kakkapriyesh Goto Github PK

acms-80770-deep-learning-with-graphs

ae-convlstm-flow-dynamics

This repository contains an Auto-encoder ConvLSTM network (Pytorch) which can be used to predict a large number of time steps (100+). The network prediction is sequence-to-sequence which works well to predict 5 to 10-time steps in one pass through the neural network. The network is trained for unsteady fluid simulations using data. Another training method tested is the physics constraint method, where governing equations of fluid motion are used to optimize loss. Few attempts to train unsteady Navier-Stokes are made, but it dint work.

convlstm-for-flow-over-cylinder

cuda-implementation-of-pic-algorithm

This is a learning exercise to use GPU for scientific computing. The code is written in CUDA and C++ to create a simulation of two electron beams passing through each other in opposite directions. An interesting phenomenon occurs, called the Two-Stream Instability.

gnn_project_chemical_kinetics_reduction

graph-pde

Using graph network to solve PDEs

meshgraphnets_pytorch

PyTorch implementations of Learning Mesh-based Simulation With Graph Networks

minimal-diffusion

A minimal yet resourceful implementation of diffusion models (along with pretrained models + synthetic images for nine datasets)

natural-convection-in-cavity-using-mpi

These codes might be useful for engineers (CFD/FEM) who wants to understand MPI.

turbulence-flow-and-heat-budgets-in-openfoam

The solver is written in OpenFOAM 6. xx, enabling on-the-fly calculation of terms required to calculate budgets. It also contains a post-processing tool to calculate the necessary gradients for the field's complete flow and heat flux budgets. The work was carried out with Prof. Kameswararao Anupindi at IIT Madras

variable_turbulent_prandtl_number_openfoam

Turbulent Prandtl number gets calculated at each time step