search

Yukiya-Umimi / ITUNet-for-PICAI-2022-Challenge

Apache License 2.0
16 stars 5 forks source link

SageMaker pipeline #10

Closed joeranbosma closed 1 year ago

joeranbosma commented 1 year ago

Dear @Yukiya-Umimi,

We have been training your models, but unfortunately ran into some issues. We would like to propose the following changes to your codebase:

Additionally, I would like to propose some minor changes:

I've put all of these proposed changes in this PR. Could you go through the changes, and if your agree, merge them into main?

Thanks!

Yukiya-Umimi commented 1 year ago

Hi,

It is OK, I have great confidence in your work!

Best, Honey K

yukiya

@. | ---- Replied Message ---- | From | Joeran @.> | | Date | 3/26/2023 20:22 | | To | @.> | | Cc | @.> , @.***> | | Subject | [Yukiya-Umimi/ITUNet-for-PICAI-2022-Challenge] SageMaker pipeline (PR #10) |

Dear @Yukiya-Umimi,

We have been training your models, but unfortunately ran into some issues. We would like to propose the following changes to your codebase:

Bugfix mha2nnunet settings to resample to 3.0x0.5x0.5 and crop to [24, 384, 384]: 443f703 Make repository pip-installable for easier importing of functions: 446ebc6 Restructure Docker container definition: a479e03 Add training code for segmentation model: 29d8416

Additionally, I would like to propose some minor changes:

Cleaned up Python cache files Created init.py files for making the repo pip-installable Change imports, e.g. from utils import ... to from segmentation.utils import ... for making the repo pip-installable Fix version of scikit-image to 0.20.0 to ensure skimage.metrics exists Remove unused imports

I've put all of these proposed changes in this PR. Could you go through the changes, and if your agree, merge them into main?

Thanks!

You can view, comment on, or merge this pull request online at:

https://github.com/Yukiya-Umimi/ITUNet-for-PICAI-2022-Challenge/pull/10

Commit Summary 8b436b9 Cleanup 4449bac Bugfix 7503fda Safer with file permissions 1780970 Cleanup 29d8416 Supervised segmentation training script bbede2f Cleanup imports df290cb Cleanup imports 9c905a5 Docker container for segmentation training 65e3468 Update scikit-image 3f82db4 Cleanup imports 8ff71f1 Add missing dependency 4c37241 Cleanup imports ed52843 Cleanup imports 1bc1ab3 Bugfix 443f703 Bugfix preprocessing resampling a878c2b Move Docker container definitions b88a08d Add preprocessing Docker container 4845190 Temporarily use this branch cea67e5 Cleanup a073b20 Cleanup & init files 446ebc6 Make repository pip-installable a479e03 Cleanup Docker containers & bugfix File Changes

(34 files)

R.dockerignore (0) ADockerfile (18) Abuild.sh (4) Aclassification/init.py (0) Dclassification/pycache/config.cpython-37.pyc (0) Dclassification/pycache/trainer.cpython-37.pyc (0) Dclassification/pycache/utils.cpython-37.pyc (0) Apreprocess/init.py (0) Mpreprocess/preprocess.py (9) Apyproject.toml (22) Arequirements.txt (16) Asegmentation/init.py (0) Dsegmentation/pycache/config.cpython-37.pyc (0) Dsegmentation/pycache/data_loader.cpython-37.pyc (0) Dsegmentation/pycache/loss.cpython-37.pyc (0) Dsegmentation/pycache/metrics.cpython-37.pyc (0) Dsegmentation/pycache/model.cpython-37.pyc (0) Dsegmentation/pycache/trainer.cpython-37.pyc (0) Dsegmentation/pycache/utils.cpython-37.pyc (0) Msegmentation/config.py (5) Msegmentation/data_loader.py (9) Msegmentation/predict_2d.py (15) Msegmentation/requirements.txt (3) Msegmentation/run.py (11) Msegmentation/trainer.py (33) Msegmentation/utils.py (8) Asetup.cfg (54) Asetup.py (18) Dsubmit_result/pycache/old_unet.cpython-37.pyc (0) Msubmit_result/process.py (20) Mtrain_1_classification.py (8) Mtrain_2_segmentation.py (56) Dtraining_dockers/1_classification/Dockerfile (23) Dtraining_dockers/1_classification/build.sh (4) Patch Links: https://github.com/Yukiya-Umimi/ITUNet-for-PICAI-2022-Challenge/pull/10.patch https://github.com/Yukiya-Umimi/ITUNet-for-PICAI-2022-Challenge/pull/10.diff

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: @.***>