SunnyMirror/DeepLearningExamples

Przemek Strzelczyk 9f4678de1b [UNet industrial/TF] Updating for Ampere		5 gadi atpakaļ
..
Colab_UNet_Industrial_TF_TFHub_export.ipynb	4e609f5555 New notebooks in UNet(industrial)/TF	6 gadi atpakaļ
Colab_UNet_Industrial_TF_TFHub_inference_demo.ipynb	4e609f5555 New notebooks in UNet(industrial)/TF	6 gadi atpakaļ
Colab_UNet_Industrial_TF_TFTRT_inference_demo.ipynb	882bc26d40 [UNet_Industrial/TF] Adding Jupyter notebooks and small fixes	6 gadi atpakaļ
README.md	4e609f5555 New notebooks in UNet(industrial)/TF	6 gadi atpakaļ
TensorFlow_UNet_Industrial_TF_train_and_inference.ipynb	9f4678de1b [UNet industrial/TF] Updating for Ampere	5 gadi atpakaļ
download_and_preprocess_dagm2007_public.sh	882bc26d40 [UNet_Industrial/TF] Adding Jupyter notebooks and small fixes	6 gadi atpakaļ

Jupyter demo notebooks

This folder contains demo notebooks for the TensorFlow UNet Industrial model.

1. TensorFlow_UNet_Industrial_TF_train_and_inference.ipynb: end to end training and inference demo.

The most convenient way to make use of the NVIDIA Tensorflow UNet model is via a docker container, which provides a self-contained, isolated and re-producible environment for all experiments. Refer to the Quick Start Guide section of the Readme documentation for a comprehensive guide. We briefly summarize the steps here.

First, clone the repository:

git clone https://github.com/NVIDIA/DeepLearningExamples.git
cd DeepLearningExamples/TensorFlow/Segmentation/UNet_Industrial

Next, build the NVIDIA UNet_Industrial container:

docker build . --rm -t unet_industrial:latest

Then launch the container with:

nvidia-docker run -it --rm \
    --shm-size=2g --ulimit memlock=-1 --ulimit stack=67108864 \
    -v /path/to/dataset:/data/dagm2007/ \
    -v /path/to/results:/results \
    unet_industrial:latest

where /path/to/dataset is the path on the host machine where the data was/is to be downloaded. More on data set preparation in the next section. /path/to/results is wher the trained model will be stored.

Within the docker interactive bash session, start Jupyter with

jupyter notebook --ip 0.0.0.0 --port 8888

Then open the Jupyter GUI interface on your host machine at http://localhost:8888. Within the container, this notebook itself is located at /workspace/unet_industrial/notebooks.

2. Colab_UNet_Industrial_TF_TFTRT_inference_demo.ipynb: inference from a pretrained UNet model with TensorFlow-TensorRT (TF-TRT).

This notebook is designed to run on Google Colab via this link

3. Colab_UNet_Industrial_TF_TFHub_export.ipynb: Colab notebook demostrating creation of TF-Hub module from NVIDIA NGC UNet model.

This notebook is designed to run on Google Colab vie this link

4. Colab_UNet_Industrial_TF_TFHub_inference_demo.ipynb: Colab notebook demostrating inference with TF-Hub UNet module.