Download
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
Categories
Machine LearningLicense
Apache License V2.0Follow FFCV
Other Useful Business Software
Host LLMs in Production With On-Demand GPUs
Deploy your model, get an endpoint, pay only for compute time. No GPU provisioning or infrastructure management required.
Rate This Project
Login To Rate This Project
User Reviews
Be the first to post a review of FFCV!
