This repository contains the DL examples ported to run on Dkube and showcase the features of Dkube platform.
Features such as -
- Deep learning training on known frameworks like Tensorflow, Pytorch...
- Custom container based training which could be using custom frameworks or simple python/C++ code
- GPU based training.
- Distributed training.
- Hyperparameter tuning to find the optimal parameter space.
- Data preprocessing using custom containers.
- Automated workflow using Pipelines.
Following examples are provided,
-
Gray scale digits classification using MNIST network.
Supports samples for GPU based training, distributed training, hyperparameter tuning & pipeline.
-
CatsDogs binary classification using RESNETV2 network.
Supports samples for GPU based training, distributed training, hyperparameter tuning & pipeline.
-
Object detection sample on cocoo pets dataset.
https://github.com/oneconvergence/dkubeio-examples/tree/master/tf/object-detection/pets/README.md
-
Clinical Regression pipeline demo.
https://github.com/oneconvergence/dkubeio-examples/blob/master/tf/clinical_reg/README.md
-
MNIST Digit Classification.
-
Stock prediction using sklearn.
https://github.com/oneconvergence/dkubeio-examples/tree/master/sklearn/README.md
-
Sonar Mines Vs Rocks Classification.
https://github.com/oneconvergence/dkubeio-examples/tree/master/R/classification/sonar/README.md
-
MNIST Digits Classification.
https://github.com/oneconvergence/dkubeio-examples/tree/master/R/classification/mnist/README.md
https://github.com/oneconvergence/dkubeio-examples/tree/master/arv-examples/README.md