microsoft/seismic-deeplearning issues and pull requests

#406 - Trying to get in touch regarding a security issue

Issue - State: closed - Opened by zidingz about 3 years ago - 3 comments

#405 - SEG 2020 and other idle fixes

Pull Request - State: open - Opened by maxkazmsft about 4 years ago

#404 - fixed tensorflow version mismatch caused by conda

Pull Request - State: closed - Opened by maxkazmsft about 4 years ago

#403 - fixed tensorflow version mismatch caused by conda

Pull Request - State: closed - Opened by maxkazmsft about 4 years ago

#402 - fix for tensorflow on staging branch

Pull Request - State: closed - Opened by maxkazmsft about 4 years ago - 1 comment

#401 - fixed ternsorflow version due to bug in tensorflow version 2.1.0

Pull Request - State: closed - Opened by maxkazmsft about 4 years ago

#400 - 0.2.1 release

Pull Request - State: closed - Opened by maxkazmsft about 4 years ago

#399 - finalized performance metrics

Pull Request - State: closed - Opened by maxkazmsft about 4 years ago

#398 - Docker README corrections and pretrained model checking

Pull Request - State: closed - Opened by maxkazmsft about 4 years ago

#397 - tensorboard notebook fix & loading of pre-trained models fix

Pull Request - State: closed - Opened by maxkazmsft about 4 years ago

#396 - resolved conflicts for 0.2 release

Pull Request - State: closed - Opened by maxkazmsft about 4 years ago - 1 comment

#395 - 0.2 release

Pull Request - State: closed - Opened by maxkazmsft about 4 years ago - 1 comment

#394 - fix test.py and notebook issues

Pull Request - State: closed - Opened by yalaudah about 4 years ago - 1 comment

#393 - debug test.py scoring of trained models - facilitates correct results when scoring trained models

Issue - State: closed - Opened by maxkazmsft about 4 years ago
Labels: Type: Bug, Prior: ShowStopper

#392 - updated documentation to close 354 and 381

Pull Request - State: closed - Opened by maxkazmsft about 4 years ago

#391 - added missing license headers and fixed formatting

Pull Request - State: closed - Opened by maxkazmsft about 4 years ago

#390 - BYOD Penobscot

Pull Request - State: closed - Opened by maxkazmsft about 4 years ago

#389 - stylistic change after debugging 385

Pull Request - State: closed - Opened by maxkazmsft about 4 years ago - 1 comment

#388 - docker build test

Pull Request - State: closed - Opened by fazamani about 4 years ago

#387 - changed tensorflow pinned version

Pull Request - State: closed - Opened by fazamani about 4 years ago

#386 - Fixing #259 by adding symmetric padding along depth direction

Pull Request - State: closed - Opened by yalaudah about 4 years ago

#385 - debug slow runtime in multi GPU training - figure out why multi-GPU training does not produce much of a speedup

Issue - State: closed - Opened by maxkazmsft about 4 years ago
Labels: Type: Enhancement, multi-GPU

#384 - check license headers again - facilitates compliance to legal terms

Issue - State: closed - Opened by maxkazmsft about 4 years ago
Labels: Type: Documentation, Prior: High, Investment: Low

#383 - determine the best configuration parameters for multi-GPU training - facilitates better multi-GPU results

Issue - State: closed - Opened by maxkazmsft about 4 years ago
Labels: Prior: High, multi-GPU, Investment: Low

#382 - fix seeds in train.py and test.py to make results perfectly reproducible on a single GPU - enhances the robustness of results

Issue - State: open - Opened by maxkazmsft about 4 years ago
Labels: Type: Enhancement, Investment: Medium, Investment: High

#381 - README documentation fixes - facilitates clear understanding of the content of the repo

Issue - State: closed - Opened by maxkazmsft about 4 years ago
Labels: Type: Documentation, Prior: High, Investment: Low

#380 - PR to fix #371 and #372

Pull Request - State: closed - Opened by fazamani about 4 years ago

#379 - fixed multi-GPU termination in train.py

Pull Request - State: closed - Opened by maxkazmsft about 4 years ago

#378 - enable multi-GPU training with Docker image - facilitates multi-GPU training on any Linux OS

