Add cudatoolkit=11.1 to environment.yml (#596)

CHANGELOG.md

@@ -57,6 +57,7 @@ gets uploaded to AzureML, by skipping all test folders.
 - ([#594](https://github.com/microsoft/InnerEye-DeepLearning/pull/594)) Pytorch is now non-deterministic by default. Upgrade to AzureML-SDK 1.36
 - ([#566](https://github.com/microsoft/InnerEye-DeepLearning/pull/566)) Update `hi-ml` dependency to `hi-ml-azure`.
 - ([#572](https://github.com/microsoft/InnerEye-DeepLearning/pull/572)) Updated to new version of hi-ml package
+- ([#596](https://github.com/microsoft/InnerEye-DeepLearning/pull/596)) Add `cudatoolkit=11.1` specification to environment.yml.
 
 ### Fixed
 - ([#593](https://github.com/microsoft/InnerEye-DeepLearning/pull/593)) Bug fix for hi-ml 0.1.11 issue (#130): empty mount point is turned into ".", which fails the AML job
@@ -98,6 +99,7 @@ in inference-only runs when using lightning containers.
  `ScalarModelBase` can be implemented to compute the loss and metrics in a task-specific manner.
 - ([#554](https://github.com/microsoft/InnerEye-DeepLearning/pull/554)) Removed cryptography from list of invalid
  packages in `test_invalid_python_packages` as it is already present as a dependency in our conda environment.
+- ([#596](https://github.com/microsoft/InnerEye-DeepLearning/pull/596)) Removed obsolete `TrainGlaucomaCV` from PR build.
 
 ### Deprecated
 

azure-pipelines/build-pr.yml

@@ -134,30 +134,6 @@ jobs:
  pytest_mark: after_training_2node
  test_run_title: tests_after_training_2node_run
 
- # Train a classification model in cross validation mode
- - job: TrainGlaucomaCV
- variables:
- - name: model
- value: 'GlaucomaPublic'
- - name: tag
- value: 'GlaucomaPR'
- - name: number_of_cross_validation_splits
- value: 2
- - name: more_switches
- value: '--num_epochs=2 --regression_test_folder=RegressionTestResults/PR_GlaucomaCV'
- pool:
- vmImage: 'ubuntu-18.04'
- steps:
- - template: train_template.yml
- parameters:
- wait_for_completion: 'True'
- pytest_mark: ''
- max_run_duration: '1h'
- - template: tests_after_training.yml
- parameters:
- pytest_mark: after_training_glaucoma_cv_run
- test_run_title: tests_after_training_glaucoma_cv_run
-
  - job: TrainHelloWorld
  variables:
  - name: model
@@ -182,7 +158,7 @@ jobs:
  - name: tag
  value: 'HelloContainerPR'
  - name: more_switches
- value: '--num_nodes=2 --regression_test_folder=RegressionTestResults/PR_HelloContainer'
+ value: '--num_nodes=2 --max_num_gpus=2 --regression_test_folder=RegressionTestResults/PR_HelloContainer'
  pool:
  vmImage: 'ubuntu-18.04'
  steps:

environment.yml

@@ -2,7 +2,9 @@ name: InnerEye
 channels:
  - defaults
  - pytorch
+ - conda-forge
 dependencies:
+ - cudatoolkit=11.1
  - pip=20.1.1
  - python=3.7.3
  - pytorch=1.8.0