facebookincubator/AITemplate

AITemplate is a Python framework which renders neural network into high performance CUDA/HIP C++ code. Specialized for FP16 TensorCore (NVIDIA GPU) and MatrixCore (AMD GPU) inference.

PythonC++CudaCShellNix
This is stars and forks stats for /facebookincubator/AITemplate repository. As of 09 May, 2024 this repository has 4272 stars and 338 forks.

AITemplate | | AITemplate (AIT) is a Python framework that transforms deep neural networks into CUDA (NVIDIA GPU) / HIP (AMD GPU) C++ code for lightning-fast inference serving. AITemplate highlights include: High performance: close to roofline fp16 TensorCore (NVIDIA GPU) / MatrixCore (AMD GPU) performance on major models, including ResNet, MaskRCNN, BERT, VisionTransformer, Stable Diffusion, etc. Unified, open, and flexible. Seamless fp16 deep neural network models for NVIDIA GPU or AMD GPU. Fully...
Read on GithubGithub Stats Page
repotechsstarsweeklyforksweekly
snaplet/postgres-wasmShellPythonPerl2.1k0680
asottile/pyupgradePython2.9k01670
rafaelvleite/centipawn_loss_calculatorPython520250
prometheus/client_pythonPython3.6k07810
HobbitLong/SupContrastPython2.6k04720
LibrePhotos/librephotosPythonShell5.9k02560
jito-labs/searcher-examplesRustDockerfileShell1160220
subconsciousnetwork/noosphereRustSwiftTypeScript553+1133+1
mozilla/uniffi-rsRustPythonKotlin1.9k01550
aurae-runtime/auraescriptRustShellMakefile1.8k0920