Guide Lightning-based ML projects using LightBox templates.
npx skills add https://github.com/neil-tan/LightBox --skill lightbox
Please help me install this Agent Skill. Skill: lightbox Source: https://github.com/neil-tan/LightBox/tree/main/%7B%7Bcookiecutter.project_slug%7D%7D/.claude/skills/lightbox Command: npx skills add https://github.com/neil-tan/LightBox --skill lightbox
Coordinating Lightning-based ML workflows—across LightningReflow, LightningTune, and DataPorter—without cohesive guidance leads to setup friction, inconsistent runtimes, and debugging headaches. This skill provides a comprehensive integration guide that helps teams implement, troubleshoot, and optimize end-to-end Lightning projects with LightBox templates.
Create a LightBox-based project template and begin training with your custom model.
High-intent search queries and answers about installing and using this skill.
Optimize hyperparameters in a Lightning ML pipeline by applying LightningTune patterns from LightBox templates to structure and execute HPO workflows. This provides safe configuration patterns for your training commands.
Manage VRAM when training PyTorch Lightning models by applying memory management techniques with torch.compile and LightBox templates. This approach details specific integration points to optimize memory usage across end-to-end workflows.
Load data efficiently for Lightning-based training workflows by integrating DataPorter within LightBox templates. This provides structured guidance for data loading configurations and best-practice patterns for production-like experiments.
Lightning training pipelines may fail during torch.compile integration due to common pitfalls in memory management and configuration. LightBox templates provide debugging tips and error-avoidance strategies to troubleshoot these end-to-end workflow issues.
Yes, you can use LightBox templates to troubleshoot LightningReflow and DataPorter integration. They offer real-world scenarios illustrating how to coordinate these components while avoiding common setup friction and runtime inconsistencies.
Tianyi-Billy-MaAutomate PyTorch training workflows with LightningModule, Trainer, and LightningDataModule.
h4vzzDeploy trained ML models as REST services on Docker or Kubernetes.
SriRamkunamsettyOrganize PyTorch training with a high-level Trainer and distributed workflows.
pluginagentmarketplaceOrchestrates end-to-end ML training pipelines with PyTorch DDP, Optima, and Kubeflow.
