DynPartition
PublicOfficial implementation of DynPartition: Automatic Optimal Pipeline Parallelism of Dynamic Neural Networks over Heterogeneous GPU Systems for Inference Tasks
Discover Popular AI-MCP Services - Find Your Perfect Match Instantly
Easy MCP Client Integration - Access Powerful AI Capabilities
Master MCP Usage - From Beginner to Expert
Top MCP Service Performance Rankings - Find Your Best Choice
Publish & Promote Your MCP Services
Official implementation of DynPartition: Automatic Optimal Pipeline Parallelism of Dynamic Neural Networks over Heterogeneous GPU Systems for Inference Tasks