SOLAR: A Highly Optimized Data Loading Framework for Distributed Training of CNN-based Scientific Surrogates.
Baixi SunXiaodong YuChengming ZhangJiannan TianSian JinKamil IskraTao ZhouTekin BicerPete BeckmanDingwen TaoPublished in: CoRR (2022)
Keyphrases
- data sets
- database
- distributed data
- data collection
- distributed data sources
- distributed environment
- data sources
- high quality
- highly optimized
- raw data
- lightweight
- wide area network
- data transfer
- synthetic data
- training samples
- data mining
- hidden markov models
- heterogeneous systems
- data quality
- data analysis
- experimental data
- labeled data
- artificial neural networks
- data processing
- data mining techniques
- distributed systems
- general purpose
- image data
- knowledge discovery
- data points
- probability distribution
- prior knowledge