1 个仓库
Awesome GitHub RepositoriesData Prefetching
Asynchronous loading of dataset samples into memory ahead of their consumption in the training pipeline.
Distinct from Database Performance Sampling: Distinct from Database Performance Sampling: focuses on asynchronous pre-loading for ML training pipelines rather than statistical retrieval for tuning.
Explore 1 awesome GitHub repository matching data & databases · Data Prefetching. Refine with filters or upvote what's useful.
Awesome Data Prefetching GitHub Repositories
用 AI 发现最棒的仓库。我们将通过 AI 为您搜索最匹配的仓库。
- facebookresearch/flashlight
facebookresearch/flashlight
5,443在 GitHub 上查看Flashlight is a C++ machine learning library and deep learning framework designed for building and training neural networks. It functions as a tensor manipulation library and an automatic differentiation engine that tracks operations to calculate gradients via backpropagation for model optimization. The project is distinguished by its role as a distributed training framework, utilizing all-reduce gradient synchronization and distributed environments to scale machine learning workloads across multiple nodes and devices. It features a backend-agnostic memory interface and RAII-based management
Implements a background thread pool to prefetch dataset samples, increasing sequential access speed during training.
C++
