest main version of the `datasets`? We had an issue that could lead to an error like this after using `rename_column` method, but it was fixed in https://github.com/huggingface/datasets/pull/5287 " ]

[ "_The documentation is not available anymore as the PR was closed or merged._" ]

[]

[ "Hi @cbarond there's already an open issue at https://github.com/dair-ai/emotion_dataset/issues/5, as the data seems to be missing now, so check that issue instead 👍🏻 ", "Thanks @cbarond for reporting and @alvarobartt for pointing to the issue we opened in the author's repo.\r\n\r\nIndeed, this issue was first raised in the \"emotion\" dataset Community tab: https://huggingface.co/datasets/emotion/discussions/3\r\n\r\nI'm closing this issue and leave the issue above for the subsequent updates.\r\n\r\nDuplicate of: https://huggingface.co/datasets/emotion/discussions/3", "try using \"SetFit/emotion\" instead", "> try using \"SetFit/emotion\" instead\r\n\r\nI' replaced \"emotion\" with \"SetFit/Emotion\", but the code is getting stuck at\r\n\r\n`emotions = load_dataset(\"SetFit/emotion\")`\r\n\r\nI pause execution using the debugger, and it takes me to filelock.py:226\r\n\r\n`with self._thread_lock:`\r\n\r\nDo you know a way to get past this issue?", "thanks @honeyimholm - worked for me", "> try using \"SetFit/emotion\" instead\r\n\r\nIt really helps a lot, thank you!", "The dataset loading script has been fixed: https://huggingface.co/datasets/emotion/discussions/4" ]

[ "_The documentation is not available anymore as the PR was closed or merged._" ]

[ "_The documentation is not available anymore as the PR was closed or merged._" ]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/datasets/pr_5339). All of your documentation changes will be reflected on that endpoint.", "@lhoestq I think I need some serious help with the tests 😅...I started this locally but it got too time consuming.\n\nOne issue I remember running into is with lossless audio encoding/decoding. I started thinking of using the underlying Audio feature instead of PyAV so I didn't have to rewrite similar logic here...but assumed that would turn into a mess w/ underlying logic" ]

[ "Hi !\r\n\r\n> It starts using all the cores (I am not sure why because I did not pass num_proc)\r\n\r\nThe tokenizer uses Rust code that is multithreaded. And maybe the `feature_extractor` might run some things in parallel as well - but I'm not super familiar with its internals.\r\n\r\n> then progress bar stops at every 1k steps. (starts using a single core)\r\n\r\nEvery 1000 examples we flush the processed examples to disk. It is this way because Arrow is a columnar format: you must write data chunk by chunk. The processing in on hold while writing right now - maybe this can be improved in the future.", "Hi @lhoestq \r\nThanks for the explanation! it was so helpful! Let me check why `feature_extractor` is running on multiple cpus." ]

[ "I like the idea of having `webdataset` as an optional dependency to ensure our loader generates web datasets the same way as the main project.", "Webdataset is the one of the most popular dataset formats for large scale computer vision tasks. Upvote for this issue. " ]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/datasets/pr_5336). All of your documentation changes will be reflected on that endpoint.", "Hi @mariosasko, @lhoestq I was wondering whether we should include `batched` as a `pytest.mark` param for the functions testing `IterableDataset.map` so as to ensure that the changes done in this PR work fine without breaking anything of the actual functionality.\r\n\r\nI've pushed updated tests just for one of the unit testing functions to be run as `pytest tests/test_iterable_dataset.py::test_mapped_examples_iterable -s --durations 0`, but some are still missing `batched` param, it was just to ask you whether we're supposed to do this for the rest of the functions or not, if it's a yes I'll push the commit as it's ready, but didn't want to push extra stuff that may be discarded later!\r\n\r\nThanks :hugs:", "Thanks for the feedback @lhoestq, I agree with keeping `Optional` instead of `Union[type, None]` for now 👍🏻" ]

