Automate SLURM workflow management for MATS GPU jobs.
npx skills add https://github.com/yulonglin/dotfiles --skill mats-slurm
Please help me install this Agent Skill. Skill: mats-slurm Source: https://github.com/yulonglin/dotfiles/tree/main/claude/local-marketplace/plugins/research-toolkit/skills/mats-slurm Command: npx skills add https://github.com/yulonglin/dotfiles --skill mats-slurm
Set up and use SLURM helpers to manage interactive and batch GPU workloads on the MATS compute cluster, providing ready-to-run aliases and utilities that simplify resource requests, monitoring, and logging.
Source the mats-slurm aliases file in your shell configuration to enable the SLURM helpers.
High-intent search queries and answers about installing and using this skill.
Use the gbatch or lbatch wrappers to submit multi-GPU batch jobs to SLURM with consistent default options for A100 and L40 partitions. These helpers apply sensible resource requests automatically.
Use the grun or lrun aliases to launch interactive A100 or L40 GPU sessions with sensible defaults. These shortcuts handle the SLURM resource requests for immediate terminal access.
SLURM monitoring and debugging workflows are simplified with helper functions that provide quick access to queue information and job logs. These utilities streamline tracking active GPU workloads.
No, you can use the provided environment templates and references to configure default partitions and log paths. Sourcing the aliases file in your shell configuration sets up reproducible run environments.
Use ready-to-run aliases and utilities that simplify resource requests, monitoring, and logging for SLURM GPU workloads. These wrappers ensure consistent options across interactive and batch submissions.
aaaxnLaunch and manage GPU-accelerated SLURM jobs on the SONIC cluster.
niznik-devOrchestrate SLURM fine-tuning jobs with torchtune and evaluate with inspect-ai.
TianyuDuOperate Stanford's Marlowe HPC cluster with Slurm job submission and monitoring.
dongzhuoyaoSubmit GPU training jobs to Slurm clusters with preflight validation and monitoring.
