Manage theseus ML training job lifecycles via YAML configs and cluster dispatch.
npx skills add https://github.com/Jemoka/theseus --skill theseus-cli
Please help me install this Agent Skill. Skill: theseus-cli Source: https://github.com/Jemoka/theseus/tree/main/.claude/skills/theseus-cli Command: npx skills add https://github.com/Jemoka/theseus --skill theseus-cli
This Skill streamlines the end-to-end workflow for running research training jobs with the theseus CLI by helping users generate and edit YAML configs, run experiments locally, dispatch to remote clusters, start remote REPLs, create bootstrap scripts, and manage checkpoints and restores.
Run uv run theseus configure gpt/train/pretrain train.yaml --chip h100 -n 8 to generate a config skeleton, inspect and edit the YAML as needed, then run or submit after confirming the full command.
High-intent search queries and answers about installing and using this skill.
To configure ML training jobs for a SLURM cluster, generate a YAML job spec from a registered schema, specify hardware requests like chip and n_chips, and submit via the CLI after confirming the full dispatch command.
Running theseus ML training experiments locally involves creating a YAML config skeleton, editing job specifications, and executing the job via the CLI with correct output root handling and optional checkpoint restoration.
Yes, you can dispatch training jobs to TPU or Volcano clusters by defining request fields like chip, n_chips, and n_shards in your YAML config and submitting through the CLI with a valid dispatch configuration.
Yes, submitting jobs remotely over SSH requires an optional ~/.theseus.yaml dispatch config file alongside a valid job schema and job name defined in your repository's YAML job specifications.
To manage checkpoints and restore failed training experiments, use the CLI to operate checkpoint files and bootstrap scripts, allowing you to resume local execution or remote cluster jobs from the last saved state.
The best way to generate YAML job specs for multi-GPU pretraining is to run the CLI configure command with a registered job schema, then merge previous configs with hardware overrides like chip type and chip count.
dlt-hubCheck dltHub Runtime job status, view logs, and cancel jobs.
randDiscover ML and AI skills during model training and inference workflows.
Deepmind666Automate pre-flight checks and execution protocols for IoT experiment runs.
ZardLi1115Provision on-demand GPU instances for ML training and inference workloads.
