Very Deep Convolutional Networks for Large-Scale Image Recognition
Citations
624 citations
623 citations
Cites methods from "Very Deep Convolutional Networks fo..."
...Inspired by common loss functions used in structured prediction (Tsochantaridis et al. (2005); Yu & Joachims (2009); Felzenszwalb et al....
[...]
622 citations
Cites methods from "Very Deep Convolutional Networks fo..."
...Some of the popular CNN configurations include: AlexNet [51], VGGNet [52], and GoogLeNet [53]....
[...]
621 citations
621 citations
Cites methods from "Very Deep Convolutional Networks fo..."
...Three neural network models—AlexNet, GoogLeNet and VggNet-A—are investigated....
[...]
...Due to the resource and time constraint, unfortunately, we aren’t able to perform the training of more DNN models like VggNet-A [43] and distributed training beyond 8 workers....
[...]
...DNNs with larger communication-to-computation ratios (e.g. AlexNet and VggNet-A) can benefit more from TernGrad than those with smaller ratios (e.g., GoogLeNet)....
[...]
...Due to the resource and time constraint, unfortunately, we aren’t able to perform the training of more DNN models like VggNet-A [39] and distributed training beyond 8 workers....
[...]
...Even on a very high-end HPC system with InfiniBand and NVLink, TernGrad is still able to double the training speed of VggNet-A on 128 nodes as shown in Figure 5(b)....
[...]
References
49,639 citations
21,729 citations
9,803 citations
9,775 citations
6,061 citations