Deploying Ray on a local kubernetes cluster
In this post, I’ll show you how to run the Ray application covered in the previous post as a Kubernetes deployment running on a local […]
Machine Learning
Kernel SHAP
In this post, I will provide the math for eliminating the constraint on the sum of Shap (SHapley Additive exPlanations) values in the KernelSHAP algorithm […]
Building a Information Retrieval system based on the Covid-19 research challenge dataset: Part 3
In the first post of this series, we went through the ETL process used to transform a collection of papers/articles about Covid-19 and other infectious […]
Building a Information Retrieval system based on the Covid-19 research challenge dataset: Part 2
In the previous post, I showed you how the cord19q ETL system parses documents in the Covid-19 corpus into a SQLite database. In this post, […]
Building a Information Retrieval system based on the Covid-19 research challenge dataset: Part 1
In these series of posts, I’ll describe how I created an Information Retrieval system based on the documents released as part of the Covid-19 research […]
Running a containerized MongoDB application using an attached EBS volume data store
Over the past few months, I’ve been learning how to build ML pipelines consisting of containerized micro services. A common task in such pipelines is […]
Time series prediction
In this post, I’ll describe my implementation and explanation of key elements of DeepAR, a probabilistic forecasting method based on autoregressive recurrent networks introduced by […]
Understanding incremental decoding in fairseq
Recently, I’ve been learning about sequence-to-sequence translation systems and going through Pytorch’s fairseq code. I’ve been focusing on the convolutional seq-to-seq method by Gehring et al. […]
Distributed data parallel training using Pytorch on AWS
In this post, I’ll describe how to use distributed data parallel techniques on multiple AWS GPU servers to speed up Machine Learning (ML) training. Along […]
Pipelining data processing and host-to-device data transfer
The purpose of this post is to show how to use multi-threading to parallelize data processing with data transfer from pageable to page-locked memory. I […]