Issue - State: closed - Opened by maxkazmsft about 4 years ago - 1 comment
Labels: Type: Feature, Prior: High, multi-GPU, Investment: Medium

#377 - add a clean-up job to remove output directories from build VM

Issue - State: open - Opened by fazamani over 4 years ago
Labels: Prior: Low, Type: Enhancement

#376 - cache Penbscot and Dutch F3 data on build VM and download weekly

Issue - State: open - Opened by fazamani over 4 years ago
Labels: Prior: Medium, Type: Enhancement

#375 - Data flow tests

Pull Request - State: closed - Opened by fazamani over 4 years ago

#374 - operator1 and operator2 close-out

Issue - State: open - Opened by maxkazmsft over 4 years ago
Labels: Prior: High

#373 - bug bash - facilitates gathering feedback from stakeholders

Issue - State: open - Opened by maxkazmsft over 4 years ago
Labels: Prior: High

#372 - investigate the reproducibility problem with patch_deconvnet for small training data set

Issue - State: closed - Opened by fazamani over 4 years ago
Labels: Type: Bug, Prior: High, Status: Done, Investment: Low

#371 - add the learning rate to the tensorboard

Issue - State: closed - Opened by fazamani over 4 years ago
Labels: Prior: High, Type: Enhancement, Status: Done, Investment: Low

#370 - Add data QC module

Issue - State: open - Opened by fazamani over 4 years ago
Labels: Type: Feature, Prior: High, Investment: Low

#369 - Dynamic global data information, reduces user-required specifications

Issue - State: open - Opened by fazamani over 4 years ago
Labels: Type: Feature, Prior: Low, Investment: Low

#368 - make tests simpler

Pull Request - State: closed - Opened by fazamani over 4 years ago

#367 - add test for multi-GPU train.py run - facilitates bug-free train.py for multi-GPU training

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Type: Feature, Prior: High, Investment: Low

#366 - Azure ML training pipeline improvement comments - facilitates better user onboarding for the pipeline

Issue - State: open - Opened by maxkazmsft over 4 years ago
Labels: Prior: High

#365 - provide pre-trained SEResNet model with high accuracy - facilitates faster onboarding with pre-trained SEResNet model

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Type: Bug, Prior: ShowStopper

#364 - add pre-trained models to test section of the HRNet notebooks - facilitates faster onboarding with more pre-trained models

Issue - State: open - Opened by maxkazmsft over 4 years ago
Labels: Prior: Low, Type: Enhancement

#363 - add UNet pre-trained model - facilitates faster onboarding with UNet model

Issue - State: open - Opened by maxkazmsft over 4 years ago
Labels: Prior: Low, Type: Enhancement

#362 - Replace HRNet with SEResNet model in the notebook

Pull Request - State: closed - Opened by maxkazmsft over 4 years ago - 1 comment

#361 - make pre-run notebooks for segyconverter utility and SEResNet available in main README - facilitates faster quick-start experience

Issue - State: open - Opened by maxkazmsft over 4 years ago
Labels: Prior: High, Type: Gap, Type: Enhancement

#360 - enable Azure ML pipeline from CSE team - adds AML training pipeline to codebase (models can be trained on AML)

Issue - State: open - Opened by maxkazmsft over 4 years ago
Labels: Type: Feature, Prior: High

#359 - Multi-GPU training support

Pull Request - State: closed - Opened by yalaudah over 4 years ago
Labels: Type: Feature, multi-GPU

#358 - Add docker tests - ensures docker image doesn't break while new features are added

Issue - State: closed - Opened by yalaudah over 4 years ago - 1 comment
Labels: Prior: High, Type: Gap, Status: Done, multi-GPU

#357 - merging work from CSE team into main staging branch

Pull Request - State: closed - Opened by maxkazmsft over 4 years ago - 1 comment

#356 - V00.01.00003 release

Pull Request - State: closed - Opened by maxkazmsft over 4 years ago - 1 comment

#355 - Remove depth channel from Tensorboard visualization - makes grayscale images appear as such in TensorBoard logging

