awesome-ai-infrastructures
Infrastructures™ for Machine Learning Training/Inference in Production.
https://github.com/1duo/awesome-ai-infrastructures
Last synced: 4 days ago
JSON representation
-
Uncategorized
-
Uncategorized
- Kubernetes
- scikit-learn
- Apache Arrow - memory data format provides high-performance, high-FPS data visualization, even with very large datasets.
- TensorFlow - node Xeon (i.e., comparable with mainstream GPU).
-
-
Large-Scale / Distributed Learning
-
Milestones
- Image Classification at Supercomputer Scale
- Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour
- recursive halving and doubling algorithm
- PowerAI DDL
- ImageNet Training in Minutes
- Don't Decay the Learning Rate, Increase the Batch Size
- Extremely Large Minibatch SGD: Training ResNet-50 on ImageNet in 15 Minutes
- Highly Scalable Deep Learning Training System with Mixed-Precision: Training ImageNet in Four Minutes
- ImageNet/ResNet-50 Training in 224 Seconds
-
-
Federated Learning
-
Milestones
- Communication-Efficient Learning of Deep Networks from Decentralized Data
- Federated Learning: Strategies for Improving Communication Efficiency
- Practical Secure Aggregation for Privacy-Preserving Machine Learning
- Federated Multi-Task Learning
- Protection Against Reconstruction and Its Applications in Private Federated Learning
- Differentially Private Distributed Learning for Language Modeling Tasks
- Differentially Private Federated Learning: A Client Level Perspective
- Federated Learning for Mobile Keyboard Prediction
- A Generic Framework for Privacy Preserving Deep Learning
-
-
Survey / Reviews
Categories
Sub Categories