Skip to content

Commit

Permalink
Address requested changes
Browse files Browse the repository at this point in the history
  • Loading branch information
@albertvillanova
albertvillanova committed Mar 31, 2022
1 parent 85f5a12 commit 37a3657
Showing 1 changed file with 3 additions and 2 deletions.
5 changes: 3 additions & 2 deletions src/datasets/load.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -580,8 +580,9 @@ def __init__(
):
self.name = name
self.revision = revision
self.download_config = download_config or DownloadConfig()
self.download_config.max_retries = 3
self.download_config = download_config.copy() or DownloadConfig()
if self.download_config.max_retries < 3:
self.download_config.max_retries = 3
self.download_mode = download_mode
self.dynamic_modules_path = dynamic_modules_path
assert self.name.count("/") == 0
Expand Down

1 comment on commit 37a3657

@github-actions
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Show benchmarks

PyArrow==5.0.0

Show updated benchmarks!

Benchmark: benchmark_array_xd.json

metric read_batch_formatted_as_numpy after write_array2d read_batch_formatted_as_numpy after write_flattened_sequence read_batch_formatted_as_numpy after write_nested_sequence read_batch_unformated after write_array2d read_batch_unformated after write_flattened_sequence read_batch_unformated after write_nested_sequence read_col_formatted_as_numpy after write_array2d read_col_formatted_as_numpy after write_flattened_sequence read_col_formatted_as_numpy after write_nested_sequence read_col_unformated after write_array2d read_col_unformated after write_flattened_sequence read_col_unformated after write_nested_sequence read_formatted_as_numpy after write_array2d read_formatted_as_numpy after write_flattened_sequence read_formatted_as_numpy after write_nested_sequence read_unformated after write_array2d read_unformated after write_flattened_sequence read_unformated after write_nested_sequence write_array2d write_flattened_sequence write_nested_sequence
new / old (diff) 0.011348 / 0.011353 (-0.000004) 0.004632 / 0.011008 (-0.006376) 0.038366 / 0.038508 (-0.000142) 0.041134 / 0.023109 (0.018025) 0.352386 / 0.275898 (0.076488) 0.384806 / 0.323480 (0.061326) 0.009467 / 0.007986 (0.001481) 0.004080 / 0.004328 (-0.000248) 0.010910 / 0.004250 (0.006660) 0.045242 / 0.037052 (0.008190) 0.332173 / 0.258489 (0.073684) 0.373145 / 0.293841 (0.079304) 0.034493 / 0.128546 (-0.094053) 0.011572 / 0.075646 (-0.064074) 0.309327 / 0.419271 (-0.109944) 0.060293 / 0.043533 (0.016760) 0.346220 / 0.255139 (0.091081) 0.362272 / 0.283200 (0.079073) 0.128058 / 0.141683 (-0.013625) 2.158522 / 1.452155 (0.706367) 2.118612 / 1.492716 (0.625896)

Benchmark: benchmark_getitem_100B.json

metric get_batch_of_1024_random_rows get_batch_of_1024_rows get_first_row get_last_row
new / old (diff) 0.333431 / 0.018006 (0.315425) 0.487152 / 0.000490 (0.486662) 0.021932 / 0.000200 (0.021732) 0.000730 / 0.000054 (0.000675)

Benchmark: benchmark_indices_mapping.json

metric select shard shuffle sort train_test_split
new / old (diff) 0.031464 / 0.037411 (-0.005947) 0.119497 / 0.014526 (0.104971) 0.129131 / 0.176557 (-0.047426) 0.175163 / 0.737135 (-0.561973) 0.131059 / 0.296338 (-0.165280)

Benchmark: benchmark_iterating.json

metric read 5000 read 50000 read_batch 50000 10 read_batch 50000 100 read_batch 50000 1000 read_formatted numpy 5000 read_formatted pandas 5000 read_formatted tensorflow 5000 read_formatted torch 5000 read_formatted_batch numpy 5000 10 read_formatted_batch numpy 5000 1000 shuffled read 5000 shuffled read 50000 shuffled read_batch 50000 10 shuffled read_batch 50000 100 shuffled read_batch 50000 1000 shuffled read_formatted numpy 5000 shuffled read_formatted_batch numpy 5000 10 shuffled read_formatted_batch numpy 5000 1000
new / old (diff) 0.462215 / 0.215209 (0.247006) 4.657541 / 2.077655 (2.579886) 2.029421 / 1.504120 (0.525301) 1.793390 / 1.541195 (0.252195) 1.862522 / 1.468490 (0.394032) 0.497535 / 4.584777 (-4.087242) 5.555005 / 3.745712 (1.809292) 2.575531 / 5.269862 (-2.694330) 1.051738 / 4.565676 (-3.513938) 0.060895 / 0.424275 (-0.363380) 0.014618 / 0.007607 (0.007011) 0.584927 / 0.226044 (0.358883) 5.754555 / 2.268929 (3.485627) 2.451483 / 55.444624 (-52.993142) 2.063797 / 6.876477 (-4.812680) 2.168949 / 2.142072 (0.026877) 0.609722 / 4.805227 (-4.195506) 0.137463 / 6.500664 (-6.363201) 0.069444 / 0.075469 (-0.006025)

Benchmark: benchmark_map_filter.json

metric filter map fast-tokenizer batched map identity map identity batched map no-op batched map no-op batched numpy map no-op batched pandas map no-op batched pytorch map no-op batched tensorflow
new / old (diff) 1.936067 / 1.841788 (0.094280) 16.479834 / 8.074308 (8.405526) 29.378570 / 10.191392 (19.187178) 0.980521 / 0.680424 (0.300097) 0.599212 / 0.534201 (0.065011) 0.550693 / 0.579283 (-0.028590) 0.599472 / 0.434364 (0.165108) 0.370284 / 0.540337 (-0.170053) 0.365211 / 1.386936 (-1.021725)
PyArrow==latest
Show updated benchmarks!

