Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[AIR] Added Ray Logging to MosaicTrainer #29620

Merged
merged 82 commits into from
Oct 27, 2022
Merged

[AIR] Added Ray Logging to MosaicTrainer #29620

merged 82 commits into from
Oct 27, 2022

Commits on Oct 11, 2022

  1. add initial trainer API files

    @ilee300a
    ilee300a committed Oct 11, 2022
    Configuration menu
    Copy the full SHA
    364d7f6 View commit details
    Browse the repository at this point in the history

  2. created a basic composer trainer wrapper and a test 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 11, 2022
    Configuration menu
    Copy the full SHA
    cfa593a View commit details
    Browse the repository at this point in the history

  3. Merge branch 'ray-project:master' into init_mosaic_trainer_api

    @ilee300a
    ilee300a authored Oct 11, 2022
    Configuration menu
    Copy the full SHA
    919aea4 View commit details
    Browse the repository at this point in the history

  4. formatted 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 11, 2022
    Configuration menu
    Copy the full SHA
    7f3b551 View commit details
    Browse the repository at this point in the history

Commits on Oct 12, 2022

  1. removed unused config key

    @ilee300a
    ilee300a committed Oct 12, 2022
    Configuration menu
    Copy the full SHA
    0871fa4 View commit details
    Browse the repository at this point in the history

  2. Merge branch 'init_mosaic_trainer_api' of https://github.com/ilee300a… 

    …/ray into init_mosaic_trainer_api

Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 12, 2022
    Configuration menu
    Copy the full SHA
    f083e1b View commit details
    Browse the repository at this point in the history

  3. merged

    @ilee300a
    ilee300a committed Oct 12, 2022
    Configuration menu
    Copy the full SHA
    536a381 View commit details
    Browse the repository at this point in the history

  4. Merge branch 'init_mosaic_trainer_api' of https://github.com/ilee300a… 

    …/ray into init_mosaic_trainer_api
    @ilee300a
    ilee300a committed Oct 12, 2022
    Configuration menu
    Copy the full SHA
    9d15416 View commit details
    Browse the repository at this point in the history

  5. addressed comments 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 12, 2022
    Configuration menu
    Copy the full SHA
    7b509da View commit details
    Browse the repository at this point in the history

  6. addressed comments 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 12, 2022
    Configuration menu
    Copy the full SHA
    2712fda View commit details
    Browse the repository at this point in the history

Commits on Oct 13, 2022

  1. Merge branch 'ray-project:master' into init_mosaic_trainer_api

    @ilee300a
    ilee300a authored Oct 13, 2022
    Configuration menu
    Copy the full SHA
    1851322 View commit details
    Browse the repository at this point in the history

  2. change the data directory 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 13, 2022
    Configuration menu
    Copy the full SHA
    ae1b9a6 View commit details
    Browse the repository at this point in the history

  3. removed progress bar by default 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 13, 2022
    Configuration menu
    Copy the full SHA
    5d599e4 View commit details
    Browse the repository at this point in the history

  4. move dataset preparation inside the trainer init function 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 13, 2022
    Configuration menu
    Copy the full SHA
    5da865d View commit details
    Browse the repository at this point in the history

  5. add mosaic trainer to the api ref 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 13, 2022
    Configuration menu
    Copy the full SHA
    0e42dec View commit details
    Browse the repository at this point in the history

  6. add fit config 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 13, 2022
    Configuration menu
    Copy the full SHA
    104075f View commit details
    Browse the repository at this point in the history

  7. remove datasets in the trainer worker loop 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 13, 2022
    Configuration menu
    Copy the full SHA
    161009e View commit details
    Browse the repository at this point in the history

  8. added a test for fit config 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 13, 2022
    Configuration menu
    Copy the full SHA
    72dc74f View commit details
    Browse the repository at this point in the history

  9. added example MosaicTrainer script and a test 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 13, 2022
    Configuration menu
    Copy the full SHA
    faa2e9b View commit details
    Browse the repository at this point in the history

  10. added mosaic trainer test to the build script 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 13, 2022
    Configuration menu
    Copy the full SHA
    810428c View commit details
    Browse the repository at this point in the history

