Skip to content

Commit bf0fbb8

Browse files
porteratzopsfoley
andauthored
Update HippMapp3r plan, requirments and readme (#1524)
* fixes Signed-off-by: porteratzo <[email protected]> * Update openfl-workspace/keras/hippmapp3r/README.md Co-authored-by: Patrick Foley <[email protected]> Signed-off-by: porteratzo <[email protected]> * Update openfl-workspace/keras/hippmapp3r/README.md Co-authored-by: Patrick Foley <[email protected]> Signed-off-by: porteratzo <[email protected]> --------- Signed-off-by: porteratzo <[email protected]> Co-authored-by: Patrick Foley <[email protected]>
1 parent ccf82d4 commit bf0fbb8

File tree

3 files changed

+13
-4
lines changed

3 files changed

+13
-4
lines changed

openfl-workspace/keras/hippmapp3r/README.md

Lines changed: 8 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -9,6 +9,14 @@ To set up the data, run the following command:
99
python src/setup_data.py --num_collaborators $MAX_NUMBER_OF_COLLABORATORS --total_dataset_size_per_col_MB $DESIRED_DATASET_SIZE
1010
```
1111

12+
## Known Issues
13+
14+
An error has been observed on 2nd Generation Intel® Xeon® Scalable Processors (codenamed Cascade Lake): "CPU implementation of Conv3D currently only supports the NHWC tensor format."
15+
16+
### Workarounds:
17+
1. Use a more recent CPU, such as 3rd Generation Intel® Xeon® Scalable Processors (codenamed Ice Lake), which supports the required tensor format.
18+
2. Alternatively, use a GPU to avoid this issue entirely.
19+
1220
## Reference
1321

1422
Goubran, M., Ntiri E., Akhavein, H., Holmes, M., Nestor, S., Ramirez, J., Adamo, S., Gao, F., Ozzoude, M., Scott, C., Martel, A., Swardfager, W., Masellis, M., Swartz, R., MacIntosh B, and Black, SE. “Hippocampal segmentation for atrophied brains using three-dimensional convolutional neural networks”. Human Brain Mapping. 2020 Feb 1;41(2):291-308. https://onlinelibrary.wiley.com/doi/full/10.1002/hbm.24811

openfl-workspace/keras/hippmapp3r/plan/plan.yaml

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -22,6 +22,7 @@ data_loader :
2222
template : src.dataloader.KerasHippmapp3rsynth
2323
settings :
2424
batch_size : 1
25+
input_shape : [160, 160, 128]
2526

2627
task_runner :
2728
defaults : plan/defaults/task_runner.yaml
Lines changed: 4 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -1,5 +1,5 @@
11
keras==2.15.0
2-
tensorflow
3-
keras-contrib @ git+https://www.github.com/keras-team/keras-contrib.git
4-
gdown
5-
scikit-learn
2+
tensorflow-cpu==2.15.1
3+
keras-contrib @ git+https://www.github.com/keras-team/keras-contrib.git@3fc5ef709e061416f4bc8a92ca3750c824b5d2b0
4+
gdown==5.2.0
5+
scikit-learn==1.6.1

0 commit comments

Comments
 (0)