Manage GPU compute job lifecycle on the Qizhi platform via kubectl-style CLI.
npx skills add https://github.com/Shallow-W/llm-wiki --skill qzcli-shallow-w
Please help me install this Agent Skill. Skill: qzcli Source: https://github.com/Shallow-W/llm-wiki/tree/main/.claude/skills/qzcli Command: npx skills add https://github.com/Shallow-W/llm-wiki --skill qzcli-shallow-w
It solves the problem of submitting, monitoring, and stopping GPU compute jobs on the Qizhi (启智) platform without manually navigating dashboards, keeping track of job IDs, or dealing with stale cached resource metadata.
Run qzcli login, then run qzcli res -u to refresh your workspace and compute-group cache.
High-intent search queries and answers about installing and using this skill.
You can submit, monitor, and stop GPU compute jobs on the Qizhi platform using a kubectl-style CLI. This approach replaces manual dashboard navigation by managing the job lifecycle and tracking job statuses directly from the command line.
Yes, you can query compute capacity and availability, including low-priority nodes, to find enough free nodes before launching distributed training. This queue-friendly check helps ensure sufficient resources are available for your job.
Run the resource refresh command to update your cached workspaces and compute groups. This ensures you target the right resources and prevents issues caused by stale cached resource metadata during job submission.
Yes, the Qizhi CLI supports command parameters for batch submission and non-interactive job creation. You can submit jobs interactively or non-interactively, then list, watch, and stop jobs by ID across available nodes.
Your Qizhi GPU jobs may fail if the cached resource metadata is stale. Refresh your workspace and compute group cache using the resource update command to ensure the CLI targets the correct resources during job creation.
dogekikiManage GPU compute jobs on the Qizhi platform via command-line interface.
NVIDIAAutomate per-action Vision-Language Model evaluation via the BP evaluation-ms HTTP API.
truongnatOrchestrate Kubernetes workloads with Helm, Kustomize, RBAC, and networking.
zhuyingqinManage GPU compute jobs on the Qizhi platform via kubectl-style qzcli commands.
