soumyadipghosh / eventgrad
Event-Triggered Communication in Parallel Machine Learning
☆25Updated 2 years ago
Related projects ⓘ
Alternatives and complementary repositories for eventgrad
- LLM training in simple, raw C/CUDA☆86Updated 6 months ago
- Fairring (FAIR + Herring) is a plug-in for PyTorch that provides a process group for distributed training that outperforms NCCL at large …☆63Updated 2 years ago
- Codebase associated with the PyTorch compiler tutorial☆44Updated 5 years ago
- ☆10Updated 4 years ago
- ☆18Updated 2 years ago
- Memory Optimizations for Deep Learning (ICML 2023)☆60Updated 8 months ago
- Some CUDA design patterns and a bit of template magic for CUDA☆146Updated last year
- A tensor-aware point-to-point communication primitive for machine learning☆249Updated last year
- benchmarking some transformer deployments☆26Updated last year
- Udacity CS344 Introduction to Parallell Programming (https://classroom.udacity.com/courses/cs344), with assignments/materials updated to …☆44Updated 3 years ago
- ☆14Updated last month
- C++ API to log data in tensorboard format.☆76Updated 2 years ago
- Demo dataset for libtorch☆54Updated 2 years ago
- PyTorch RFCs (experimental)☆130Updated 2 months ago
- Introduction to CUDA programming☆113Updated 7 years ago
- Cooperative Primitives for CUDA C++ Kernel Authors. This repository contains CUB PRs from Q4 2019 until Q4 2020.☆22Updated 4 years ago
- A tracing JIT compiler for PyTorch☆12Updated 2 years ago
- ☆22Updated 10 months ago
- CUDA kernel author's tools☆109Updated 2 years ago
- Example code to create and train a Pytorch model using the new C++ frontend.☆17Updated 5 years ago
- Automatic Differentiation for high-performance stencil loops☆12Updated 3 years ago
- Hacks for PyTorch☆17Updated last year
- A library for syntactically rewriting Python programs, pronounced (sinner).☆70Updated 2 years ago
- CUDA templates for tile-sparse matrix multiplication based on CUTLASS.☆49Updated 6 years ago
- a lightweight transformer library for PyTorch☆72Updated 3 years ago
- JMP is a Mixed Precision library for JAX.☆187Updated 6 months ago
- Template for starting CUDA/C++ project using CMake with Github Action for CI☆29Updated last year
- Customized matrix multiplication kernels☆53Updated 2 years ago
- Better bindings for Python☆17Updated last year
- ☆36Updated 10 months ago