Search Results for "double"

...For example, on BERT-large training, BytePS can achieve ~90% scaling efficiency with 256 GPUs (see below), which is much higher than Horovod+NCCL. In certain scenarios, BytePS can double the training speed compared with Horovod+NCCL. We show our experiment on BERT-large training, which is based on GluonNLP toolkit. The model uses mixed precision. We use Tesla V100 32GB GPUs and set batch size equal to 64 per GPU. Each machine has 8 V100 GPUs (32GB memory) with NVLink-enabled. Machines are inter-connected with 100 Gbps RDMA network. ...

The SPAWNN toolkit is an innovative toolkit for spatial analysis with self-organizing neural networks which is particularily useful for spatial analysis, visualization and geographical data mining. To run the toolkit, simply download and execute (double-click) the jar-file. Please cite: - Hagenauer, J., & Helbich, M. (2016). SPAWNN: A Toolkit for SPatial Analysis With Self-Organizing Neural Networks. Transactions in GIS, 20(5), 755-775. Other related publications: - Hagenauer, J. (2016). Weighted merge context for clustering and quantizing spatial data with self-organizing neural networks. ...

...Performance evaluation reveals that on a single Tesla K40c GPU, LightSpMV is superior to both CUSP and cuSPARSE, with a speedup of up to 2.60 and 2.63 over CUSP, and up to 1.93 and 1.79 over cuSPARSE for single and double precision, respectively.