[ "_The documentation is not available anymore as the PR was closed or merged._", "I think the only place where we need to add it is here https://github.com/huggingface/hub-docs/blob/main/js/src/lib/interfaces/Types.ts\r\n\r\nAnd I think we can remove tasks.json completely from this repo", "Isn't tasks.json used anymore in this repo?", "> I think the only place where we need to add it is here https://github.com/huggingface/hub-docs/blob/main/js/src/lib/interfaces/Types.ts\r\n> \r\n> And I think we can remove tasks.json completely from this repo\r\n\r\nWhat about the warning I mentioned in https://github.com/huggingface/datasets/issues/5255#issuecomment-1339013527? Also, the depth estimation entry is already present in https://github.com/huggingface/hub-docs/blob/main/js/src/lib/interfaces/Types.ts. ", "The update is based on what I received in the output of the export job (c.f. https://github.com/huggingface/datasets/issues/5255#issuecomment-1339107195). \r\n\r\nEdit: Oh, are you referring to the dataset card of NYU Depth V2?", "Yes, my suggestion was for the dataset card: you got the error message because you tried to set `depth-estimation` in `class_ids` instead of `class_categories`.", "> What about the warning I mentioned in https://github.com/huggingface/datasets/issues/5255#issuecomment-1339013527? Also, the depth estimation entry is already present in https://github.com/huggingface/hub-docs/blob/main/js/src/lib/interfaces/Types.ts.\r\n\r\nif you place it in `task_categories` you should be good :)", "yes i would suggest rm'ing tasks.json here for clarity", "Closing it. ", "It's not clear if we can remove it btw, since old versions of `evaluate` rely on it (see https://github.com/huggingface/evaluate/pull/309)\r\n\r\ncc @lvwerra ", "Actually it can be removed without incidence in old versions of evaluate since we kept an hardcoded `known_task_ids` that is marked \"DEPRECATED\"" ]

[ "_The documentation is not available anymore as the PR was closed or merged._", "Thanks ! Let us know if we can help :)\r\n\r\nSmall pref for having multiple PRs", "Awesome, thanks! Sorry this one is a little big, I'll open some smaller ones next :)" ]

[ "_The documentation is not available anymore as the PR was closed or merged._" ]

[ "Should `datasets` allow `ClassLabel` input parameter to be an `np.array` even though internally we need to cast it to a Python list? @lhoestq @mariosasko ", "Hi! No, I don't think so. The `names` parameter is [annotated](https://github.com/huggingface/datasets/blob/582236640b9109988e5f7a16a8353696ffa09a16/src/datasets/features/features.py#L892) as `List[str]` (**NumPy arrays are not lists**), and considering that type checking is not a common practice in Python, I think we can leave the code as-is.", "I appreciate it is the wrong type, and that type checking is not common, but I think there's a few circumstances that make it a good idea from a usability perspective.\r\n\r\nIt's quite a difficult error to debug because it comes from a utility function (so it's not immediately obvious which parameter caused it). What makes it even more difficult is the exception happens when the features instance is used to instantiate the dataset, **not** when when the wrong type is actually passed when the features is instantiated. When I was debugging the error, I didn't really consider it could be an issue with the features instance because it had instantiated fine. It's also not one of the more common exceptions caused by trying to use a non-list as a list.\r\n\r\nIt's also relatively easy to accidentally get a numpy array of class types (e.g. calling `unique()` on a pandas dataframe column). Additionally, passing in a `set` instead of the list (again, relatively easy because people may run `set(classes)` to generate uniques) causes an error when the features instance is used, albeit a slightly more obvious one.\r\n\r\nThe names list is already being processed and validated in the `__post_init__` method anyway, so it would not really be adding any complexity to check it is actually a list here too. I'm happy to contribute this change if you change your mind about whether it's worthwhile.", "I agree that it's not easy to debug this issue, so perhaps we could add some basic type checking (e.g. `not isinstance(names, list)` -> error) to make debugging easier. Feel free to submit a PR.\r\n\r\n> Additionally, passing in a set instead of the list (again, relatively easy because people may run set(classes) to generate uniques) causes an error when the features instance is used, albeit a slightly more obvious one.\r\n\r\n`set` is an unordered structure (it's ordered in Python 3.6+, but this is CPython's implementation detail), and the order of ClassLabel `names` matters, so this doesn't require a fix.", "What about checking for `Sequence` instead? I think users can pass a list or a tuple as well." ]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/datasets/pr_5331). All of your documentation changes will be reflected on that endpoint." ]

