Delete all training and sampler checkpoints from a Tinker account.
npx skills add https://github.com/kfallah/CLaaS --skill clear-tinker-storage
Please help me install this Agent Skill. Skill: clear-tinker-storage Source: https://github.com/kfallah/CLaaS/tree/main/.claude/skills/clear-tinker-storage Command: npx skills add https://github.com/kfallah/CLaaS --skill clear-tinker-storage
This Skill helps manage cloud storage costs by deleting old checkpoints from your Tinker account, preventing unnecessary charges.
Run the clear-tinker-storage skill to delete all checkpoints from your Tinker account.
High-intent search queries and answers about installing and using this skill.
You can reduce Tinker storage costs by running a cleanup process that automatically identifies and deletes all training and sampler checkpoints associated with your account, freeing up space and lowering your bill.
Cleaning up Tinker storage performs a destructive deletion of all training and sampler checkpoints. Once executed, the accumulated checkpoints from your model experiments are permanently removed to free up space.
Yes, you need a valid TINKER_API_KEY to authenticate the storage cleanup process. The API key is required to access your Tinker Machine Learning account and perform the destructive deletion of checkpoints.
You should run a Tinker storage cleanup when you have accumulated many checkpoints from different model training runs and want to prevent unnecessary charges. This directly reduces your cloud resource consumption.
The primary limitation is that the deletion is destructive and permanent. You should not use this approach if you need to retain or resume experiments from specific training or sampler checkpoints, as they cannot be recovered.
jackspaceOptimize high-throughput PPO+LSTM reinforcement learning training with vectorized environments.
nimadorostkarAnalyze AWS Cost Explorer data to identify and mitigate cost drivers.
wgpsecGuide cloud sizing, lifecycle, and cost controls for red-team infrastructure.
thinking-machines-labSave and manage ML training checkpoints via REST API and CLI.
