GitHub / NVIDIA / TransformerEngine
A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit floating point (FP8) precision on Hopper, Ada and Blackwell GPUs, to provide better performance with lower memory utilization in both training and inference.
JSON API: http://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/NVIDIA%2FTransformerEngine
Stars: 2,460
Forks: 429
Open issues: 263
License: apache-2.0
Language: Python
Size: 10.3 MB
Dependencies parsed at: Pending
Created at: over 2 years ago
Updated at: about 14 hours ago
Pushed at: about 14 hours ago
Last synced at: about 10 hours ago
Topics: cuda, deep-learning, fp8, gpu, jax, machine-learning, python, pytorch