Skip to content

Commit

Permalink
update CI to use pyarrow 3.0.0
Browse files Browse the repository at this point in the history
  • Loading branch information
@lhoestq
lhoestq committed Jun 23, 2021
1 parent 17301d8 commit ebeb9c1
Showing 1 changed file with 6 additions and 6 deletions.
12 changes: 6 additions & 6 deletions .circleci/config.yml
View file
Original file line number Diff line number Diff line change
Expand Up @@ -18,7 +18,7 @@ jobs:
- run: pip install pyarrow --upgrade
- run: HF_SCRIPTS_VERSION=master python -m pytest -sv ./tests/

run_dataset_script_tests_pyarrow_1:
run_dataset_script_tests_pyarrow_3:
working_directory: ~/datasets
docker:
- image: circleci/python:3.6
Expand All @@ -29,7 +29,7 @@ jobs:
- run: source venv/bin/activate
- run: pip install .[tests]
- run: pip install -r additional-tests-requirements.txt --no-deps
- run: pip install pyarrow==1.0.0
- run: pip install pyarrow==3.0.0
- run: HF_SCRIPTS_VERSION=master python -m pytest -sv ./tests/

run_dataset_script_tests_pyarrow_latest_WIN:
Expand All @@ -50,7 +50,7 @@ jobs:
- run: $env:HF_SCRIPTS_VERSION="master"
- run: python -m pytest -sv ./tests/

run_dataset_script_tests_pyarrow_1_WIN:
run_dataset_script_tests_pyarrow_3_WIN:
working_directory: ~/datasets
executor:
name: win/default
Expand All @@ -64,7 +64,7 @@ jobs:
- run: "& venv/Scripts/activate.ps1"
- run: pip install .[tests]
- run: pip install -r additional-tests-requirements.txt --no-deps
- run: pip install pyarrow==1.0.0
- run: pip install pyarrow==3.0.0
- run: $env:HF_SCRIPTS_VERSION="master"
- run: python -m pytest -sv ./tests/

Expand Down Expand Up @@ -116,8 +116,8 @@ workflows:
jobs:
- check_code_quality
- run_dataset_script_tests_pyarrow_latest
- run_dataset_script_tests_pyarrow_1
- run_dataset_script_tests_pyarrow_3
- run_dataset_script_tests_pyarrow_latest_WIN
- run_dataset_script_tests_pyarrow_1_WIN
- run_dataset_script_tests_pyarrow_3_WIN
- build_doc
- deploy_doc: *workflow_filters

1 comment on commit ebeb9c1

@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==1.0.0

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) 0.020041 / 0.011353 (0.008688) 0.012439 / 0.011008 (0.001430) 0.040389 / 0.038508 (0.001881) 0.042272 / 0.023109 (0.019163) 0.291320 / 0.275898 (0.015422) 0.300063 / 0.323480 (-0.023416) 0.009243 / 0.007986 (0.001257) 0.004599 / 0.004328 (0.000270) 0.009717 / 0.004250 (0.005466) 0.055131 / 0.037052 (0.018078) 0.292139 / 0.258489 (0.033650) 0.301447 / 0.293841 (0.007606) 0.134934 / 0.128546 (0.006388) 0.097549 / 0.075646 (0.021903) 0.357886 / 0.419271 (-0.061386) 1.783182 / 0.043533 (1.739650) 0.282171 / 0.255139 (0.027032) 0.341104 / 0.283200 (0.057905) 0.084977 / 0.141683 (-0.056706) 1.419730 / 1.452155 (-0.032425) 1.650818 / 1.492716 (0.158102)

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) 0.011603 / 0.018006 (-0.006404) 0.544171 / 0.000490 (0.543681) 0.002986 / 0.000200 (0.002786) 0.000070 / 0.000054 (0.000015)

Benchmark: benchmark_indices_mapping.json

metric select shard shuffle sort train_test_split
new / old (diff) 0.039279 / 0.037411 (0.001867) 0.024287 / 0.014526 (0.009761) 0.024949 / 0.176557 (-0.151607) 0.042757 / 0.737135 (-0.694378) 0.026395 / 0.296338 (-0.269944)