  11. removed unnecessary line 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 13, 2022
    Configuration menu
    Copy the full SHA
    3d32969 View commit details
    Browse the repository at this point in the history

  12. reformatted passing in config and use regex for error message testing 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 13, 2022
    Configuration menu
    Copy the full SHA
    b96f384 View commit details
    Browse the repository at this point in the history

  13. Fixed typo 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 13, 2022
    Configuration menu
    Copy the full SHA
    1fe7989 View commit details
    Browse the repository at this point in the history

Commits on Oct 17, 2022

  1. remove composer loggers in training callback 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 17, 2022
    Configuration menu
    Copy the full SHA
    d39b63d View commit details
    Browse the repository at this point in the history

  2. added a test for logger removal 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 17, 2022
    Configuration menu
    Copy the full SHA
    de77a64 View commit details
    Browse the repository at this point in the history

  3. addressed comments and removed fit_config 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 17, 2022
    Configuration menu
    Copy the full SHA
    b297926 View commit details
    Browse the repository at this point in the history

  4. fixed a typo 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 17, 2022
    Configuration menu
    Copy the full SHA
    00bd1f8 View commit details
    Browse the repository at this point in the history

  5. added further description for trainier_init_config 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 17, 2022
    Configuration menu
    Copy the full SHA
    be4c311 View commit details
    Browse the repository at this point in the history

  6. fixed typo 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 17, 2022
    Configuration menu
    Copy the full SHA
    a93ddf9 View commit details
    Browse the repository at this point in the history

  7. fixed mosaic test_example 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 17, 2022
    Configuration menu
    Copy the full SHA
    8b324af View commit details
    Browse the repository at this point in the history

Commits on Oct 18, 2022

  1. addressed comments 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 18, 2022
    Configuration menu
    Copy the full SHA
    65bf8bb View commit details
    Browse the repository at this point in the history

  2. added composer to tune requirements 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 18, 2022
    Configuration menu
    Copy the full SHA
    2b57038 View commit details
    Browse the repository at this point in the history

  3. fixed library import 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 18, 2022
    Configuration menu
    Copy the full SHA
    5c96046 View commit details
    Browse the repository at this point in the history

  4. updated torch and torchvision versions for composer compatibility 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 18, 2022
    Configuration menu
    Copy the full SHA
    7177134 View commit details
    Browse the repository at this point in the history

  5. removed mosaic from requirements and install it separately in test file 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 18, 2022
    Configuration menu
    Copy the full SHA
    3617626 View commit details
    Browse the repository at this point in the history

  6. added mosaicml library to the doc requirements 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 18, 2022
    Configuration menu
    Copy the full SHA
    1563395 View commit details
    Browse the repository at this point in the history

  7. add mosaicml in custom directives instead 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 18, 2022
    Configuration menu
    Copy the full SHA
    9fb684e View commit details
    Browse the repository at this point in the history

  8. changed mock module name 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 18, 2022
    Configuration menu
    Copy the full SHA
    d27c20b View commit details
    Browse the repository at this point in the history

  9. added more composer libraries to custom directives 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 18, 2022
    Configuration menu
    Copy the full SHA
    e1e92e4 View commit details
    Browse the repository at this point in the history

  10. reformatted 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 18, 2022
    Configuration menu
    Copy the full SHA
    8c64463 View commit details
    Browse the repository at this point in the history

Commits on Oct 19, 2022

  1. updated for tests 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 19, 2022
    Configuration menu
    Copy the full SHA
    4fec09e View commit details
    Browse the repository at this point in the history

  2. added imports to the docstring example 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 19, 2022
    Configuration menu
    Copy the full SHA
    652c12f View commit details
    Browse the repository at this point in the history

  3. fixed indentation 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 19, 2022
    Configuration menu
    Copy the full SHA
    31914cf View commit details
    Browse the repository at this point in the history

Commits on Oct 20, 2022

  1. moved imports inside the functions 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 20, 2022
    Configuration menu
    Copy the full SHA
    706d971 View commit details
    Browse the repository at this point in the history

  2. added doctest skip and moved MosaicTrainer import in tests 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 20, 2022
    Configuration menu
    Copy the full SHA
    797b4df View commit details
    Browse the repository at this point in the history