Benchmark: benchmark_array_xd.json

metric read_batch_formatted_as_numpy after write_array2d read_batch_formatted_as_numpy after write_flattened_sequence read_batch_formatted_as_numpy after write_nested_sequence read_batch_unformated after write_array2d read_batch_unformated after write_flattened_sequence read_batch_unformated after write_nested_sequence read_col_formatted_as_numpy after write_array2d read_col_formatted_as_numpy after write_flattened_sequence read_col_formatted_as_numpy after write_nested_sequence read_col_unformated after write_array2d read_col_unformated after write_flattened_sequence read_col_unformated after write_nested_sequence read_formatted_as_numpy after write_array2d read_formatted_as_numpy after write_flattened_sequence read_formatted_as_numpy after write_nested_sequence read_unformated after write_array2d read_unformated after write_flattened_sequence read_unformated after write_nested_sequence write_array2d write_flattened_sequence write_nested_sequence
new / old (diff) 0.009855 / 0.011353 (-0.001497) 0.004378 / 0.011008 (-0.006630) 0.035814 / 0.038508 (-0.002694) 0.040195 / 0.023109 (0.017086) 0.372426 / 0.275898 (0.096528) 0.435192 / 0.323480 (0.111712) 0.007096 / 0.007986 (-0.000889) 0.005449 / 0.004328 (0.001121) 0.008045 / 0.004250 (0.003794) 0.045923 / 0.037052 (0.008870) 0.327090 / 0.258489 (0.068601) 0.385308 / 0.293841 (0.091467) 0.039534 / 0.128546 (-0.089012) 0.012042 / 0.075646 (-0.063604) 0.300803 / 0.419271 (-0.118469) 0.061368 / 0.043533 (0.017835) 0.351846 / 0.255139 (0.096707) 0.395907 / 0.283200 (0.112707) 0.112526 / 0.141683 (-0.029156) 2.174275 / 1.452155 (0.722120) 2.206038 / 1.492716 (0.713322)

Benchmark: benchmark_getitem_100B.json

metric get_batch_of_1024_random_rows get_batch_of_1024_rows get_first_row get_last_row
new / old (diff) 0.306697 / 0.018006 (0.288691) 0.476022 / 0.000490 (0.475533) 0.012360 / 0.000200 (0.012160) 0.000328 / 0.000054 (0.000274)

Benchmark: benchmark_indices_mapping.json

metric select shard shuffle sort train_test_split
new / old (diff) 0.028743 / 0.037411 (-0.008669) 0.119005 / 0.014526 (0.104480) 0.133045 / 0.176557 (-0.043511) 0.189008 / 0.737135 (-0.548127) 0.129923 / 0.296338 (-0.166415)

Benchmark: benchmark_iterating.json

metric read 5000 read 50000 read_batch 50000 10 read_batch 50000 100 read_batch 50000 1000 read_formatted numpy 5000 read_formatted pandas 5000 read_formatted tensorflow 5000 read_formatted torch 5000 read_formatted_batch numpy 5000 10 read_formatted_batch numpy 5000 1000 shuffled read 5000 shuffled read 50000 shuffled read_batch 50000 10 shuffled read_batch 50000 100 shuffled read_batch 50000 1000 shuffled read_formatted numpy 5000 shuffled read_formatted_batch numpy 5000 10 shuffled read_formatted_batch numpy 5000 1000
new / old (diff) 0.449210 / 0.215209 (0.234001) 4.679998 / 2.077655 (2.602344) 2.076047 / 1.504120 (0.571927) 1.861903 / 1.541195 (0.320708) 1.967866 / 1.468490 (0.499376) 0.497833 / 4.584777 (-4.086944) 5.569052 / 3.745712 (1.823340) 4.274917 / 5.269862 (-0.994945) 1.076204 / 4.565676 (-3.489473) 0.059060 / 0.424275 (-0.365216) 0.013271 / 0.007607 (0.005664) 0.570666 / 0.226044 (0.344621) 5.858795 / 2.268929 (3.589866) 2.603997 / 55.444624 (-52.840627) 2.186464 / 6.876477 (-4.690013) 2.284682 / 2.142072 (0.142610) 0.638461 / 4.805227 (-4.166766) 0.144454 / 6.500664 (-6.356210) 0.069649 / 0.075469 (-0.005820)

Benchmark: benchmark_map_filter.json

metric filter map fast-tokenizer batched map identity map identity batched map no-op batched map no-op batched numpy map no-op batched pandas map no-op batched pytorch map no-op batched tensorflow
new / old (diff) 1.852217 / 1.841788 (0.010429) 15.975579 / 8.074308 (7.901271) 29.841165 / 10.191392 (19.649773) 0.957867 / 0.680424 (0.277443) 0.595943 / 0.534201 (0.061742) 0.547112 / 0.579283 (-0.032172) 0.606491 / 0.434364 (0.172127) 0.359188 / 0.540337 (-0.181149) 0.380400 / 1.386936 (-1.006536)

CML watermark

Please sign in to comment.