Skip to content

Commit

Permalink
Fix dummy data
Browse files Browse the repository at this point in the history
  • Loading branch information
@albertvillanova
albertvillanova committed Nov 30, 2021
1 parent f7ba594 commit b1d63ca
Showing 1 changed file with 0 additions and 0 deletions.
Binary file modified datasets/the_pile/dummy/all/0.0.0/dummy_data.zip
View file
Binary file not shown.

1 comment on commit b1d63ca

@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==3.0.0

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) 0.086874 / 0.011353 (0.075521) 0.004987 / 0.011008 (-0.006021) 0.042657 / 0.038508 (0.004149) 0.045102 / 0.023109 (0.021993) 0.398918 / 0.275898 (0.123020) 0.469944 / 0.323480 (0.146464) 0.093020 / 0.007986 (0.085035) 0.005550 / 0.004328 (0.001221) 0.011770 / 0.004250 (0.007519) 0.048057 / 0.037052 (0.011004) 0.424228 / 0.258489 (0.165739) 0.463757 / 0.293841 (0.169916) 0.113427 / 0.128546 (-0.015119) 0.014243 / 0.075646 (-0.061404) 0.347336 / 0.419271 (-0.071935) 0.064774 / 0.043533 (0.021241) 0.411532 / 0.255139 (0.156393) 0.430252 / 0.283200 (0.147053) 0.107962 / 0.141683 (-0.033721) 2.307518 / 1.452155 (0.855364) 2.377545 / 1.492716 (0.884829)

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) 0.309785 / 0.018006 (0.291778) 0.568776 / 0.000490 (0.568286) 0.013036 / 0.000200 (0.012836) 0.000140 / 0.000054 (0.000085)

Benchmark: benchmark_indices_mapping.json

metric select shard shuffle sort train_test_split
new / old (diff) 0.046962 / 0.037411 (0.009551) 0.027647 / 0.014526 (0.013121) 0.033849 / 0.176557 (-0.142707) 0.276037 / 0.737135 (-0.461099) 0.036003 / 0.296338 (-0.260335)

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) 0.669424 / 0.215209 (0.454215) 6.622835 / 2.077655 (4.545180) 2.542813 / 1.504120 (1.038693) 2.164679 / 1.541195 (0.623484) 2.166673 / 1.468490 (0.698183) 0.755895 / 4.584777 (-3.828882) 7.218648 / 3.745712 (3.472936) 3.343295 / 5.269862 (-1.926567) 1.529890 / 4.565676 (-3.035787) 0.086326 / 0.424275 (-0.337949) 0.014308 / 0.007607 (0.006701) 0.818876 / 0.226044 (0.592832) 8.130424 / 2.268929 (5.861496) 3.239341 / 55.444624 (-52.205284) 2.505084 / 6.876477 (-4.371392) 2.709017 / 2.142072 (0.566944) 0.908058 / 4.805227 (-3.897170) 0.190030 / 6.500664 (-6.310634) 0.071573 / 0.075469 (-0.003896)

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) 2.195364 / 1.841788 (0.353576) 16.788987 / 8.074308 (8.714679) 46.123243 / 10.191392 (35.931851) 1.127997 / 0.680424 (0.447574) 0.777358 / 0.534201 (0.243157) 0.536192 / 0.579283 (-0.043091) 0.790257 / 0.434364 (0.355893) 0.372847 / 0.540337 (-0.167490) 0.408559 / 1.386936 (-0.978377)
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) 0.082015 / 0.011353 (0.070662) 0.004852 / 0.011008 (-0.006156) 0.039177 / 0.038508 (0.000669) 0.042789 / 0.023109 (0.019680) 0.401935 / 0.275898 (0.126037) 0.445513 / 0.323480 (0.122033) 0.128328 / 0.007986 (0.120342) 0.005229 / 0.004328 (0.000901) 0.009663 / 0.004250 (0.005412) 0.059081 / 0.037052 (0.022029) 0.407465 / 0.258489 (0.148976) 0.459012 / 0.293841 (0.165171) 0.105461 / 0.128546 (-0.023085) 0.019033 / 0.075646 (-0.056614) 0.347684 / 0.419271 (-0.071588) 0.059760 / 0.043533 (0.016227) 0.418317 / 0.255139 (0.163178) 0.458597 / 0.283200 (0.175398) 0.096608 / 0.141683 (-0.045074) 2.261676 / 1.452155 (0.809521) 2.278688 / 1.492716 (0.785971)

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) 0.312440 / 0.018006 (0.294434) 0.588639 / 0.000490 (0.588149) 0.002989 / 0.000200 (0.002789) 0.000105 / 0.000054 (0.000051)

Benchmark: benchmark_indices_mapping.json

metric select shard shuffle sort train_test_split
new / old (diff) 0.039869 / 0.037411 (0.002457) 0.032680 / 0.014526 (0.018154) 0.034218 / 0.176557 (-0.142338) 0.239582 / 0.737135 (-0.497553) 0.038908 / 0.296338 (-0.257430)

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) 0.659104 / 0.215209 (0.443895) 6.561209 / 2.077655 (4.483554) 2.497962 / 1.504120 (0.993842) 2.187141 / 1.541195 (0.645946) 2.264264 / 1.468490 (0.795774) 0.721351 / 4.584777 (-3.863426) 7.271484 / 3.745712 (3.525772) 5.153908 / 5.269862 (-0.115954) 1.481887 / 4.565676 (-3.083790) 0.082717 / 0.424275 (-0.341558) 0.015745 / 0.007607 (0.008138) 0.841488 / 0.226044 (0.615444) 8.308669 / 2.268929 (6.039741) 3.301208 / 55.444624 (-52.143416) 2.613905 / 6.876477 (-4.262572) 2.675918 / 2.142072 (0.533846) 0.893327 / 4.805227 (-3.911900) 0.178913 / 6.500664 (-6.321751) 0.082516 / 0.075469 (0.007046)

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) 2.050032 / 1.841788 (0.208244) 15.480580 / 8.074308 (7.406272) 45.691921 / 10.191392 (35.500529) 0.935999 / 0.680424 (0.255575) 0.748489 / 0.534201 (0.214288) 0.486147 / 0.579283 (-0.093136) 0.770389 / 0.434364 (0.336025) 0.343788 / 0.540337 (-0.196550) 0.348109 / 1.386936 (-1.038827)

CML watermark

Please sign in to comment.