ffcv is a drop-in data loading system that dramatically increases data throughput in model training. From gridding to benchmarking to fast research iteration, there are many reasons to want faster model training. Below we present premade codebases for training on ImageNet and CIFAR, including both (a) extensible codebases and (b) numerous premade training configurations.

Features

  • Train an ImageNet model on one GPU in 35 minutes (98¢/model on AWS)
  • Train a CIFAR-10 model on one GPU in 36 seconds (2¢/model on AWS)
  • Train a $YOUR_DATASET model $REALLY_FAST (for $WAY_LESS)
  • Keep your training algorithm the same, just replace the data loader
  • ffcv also comes prepacked with fast, simple code for standard vision benchmarks
  • Accelerate any learning system with ffcv

Project Samples

Project Activity

See All Activity >

Categories

Machine Learning

License

Apache License V2.0

Follow FFCV

FFCV Web Site

Other Useful Business Software
Build Agents and Models on One Platform Icon
Build Agents and Models on One Platform

Everything you need to build production-ready agents and models. Access 200+ Google and third-party AI models and tools.

Gemini Enterprise Agent Platform is Google Cloud's comprehensive platform for developers to build, scale, govern, and optimize agents and models. Choose from Google's most advanced models and third-party models like Anthropic's Claude Model Family.
Try It Free
Rate This Project
Login To Rate This Project

User Reviews

Be the first to post a review of FFCV!

Additional Project Details

Operating Systems

Linux, Mac, Windows

Programming Language

Python

Related Categories

Python Machine Learning Software

Registered

2024-08-07