![Artificial Intelligence on Twitter: "Machine learning #MachineLearning https://t.co/pSHPaI1LOA https://t.co/tiq9P0EFm2" / Twitter Artificial Intelligence on Twitter: "Machine learning #MachineLearning https://t.co/pSHPaI1LOA https://t.co/tiq9P0EFm2" / Twitter](https://pbs.twimg.com/media/FL9YZZjXEAEeRQs.jpg)
Artificial Intelligence on Twitter: "Machine learning #MachineLearning https://t.co/pSHPaI1LOA https://t.co/tiq9P0EFm2" / Twitter
![Konduit on Twitter: "Deeplearning4j on Spark: Introduction Spark should be used when you have a cluster of multi-GPU machines for training and a large enough network to justify a distributed implementation. https://t.co/Sc1pldyWW0 # Konduit on Twitter: "Deeplearning4j on Spark: Introduction Spark should be used when you have a cluster of multi-GPU machines for training and a large enough network to justify a distributed implementation. https://t.co/Sc1pldyWW0 #](https://pbs.twimg.com/media/Ek17CVZVkAEE1ae.jpg)
Konduit on Twitter: "Deeplearning4j on Spark: Introduction Spark should be used when you have a cluster of multi-GPU machines for training and a large enough network to justify a distributed implementation. https://t.co/Sc1pldyWW0 #
![With a GPU, memory continues to grow without releasing · Issue #8236 · eclipse/deeplearning4j · GitHub With a GPU, memory continues to grow without releasing · Issue #8236 · eclipse/deeplearning4j · GitHub](https://user-images.githubusercontent.com/30429640/64839376-e2e25e00-d629-11e9-9ac9-398a27e6dc89.png)