https://github.com/willguimont/sjm

Manage SLURM jobs on multiple clusters easily.
https://github.com/willguimont/sjm

cli machine-learning ocaml slurm tools

Last synced: about 11 hours ago
JSON representation

Manage SLURM jobs on multiple clusters easily.

• Host: GitHub
• URL: https://github.com/willguimont/sjm
• Owner: willGuimont
• License: mit
• Created: 2023-05-02T17:14:10.000Z (almost 2 years ago)
• Default Branch: master
• Last Pushed: 2024-04-26T13:37:27.000Z (10 months ago)
• Last Synced: 2024-04-26T14:45:20.643Z (10 months ago)
• Topics: cli, machine-learning, ocaml, slurm, tools
• Language: OCaml
• Homepage:
• Size: 16.6 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# sjm: **S**LURM **J**ob **M**anager

Manage SLURM jobs on multiple clusters easily.

## Installation

```bash

git clone https://github.com/willGuimont/sjm

opam install .

```

## Usage

```bash

# Add remote (ssh-copy-id is required)

sjm add  

# List remotes

sjm ls

# Remove remote

sjm rm 

# List jobs on remote

sjm ps 

# Submit job on remote

sjm run   [list of replace patterns of the form "pattern=value" that will be place each $pattern in the script by value]

# Clear tmp jobs cache

sjm clr

# Clear tmp jobs cache on host

sjm clr-remote 

# git pull in a directory on a remote

sjm pull  

```

## Example

```bash

sjm add mycluster [email protected]

# See test_job.sh for an example of a job script

sjm pull mycluster my_project

sjm run mycluster test_job.sh NUMBER_GPU=2 CONFIG=my_config.yml

sjm ps mycluster

```