DeepSpeed: System Optimizations Enable Training Deep Learning Models with Over 100 Billion Parameters.
Jeff RasleySamyam RajbhandariOlatunji RuwaseYuxiong HePublished in: KDD (2020)
Keyphrases
- learning models
- learning tasks
- machine learning
- machine learning algorithms
- learning algorithm
- classification models
- training set
- supervised learning
- learning problems
- conditional random fields
- semi supervised learning
- loss function
- training examples
- maximum likelihood
- sparse metric learning
- machine learning models
- neural network
- probabilistic model
- real world