TAP: Accelerating Large-Scale DNN Training Through Tensor Automatic Parallelisation

1 February 2023

Ziji Shi

Yong Li

Papers citing "TAP: Accelerating Large-Scale DNN Training Through Tensor Automatic Parallelisation"

3 / 3 papers shown

Title
ROAM: memory-efficient large DNN training via optimized operator ordering and memory layout Huiyao Shu Ang Wang Ziji Shi Hanyu Zhao Yong Li Lu Lu OffRL 34 2 0 30 Oct 2023
ZeRO-Offload: Democratizing Billion-Scale Model Training Jie Ren Samyam Rajbhandari Reza Yazdani Aminabadi Olatunji Ruwase Shuangyang Yang Minjia Zhang Dong Li Yuxiong He MoE 177 416 0 18 Jan 2021
Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism M. Shoeybi M. Patwary Raul Puri P. LeGresley Jared Casper Bryan Catanzaro MoE 245 1,826 0 17 Sep 2019