[ "_The documentation is not available anymore as the PR was closed or merged._", "I think it's also reasonable to add the same note to the AudioFolder decription", "Thank you ! I think \"regular\" is more appropriate than \"small\". It can easily scale to a few thousands of images - just not millions x)", "Replaced \"small\" with \"several thousand\" since what is considered \"regular\" and even \"small\" can be kind of vague!" ]

[ "_The documentation is not available anymore as the PR was closed or merged._", "EDIT\r\nAt least the docs for ~~main~~ PR branch are now built:\r\n- https://github.com/huggingface/datasets/actions/runs/3594847760/jobs/6053620813", "Build documentation for main branch was triggered after this PR being merged: https://github.com/huggingface/datasets/actions/runs/3603370082/jobs/6071482470" ]

[ "The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/datasets/pr_5327). All of your documentation changes will be reflected on that endpoint." ]

[]

[ "Hi! I agree it makes sense for `IterableDataset.map` to support the `batch_size=None` case. This should be super easy to fix.", "@mariosasko as this is something simple maybe I can include it as part of https://github.com/huggingface/datasets/pull/5311? Let me know :+1:", "#self-assign", "Feel free to close this @lhoestq as part of https://github.com/huggingface/datasets/pull/5336 :hugs:", "Thanks again :)\r\n\r\n> For practical usages, an alternative to this would be to convert from an iterable dataset to a map-style dataset, but it is not obvious how to do this.\r\n\r\nThis is interesting as well, if anyone wants to explore" ]

[ "I agree we have a mess with docstrings...", "Ok, I believe we've cleaned up most of the old syntax we were using for the user-facing docs! There are still a couple of `:obj:`'s and `:class:` floating around in the docstrings we don't expose that I'll track down :)" ]

[ "Thanks for reporting, @liaeh.\r\n\r\nWe are having a look at it. ", "I have transferred the discussion to the Community tab of the dataset: https://huggingface.co/datasets/taskmaster2/discussions/1" ]

[ "_The documentation is not available anymore as the PR was closed or merged._" ]

[ "_The documentation is not available anymore as the PR was closed or merged._", "> Do you know why this stopped working?\r\n\r\nIt comes from the changes in https://github.com/huggingface/datasets/pull/5107/files#diff-355ae5c229f95f86895404b72378ecd6e966c41cbeebb674af6fe6e9611bc126" ]

[ "_The documentation is not available anymore as the PR was closed or merged._" ]

[ "_The documentation is not available anymore as the PR was closed or merged._" ]

[ "_The documentation is not available anymore as the PR was closed or merged._", "please review :) @lhoestq @ola13 thankoo", "Thanks :) I just updated the test to make sure it works even when there's a column missing, and did a minor change to json.py to add the missing columns for the other kinds of JSON files as well (I moved the code to`self._cast_table`)", "Thanks Unso! If @lhoestq is happy then I'm also happy :D", "When I noticed the ping, this PR had already been merged...\r\n\r\nLuckily, PyArrow's `read_json` behaves the same when `explicit_schema` is given via `ParseOptions`, so I'm okay with this change (our JSON loader doesn't use `read_json` for decoding JSON in some scenarios, so this manual approach is the right one).\r\n" ]