Issue - State: open - Opened by yalaudah over 4 years ago
Labels: Type: Feature, Prior: High

#354 - update documentation with SEResNet and data requirements - facilitates better user onboarding experience

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Type: Documentation, Prior: High

#353 - follow up with legal regarding temporary hosting of datasets for builds - facilitates faster setup job runtime

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Prior: High, Type: Enhancement

#352 - extend config object in train and test and pass it to data loaders - facilitates the use the config variables by lower level functions

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Prior: High, Type: Enhancement, data-loaders

#351 - run Penobscot dataset through tensor format - facilitates uniform data loader and converter sample code

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Prior: High, data-loaders

#350 - change HRNet notebook to use SEResNet - removes possibility of model malfunction in the notebook

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Type: Bug, Prior: High

#349 - move contributions from CSE team into main folders - facilitates more user friendly repository with code samples

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Type: Feature, Prior: High

#348 - add GPU profile information to multi-GPU train and test scripts - facilitates ability to view possible code bottlenecks

Issue - State: open - Opened by maxkazmsft over 4 years ago
Labels: Type: Feature, Prior: High

#347 - PR to fix #342

Pull Request - State: closed - Opened by fazamani over 4 years ago

#346 - make tests simpler (reduce test data size) - facilitates faster builds for PRs

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Prior: High, Type: Enhancement

#345 - Minor fix to address #326

Pull Request - State: closed - Opened by yalaudah over 4 years ago

#344 - Images from the binary dataset result in multiple classes after augmentation

Issue - State: closed - Opened by yalaudah over 4 years ago - 1 comment
Labels: Type: Wontfix

#343 - replace hrnet with seresnet in experiments - provides stable default model

Pull Request - State: closed - Opened by yalaudah over 4 years ago - 2 comments

#342 - image normalization should be wrt the volume image - facilitates correctly displayed image colormap range

Issue - State: closed - Opened by fazamani over 4 years ago - 2 comments
Labels: Type: Bug, Prior: High, Status: In-progress, Type: Correctness

#340 - fix 333

Pull Request - State: closed - Opened by fazamani over 4 years ago

#339 - fixes 318

Pull Request - State: closed - Opened by maxkazmsft over 4 years ago

#338 - Fixing #324 and #325

Pull Request - State: closed - Opened by yalaudah over 4 years ago

#333 - gradient synthetic dataset does not produce gradient-like images in the train loader - facilitates data correctness

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Type: Bug, Prior: High, Status: In-progress, Type: Correctness

#332 - update READMEs

Issue - State: closed - Opened by fazamani over 4 years ago
Labels: Type: Documentation, Prior: Medium

#330 - debug test.py patch padding before model scoring - facilitates scoring correctness and higher scoring accuracy

Issue - State: open - Opened by maxkazmsft over 4 years ago
Labels: Type: Bug, Prior: Medium, Type: Accuracy

#329 - create a backlog of correctness items to fix in the codebase - facilitates correct training pipeline

Issue - State: closed - Opened by maxkazmsft over 4 years ago

#328 - debug checkerboard synthetic dataset having empty padded images as input - facilitates higher training accuracy and training correctness

Issue - State: closed - Opened by maxkazmsft over 4 years ago - 1 comment
Labels: Type: Bug, Prior: High, Type: Accuracy

#326 - debug multiple classes showing up on binary synthetic data - facilitates correct mask visualizations representative of the number of classes

Issue - State: closed - Opened by yalaudah over 4 years ago - 1 comment
Labels: Type: Bug, Prior: ShowStopper, Status: In-progress, Type: Correctness

#325 - Model predictions should not be normalized before saving to disk, leads to incorrect results

Issue - State: closed - Opened by yalaudah over 4 years ago
Labels: Type: Bug, Prior: ShowStopper, Type: Correctness

#324 - The colormap used in mask_to_disk does not support more than 12 classes, will leads to incorrect results for datasets with more than 12 classes

