Lightning-AI/pytorch-lightning

Save training metadata with the fault tolerance checkpoint

Open

#9,123 opened on Aug 26, 2021

View on GitHub
 (3 comments) (0 reactions) (0 assignees)Python (3,233 forks)batch import
fault tolerancehelp wantedlet's do it!

Repository metrics

Stars
 (26,687 stars)
PR merge metrics
 (Avg merge 9d 15h) (3 merged PRs in 30d)

Description

🚀 Feature

See title

Motivation

Since we will only guarantee fault-tolerance restart for the same number of GPUs and workers (among others), we might want to save metadata about those for archival and error checking.

Pitch

Add extra fields to the checkpoint generated in _on_exception

https://github.com/PyTorchLightning/pytorch-lightning/blob/9d62f248476c6358d8707188f7b20fafa79f8a4f/pytorch_lightning/trainer/trainer.py#L1376-L1381

Additional context

This kind of training "metadata" should get saved with the checkpoint. For example, we will also want to know this for fault-tolerance to fail if the trainer configuration has changed between runs and the user is trying to restore mid-batch.

Originally posted by @carmocca in https://github.com/PyTorchLightning/pytorch-lightning/pull/8515#r677317201

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

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

  • Flash: The fastest way to get a Lightning baseline! A collection of tasks for fast prototyping, baselining, finetuning 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.

Contributor guide

Research direction
Understand the fault tolerance checkpoint mechanism in ` on exception` and identify where to add metadata fields. Then implement the changes to store relevant training configuration such as number of GPUs, workers, etc.
Tech stack
pythonpytorch
Domain
machine learningai
Issue type
Feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
3
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
Half day
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
PythonPyTorch
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
75