Skip to content
Undercurrent
Benchmarks
Initializing search
wrynx/undercurrent
Undercurrent
wrynx/undercurrent
Home
Getting started
Getting started
Installation
Quickstart
Concepts
Concepts
Architecture
Extraction points & specs
Position selectors
Probes & lifecycle
Execution modes: inline vs async
Interventions
Guides
Guides
Write a custom probe
Production
Production
Deploy with vLLM
Embed in your serving stack
Async execution & backpressure
Intervention policies & timeouts
Observation sinks
Metrics
How it works
How it works
The vLLM adapter
Tensor & pipeline parallelism
Examples
API reference
API reference
Overview
undercurrent.model
undercurrent.spec
undercurrent.core
undercurrent.router
undercurrent.sinks
undercurrent.adapters
CLI
Compatibility
Benchmarks
Comparison
About
About
Intended use & limitations
Privacy & data handling
Changelog
Contributing
Benchmarks
¶
Work in progress
This page is being written.
Back to top