Issue - State: closed - Opened by yalaudah over 4 years ago
Labels: Type: Bug, Prior: High, Type: Enhancement

#323 - set clean staging history - facilitates easy release merges of staging into master

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Prior: High, Type: Correctness

#321 - bring back Penobscot dataset and treat it as BYOD - adds another dataset which we can quote results on

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Type: Feature, Prior: High

#320 - add multi-GPU training to train.py Dutch F3 local - facilitates faster training times

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Prior: High, multi-GPU

#319 - add unit test to check that each input slice has uniform distribution - facilitates uniform sampling of training data for better model accuracy

Issue - State: open - Opened by maxkazmsft over 4 years ago - 1 comment
Labels: Prior: High, Type: Accuracy

#318 - add deconvnet and unet correctness train, val and test metrics as a unit test - facilitates certainty in results

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Type: Bug, Prior: High, Type: Correctness

#317 - add correctness data and class size unit test metrics - facilitates certainty in correct data flow

Issue - State: closed - Opened by maxkazmsft over 4 years ago - 1 comment
Labels: Prior: High, Type: Enhancement

#316 - add code coverage badge to github front page - facilitates visual certainty in unit test code coverage

Issue - State: open - Opened by maxkazmsft over 4 years ago
Labels: Type: Feature, Prior: High

#315 - run master and provide feedback on state of the repo - makes repo friendly to new users

Issue - State: closed - Opened by maxkazmsft over 4 years ago - 1 comment
Labels: Prior: Medium, Type: Correctness

#314 - merge staging into master and cut a 0.1.3 release - keeps master as latest stable branch

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Type: Feature, Prior: High

#313 - add v0.1.3 release notes - facilitates clean release tracking

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Type: Documentation, Prior: High

#304 - debug test.py returns zero output - determine why deconvnet and unet model snapshots produce zero output (same class predictions)

Issue - State: closed - Opened by yalaudah over 4 years ago - 1 comment
Labels: Type: Bug, Prior: ShowStopper, Type: Correctness

#303 - Add serialization/deserialization unit tests - checks whether writing model snapshots to disc actually works properly

Issue - State: open - Opened by yalaudah over 4 years ago
Labels: Prior: Medium

#298 - Update main README for V0.2 release - facilitates correct understanding

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Type: Documentation, Prior: High

#286 - Generated data splits should be tracked along with model outputs, and not stored with data.

Issue - State: open - Opened by yalaudah over 4 years ago
Labels: Type: Enhancement

#276 - make output transform unpad and scale down mask in penobscot metrics - facilitates correctly reported inline mIoU

Issue - State: open - Opened by maxkazmsft over 4 years ago
Labels: Prior: Medium, Type: Enhancement, Type: Accuracy

#259 - prepare_dutchf3.py vertical sample locations drops patches at the bottom of the volume - leads to worse results in that region

Issue - State: closed - Opened by yalaudah over 4 years ago
Labels: Prior: High, Type: Accuracy

#255 - create dev/user Docker version - enables repo to work in platform-agnostic way

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Type: Feature, Prior: ShowStopper

#225 - add Azure ML training pipeline notebook - facilitates the user to get quickly up to speed with AML training pipelines

Issue - State: open - Opened by maxkazmsft over 4 years ago
Labels: Prior: High, Type: Enhancement

#223 - merge contrib into staging to make a release - release with contributions from CSE team

Issue - State: closed - Opened by maxkazmsft over 4 years ago
Labels: Prior: High, Type: Enhancement

#158 - generate whole 3D volumes on test sets - sets uniform data comparison across multiple datasets

Issue - State: open - Opened by maxkazmsft over 4 years ago
Labels: Type: Feature, Prior: High

#157 - create a data loader class with works with generic segy files - facilitates better user experience of using custom datasets with the repo

Issue - State: open - Opened by maxkazmsft over 4 years ago
Labels: Type: Feature, Prior: Low

GitHub / microsoft/seismic-deeplearning issues and pull requests