Kirthi Shankar Sivamani's repositories
adversarial-autoencoders
Sklearn compatible estimators for securing deep learning models from malicious attacks.
Transfer-image-styling
This project uses the VGG-19 network to transfer the style/texture of an image onto the content of another image. Implementation of Neural style transfer.
domain-align
Original Source code for our paper: Unsupervised Domain Alignment to Mitigate low-level Dataset Biases
TransformerEngine
A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit floating point (FP8) precision on Hopper GPUs, to provide better performance with lower memory utilization in both training and inference.
efficientCamScan
An implementation: https://nomis80.org/ctmf.pdf in C (pthreads have been used instead of SMID instructions).
observer-networks
Non-Intrusive Detection of Adversarial Deep Learning Attacks via Observer Networks
You-Only-Look-Once
A PyTorch Object detector inspired by the YOLO algorithm
accelerate
🚀 A simple way to train and use PyTorch models with multi-GPU, TPU, mixed-precision
adversarial-robustness-toolbox
Python library for adversarial attacks and defenses for neural networks, logistic regression, decision trees, SVM, gradient boosted trees, and more with multiple framework support
apex
A PyTorch Extension: Tools for easy mixed precision and distributed training in Pytorch
ColossalAI
Making large AI models cheaper, faster and more accessible
DeepLearningExamples
Deep Learning Examples
ECE-463-Proj-2
How does sizeof work?
fast_adversarial
Code for the CVPR 2019 article "Decoupling Direction and Norm for Efficient Gradient-Based L2 Adversarial Attacks and Defenses"
flash-attention
Fast and memory-efficient exact attention
Hierarchical-clustering
A method to classify n dimensional data using hierarchical clustering
k-means-clustering
A method of clustering high dimension data using k-means clustering
linux
Linux kernel source tree
Megatron-LM
Ongoing research training transformer models at scale
miniCamScan
An algorithms that performs input image denoising using mean kernel filtering and adaptive thresholding.
opencv
Open Source Computer Vision Library
periflow-python-sdk
PeriFlow Python SDK
Purdue-Coursework
Repository of Purdue coursework.
pytorch
Tensors and Dynamic neural networks in Python with strong GPU acceleration
scikit-learn
scikit-learn: machine learning in Python
SmartCityApp
Smart City App for the Smart City Project in Purdue University
thermocouple-data-analysis
Analysis of 5 types of thermocouple designs for First Order Systems Inc.
TransformerEngine-1
A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit floating point (FP8) precision on Hopper GPUs, to provide better performance with lower memory utilization in both training and inference.