ML Performance Reading Group Session 8: Megatron-LM

Name: ML Performance Reading Group Session 8: Megatron-LM
Uploaded: Mar 10, 2025
Duration: 4160 s
Description: ML Performance Reading Group Session 8, where we covered the paper "Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism" (https://arxiv.org/abs/1909.08053) Presenter: Daniel Vega-Myhre

EleutherAI2.49K subscribers

1.2K views

Mar 10, 2025

1:09:20

ML Performance Reading Group Session 8, where we covered the paper "Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism" (https://arxiv.org/abs/1909.08053) Presenter: Daniel Vega-Myhre

Download

1 formats

Video Formats

360pmp4114.3 MB

Download

Right-click 'Download' and select 'Save Link As' if the file opens in a new tab.