Picture for Sam Ade Jacobs

Sam Ade Jacobs

Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone

Apr 23, 2024
Viaarxiv icon

DeepSpeed Ulysses: System Optimizations for Enabling Training of Extreme Long Sequence Transformer Models

Sep 25, 2023
Viaarxiv icon

ZeRO++: Extremely Efficient Collective Communication for Giant Model Training

Add code
Jun 16, 2023
Figure 1 for ZeRO++: Extremely Efficient Collective Communication for Giant Model Training
Figure 2 for ZeRO++: Extremely Efficient Collective Communication for Giant Model Training
Figure 3 for ZeRO++: Extremely Efficient Collective Communication for Giant Model Training
Figure 4 for ZeRO++: Extremely Efficient Collective Communication for Giant Model Training
Viaarxiv icon

Learning Interpretable Models Through Multi-Objective Neural Architecture Search

Add code
Dec 16, 2021
Figure 1 for Learning Interpretable Models Through Multi-Objective Neural Architecture Search
Figure 2 for Learning Interpretable Models Through Multi-Objective Neural Architecture Search
Figure 3 for Learning Interpretable Models Through Multi-Objective Neural Architecture Search
Figure 4 for Learning Interpretable Models Through Multi-Objective Neural Architecture Search
Viaarxiv icon

Merlin: Enabling Machine Learning-Ready HPC Ensembles

Add code
Dec 05, 2019
Figure 1 for Merlin: Enabling Machine Learning-Ready HPC Ensembles
Figure 2 for Merlin: Enabling Machine Learning-Ready HPC Ensembles
Figure 3 for Merlin: Enabling Machine Learning-Ready HPC Ensembles
Figure 4 for Merlin: Enabling Machine Learning-Ready HPC Ensembles
Viaarxiv icon

Parallelizing Training of Deep Generative Models on Massive Scientific Datasets

Add code
Oct 05, 2019
Figure 1 for Parallelizing Training of Deep Generative Models on Massive Scientific Datasets
Figure 2 for Parallelizing Training of Deep Generative Models on Massive Scientific Datasets
Figure 3 for Parallelizing Training of Deep Generative Models on Massive Scientific Datasets
Figure 4 for Parallelizing Training of Deep Generative Models on Massive Scientific Datasets
Viaarxiv icon

Scalable Topological Data Analysis and Visualization for Evaluating Data-Driven Models in Scientific Applications

Add code
Jul 19, 2019
Figure 1 for Scalable Topological Data Analysis and Visualization for Evaluating Data-Driven Models in Scientific Applications
Figure 2 for Scalable Topological Data Analysis and Visualization for Evaluating Data-Driven Models in Scientific Applications
Figure 3 for Scalable Topological Data Analysis and Visualization for Evaluating Data-Driven Models in Scientific Applications
Figure 4 for Scalable Topological Data Analysis and Visualization for Evaluating Data-Driven Models in Scientific Applications
Viaarxiv icon

Distinguishing between Normal and Cancer Cells Using Autoencoder Node Saliency

Jan 30, 2019
Figure 1 for Distinguishing between Normal and Cancer Cells Using Autoencoder Node Saliency
Figure 2 for Distinguishing between Normal and Cancer Cells Using Autoencoder Node Saliency
Figure 3 for Distinguishing between Normal and Cancer Cells Using Autoencoder Node Saliency
Figure 4 for Distinguishing between Normal and Cancer Cells Using Autoencoder Node Saliency
Viaarxiv icon