Ability to pre-warm compute instances in Lightning AI

[krishnakalyan3]

🚀 Feature

As an LAI user, I want to avoid the long startup times when I submit my application. It would be great if I could somehow keep an instance prewarmed during my working hours so that I can deploy my applications quickly.

Motivation

The prewarmed instance will result in developer productivity.

Pitch

Have a check box to prewarm instances

• Number of Instances
• Type of Instance
• Duration / a certain time period to be prewarmed

Alternatives

Additional context

---

If you enjoy Lightning, check out our other projects! ⚡

• Metrics: Machine learning metrics for distributed, scalable PyTorch applications.

• Lite: enables pure PyTorch users to scale their existing code on any kind of device while retaining full control over their own loops and optimization logic.

• Flash: The fastest way to get a Lightning baseline! A collection of tasks for fast prototyping, baselining, fine-tuning, and solving problems with deep learning.

• Bolts: Pretrained SOTA Deep Learning models, callbacks, and more for research and production with PyTorch Lightning and PyTorch.

• Lightning Transformers: Flexible interface for high-performance research using SOTA Transformers leveraging PyTorch Lightning, Transformers, and Hydra.

cc @Borda