Ankit Shah's Projects
DIVA project repository
本项目将《动手学深度学习》(Dive into Deep Learning)原书中的MXNet实现改为PyTorch实现。
Code for the paper "Rethinking Importance Weighting for Deep Learning under Distribution Shift".
A solution to https://docs.djangoproject.com/en/dev/intro/tutorial01/
Notebooks for learning deep learning
An implementation of a deep learning recommendation model (DLRM)
Development repository for integrating FlexFlow (A distributed deep learning framework that supports flexible parallelization strategies) with DLRM (Deep Learning Recommendation Model)
JAX-based neural network library
The DM Control Suite and Package is a tool for developing and testing reinforcement learning agents for the MuJoCo physics engine.
A Python interface for reinforcement learning environments
Dynamic memory network tensorflow tf.data tf.estimator
A TensorFlow implementation of the Differentiable Neural Computer.
Universal Deep neural network based speech enhancement demo and tools, well pre-trained DNN model
This repo contains the scripts, models, and required files for the Deep Noise Suppression (DNS) Challenge.
Docker images for fastai
Azure Monitor for Containers
Code associated with the Don't Stop Pretraining ACL 2020 paper
Double base number system (DBNS) is an alternative number system besides the binary system. Its representation is similar to the radix number system together with two bases, usually be two and three. DBNS preserves the two important properties: redundancy and sparseness. The redundancy is the property accommodating with the parallelism. In this research, we are interested in parallel addition algorithm on DBNS. Our theoretical result shows that parallel addition in DBNS can be performed. An addition algorithm together with the proof of correctness is described in this paper. In generally, we study the generalization form of DBNS addition algorithms in any sizes.
A Deep-learning based dOcking decoy eValuation mEthod
DoWhy is a Python library for causal inference that supports explicit modeling and testing of causal assumptions. DoWhy is based on a unified language for causal inference, combining causal graphical models and potential outcomes frameworks.
Implementation of "Deep Private-Feature Extraction (DPFE)"
Lua/Torch implementation of DQN (Nature, 2015)
DronaMaps | centimeter level accurate 3D mapping using drones, at your fingertips!
Dataset to assess the disentanglement properties of unsupervised learning methods
dstc7-noesis
DSTC10 Track1 - MOD: Internet Meme Incorporated Open-domain Dialog