  3. Merge branch 'ray-project:master' into init_mosaic_trainer_api

    @ilee300a
    ilee300a authored Oct 20, 2022
    Configuration menu
    Copy the full SHA
    715c37f View commit details
    Browse the repository at this point in the history

  4. added more doc test skips 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 20, 2022
    Configuration menu
    Copy the full SHA
    175aeaa View commit details
    Browse the repository at this point in the history

  5. move composer import 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 20, 2022
    Configuration menu
    Copy the full SHA
    82305c4 View commit details
    Browse the repository at this point in the history

Commits on Oct 21, 2022

  1. moved mosaic test example imports 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 21, 2022
    Configuration menu
    Copy the full SHA
    883d2e5 View commit details
    Browse the repository at this point in the history

  2. added mosaicml to requirements_train 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 21, 2022
    Configuration menu
    Copy the full SHA
    8c3306c View commit details
    Browse the repository at this point in the history

  3. Merge branch 'ray-project:master' into init_mosaic_trainer_api

    @ilee300a
    ilee300a authored Oct 21, 2022
    Configuration menu
    Copy the full SHA
    69f60f9 View commit details
    Browse the repository at this point in the history

  4. updated torch and torchvision versions 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 21, 2022
    Configuration menu
    Copy the full SHA
    f168f6c View commit details
    Browse the repository at this point in the history

  5. Merge branch 'init_mosaic_trainer_api' of https://github.com/ilee300a… 

    …/ray into init_mosaic_trainer_api
    @ilee300a
    ilee300a committed Oct 21, 2022
    Configuration menu
    Copy the full SHA
    9e3d119 View commit details
    Browse the repository at this point in the history

  6. updated ml docker to match torch version 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 21, 2022
    Configuration menu
    Copy the full SHA
    858973b View commit details
    Browse the repository at this point in the history

  7. update 

    Signed-off-by: Amog Kamsetty <[email protected]>
    @amogkam
    amogkam committed Oct 21, 2022
    Configuration menu
    Copy the full SHA
    4654631 View commit details
    Browse the repository at this point in the history

  8. update 

    Signed-off-by: Amog Kamsetty <[email protected]>
    @amogkam
    amogkam committed Oct 21, 2022
    Configuration menu
    Copy the full SHA
    48b48bb View commit details
    Browse the repository at this point in the history

  9. Update .buildkite/pipeline.ml.yml 

    Signed-off-by: Amog Kamsetty <[email protected]>
    @amogkam
    amogkam authored Oct 21, 2022
    Configuration menu
    Copy the full SHA
    20dc4f8 View commit details
    Browse the repository at this point in the history

Commits on Oct 22, 2022

  1. Update .buildkite/pipeline.ml.yml 

    Signed-off-by: Amog Kamsetty <[email protected]>
    @amogkam
    amogkam authored Oct 22, 2022
    Configuration menu
    Copy the full SHA
    e401eff View commit details
    Browse the repository at this point in the history

Commits on Oct 24, 2022

  1. Update python/requirements/ml/requirements_dl.txt 

    Signed-off-by: Amog Kamsetty <[email protected]>
    @amogkam
    amogkam authored Oct 24, 2022
    Configuration menu
    Copy the full SHA
    eb0c97d View commit details
    Browse the repository at this point in the history

  2. add conftest 

    Signed-off-by: Amog Kamsetty <[email protected]>
    @amogkam
    amogkam committed Oct 24, 2022
    Configuration menu
    Copy the full SHA
    928b529 View commit details
    Browse the repository at this point in the history

  3. added ray logging 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 24, 2022
    Configuration menu
    Copy the full SHA
    3529b77 View commit details
    Browse the repository at this point in the history

  4. Merge branch 'init_mosaic_trainer_api' into ray_logging

    @ilee300a
    ilee300a committed Oct 24, 2022
    Configuration menu
    Copy the full SHA
    5e956cc View commit details
    Browse the repository at this point in the history

  5. remove loggers after trainer instantiation as well 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 24, 2022
    Configuration menu
    Copy the full SHA
    7ddee9b View commit details
    Browse the repository at this point in the history