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) 0.327611 / 0.215209 (0.112402) 3.323778 / 2.077655 (1.246123) 1.715628 / 1.504120 (0.211508) 1.560511 / 1.541195 (0.019316) 1.570029 / 1.468490 (0.101539) 5.232217 / 4.584777 (0.647440) 4.443875 / 3.745712 (0.698163) 6.792419 / 5.269862 (1.522557) 5.555997 / 4.565676 (0.990320) 0.477780 / 0.424275 (0.053505) 0.008401 / 0.007607 (0.000794) 0.425640 / 0.226044 (0.199595) 4.173108 / 2.268929 (1.904180) 2.045341 / 55.444624 (-53.399283) 1.788058 / 6.876477 (-5.088419) 1.796086 / 2.142072 (-0.345986) 5.300534 / 4.805227 (0.495307) 4.483774 / 6.500664 (-2.016890) 8.077808 / 0.075469 (8.002338)

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) 9.836074 / 1.841788 (7.994286) 12.129513 / 8.074308 (4.055204) 23.749692 / 10.191392 (13.558300) 0.737721 / 0.680424 (0.057298) 0.463129 / 0.534201 (-0.071072) 0.658128 / 0.579283 (0.078845) 0.500773 / 0.434364 (0.066409) 0.575284 / 0.540337 (0.034947) 1.336931 / 1.386936 (-0.050005)
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) 0.020061 / 0.011353 (0.008709) 0.013610 / 0.011008 (0.002602) 0.041993 / 0.038508 (0.003485) 0.124277 / 0.023109 (0.101168) 0.294466 / 0.275898 (0.018568) 0.316530 / 0.323480 (-0.006950) 0.010499 / 0.007986 (0.002513) 0.004936 / 0.004328 (0.000607) 0.010579 / 0.004250 (0.006329) 0.052997 / 0.037052 (0.015945) 0.289738 / 0.258489 (0.031249) 0.328498 / 0.293841 (0.034657) 0.133964 / 0.128546 (0.005418) 0.100340 / 0.075646 (0.024694) 0.360412 / 0.419271 (-0.058860) 1.709139 / 0.043533 (1.665606) 0.284759 / 0.255139 (0.029620) 0.307899 / 0.283200 (0.024700) 0.082100 / 0.141683 (-0.059583) 1.622745 / 1.452155 (0.170590) 1.643398 / 1.492716 (0.150681)

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) 0.080490 / 0.018006 (0.062484) 0.526424 / 0.000490 (0.525935) 0.037197 / 0.000200 (0.036997) 0.000379 / 0.000054 (0.000324)

Benchmark: benchmark_indices_mapping.json

metric select shard shuffle sort train_test_split
new / old (diff) 0.036393 / 0.037411 (-0.001018) 0.023649 / 0.014526 (0.009123) 0.024575 / 0.176557 (-0.151981) 0.042025 / 0.737135 (-0.695110) 0.025778 / 0.296338 (-0.270560)

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) 0.347273 / 0.215209 (0.132064) 3.459770 / 2.077655 (1.382115) 1.744083 / 1.504120 (0.239963) 1.541186 / 1.541195 (-0.000009) 1.574775 / 1.468490 (0.106285) 5.201991 / 4.584777 (0.617214) 4.426882 / 3.745712 (0.681170) 6.788922 / 5.269862 (1.519061) 5.378240 / 4.565676 (0.812564) 0.472780 / 0.424275 (0.048505) 0.009063 / 0.007607 (0.001456) 0.401774 / 0.226044 (0.175730) 4.012736 / 2.268929 (1.743807) 1.909793 / 55.444624 (-53.534832) 1.598791 / 6.876477 (-5.277686) 1.629085 / 2.142072 (-0.512988) 4.880297 / 4.805227 (0.075070) 4.929781 / 6.500664 (-1.570883) 8.461287 / 0.075469 (8.385818)

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) 9.981333 / 1.841788 (8.139546) 11.719334 / 8.074308 (3.645026) 26.053776 / 10.191392 (15.862384) 0.788176 / 0.680424 (0.107753) 0.496255 / 0.534201 (-0.037946) 0.657141 / 0.579283 (0.077858) 0.483033 / 0.434364 (0.048669) 0.564105 / 0.540337 (0.023768) 1.330070 / 1.386936 (-0.056866)

CML watermark

Please sign in to comment.