Skip to content

Files

Latest commit

 

History

History
28 lines (22 loc) · 679 Bytes

README.md

File metadata and controls

28 lines (22 loc) · 679 Bytes

SWNM - Sliding Window Pipeline

This repository started to keep track of the NM Sliding Window Pipeline for training the main BERT model.

Note that currently you can't close the ssh connection before the script finishes.

Connect to DGX

  1. Connect using SSH:
ssh [user]@[public-ip] -p [ssh-port]

Usage

  1. Build the image:
sudo docker build -t nextgpt . --force-rm --no-cache
  1. Run the container in interactive/detached mode:
sudo docker run -it -d --name="gabrieldamata" --cpus="8.0" --memory="64g" nextgpt
  1. Enter exec mode using the container ID
docker ps --filter ancestor=nextgpt

docker exec -it [container-id] /bin/bash