  6. refactor 

    Signed-off-by: Amog Kamsetty <[email protected]>
    @amogkam
    amogkam committed Oct 24, 2022
    Configuration menu
    Copy the full SHA
    9d1a302 View commit details
    Browse the repository at this point in the history

Commits on Oct 25, 2022

  1. suppress console to log 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 25, 2022
    Configuration menu
    Copy the full SHA
    1a2c4b1 View commit details
    Browse the repository at this point in the history

  2. check the epoch and accuracy data for test example 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 25, 2022
    Configuration menu
    Copy the full SHA
    7fcb7b0 View commit details
    Browse the repository at this point in the history

  3. Merge branch 'init_mosaic_trainer_api' into ray_logging

    @ilee300a
    ilee300a committed Oct 25, 2022
    Configuration menu
    Copy the full SHA
    9b3d9e7 View commit details
    Browse the repository at this point in the history

  4. updated fixtures 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 25, 2022
    Configuration menu
    Copy the full SHA
    4cf8fa0 View commit details
    Browse the repository at this point in the history

  5. Merge branch 'master' into init_mosaic_trainer_api

    @ilee300a
    ilee300a committed Oct 25, 2022
    Configuration menu
    Copy the full SHA
    cbc1c6e View commit details
    Browse the repository at this point in the history

  6. Merge branch 'init_mosaic_trainer_api' into ray_logging

    @ilee300a
    ilee300a committed Oct 25, 2022
    Configuration menu
    Copy the full SHA
    1fc42f4 View commit details
    Browse the repository at this point in the history

  7. added custom directives for mosaic util library 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 25, 2022
    Configuration menu
    Copy the full SHA
    acbcc2a View commit details
    Browse the repository at this point in the history

  8. updated removing composer loggers 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 25, 2022
    Configuration menu
    Copy the full SHA
    0c79025 View commit details
    Browse the repository at this point in the history

  9. updated -- add ray logger after trainer init without using loggers key 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 25, 2022
    Configuration menu
    Copy the full SHA
    3d84dcb View commit details
    Browse the repository at this point in the history

Commits on Oct 27, 2022

  1. fixed typo 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 27, 2022
    Configuration menu
    Copy the full SHA
    2ce2b41 View commit details
    Browse the repository at this point in the history

  2. updated ray logging interval 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 27, 2022
    Configuration menu
    Copy the full SHA
    1993c0f View commit details
    Browse the repository at this point in the history

  3. updated ray logging tests with suggested changes 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 27, 2022
    Configuration menu
    Copy the full SHA
    3f15a5a View commit details
    Browse the repository at this point in the history

  4. made test dataset size scaled by predefined batch size 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 27, 2022
    Configuration menu
    Copy the full SHA
    36393e9 View commit details
    Browse the repository at this point in the history

  5. updated logging interval 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 27, 2022
    Configuration menu
    Copy the full SHA
    09acf07 View commit details
    Browse the repository at this point in the history

  6. Update python/ray/train/mosaic/_mosaic_utils.py 

    Co-authored-by: Amog Kamsetty <[email protected]>
Signed-off-by: ilee300a <[email protected]>
    @ilee300a @amogkam
    ilee300a and amogkam authored Oct 27, 2022
    Configuration menu
    Copy the full SHA
    870b8c5 View commit details
    Browse the repository at this point in the history

  7. updated logging -- flush data and added comments 

    Signed-off-by: ilee300a <[email protected]>
    @ilee300a
    ilee300a committed Oct 27, 2022
    Configuration menu
    Copy the full SHA
    18392ff View commit details
    Browse the repository at this point in the history

  8. Merge branch 'ray_logging' of https://github.com/ilee300a/ray into ra… 

    …y_logging
    @ilee300a
    ilee300a committed Oct 27, 2022
    Configuration menu
    Copy the full SHA
    875dacc View commit details
    Browse the repository at this point in the history

  9. Merge branch 'ray-project:master' into ray_logging

    @ilee300a
    ilee300a authored Oct 27, 2022
    Configuration menu
    Copy the full SHA
    ac07acc View commit details
    Browse the repository at this point in the history