Skip to content

nnsight

model_runners

Initializing search

ndif-team/nnsight

Home
Getting Started
Documentation
Features
Tutorials
Blog
Status
About

nnsight

ndif-team/nnsight

Home
Getting Started
Getting Started
- Installation
- Quick Start
Documentation
Documentation
- intervention
  intervention
  - envoy
  - inject
  - batching
  - interleaver
  - serialization
  - backends
    
    backends
    
    base
    
    editing
    
    execution
    
    remote
  - tracing
    
    tracing
    
    base
    
    backwards
    
    editing
    
    globals
    
    invoker
    
    iterator
    
    tracer
    
    util
- modeling
  modeling
  - base
  - diffusion
  - huggingface
  - language
  - transformers
  - mixins
    
    mixins
    
    loadable
    
    meta
    
    remoteable
  - vllm
    
    vllm
    
    vllm
    
    batching
    
    sampling
    
    engines
    
    engines
    
    engine
    
    model_runners
    
    model_runners
    
    GPUModelRunner
    
    workers
    
    workers
    
    GPUWorker
- schema
  schema
  - config
  - request
  - response
- ndif
- util
Features
Features
Tutorials
Tutorials
- Tutorials
  Tutorials
  - Getting Started
    Getting Started
    
    Walkthrough
    
    Chat Templates
    
    Remote Access
  - Causal Tracing
    Causal Tracing
    
    Causal Models Intro
    
    Activation Patching
    
    Attribution Patching
    
    Causal Mediation I
    
    Causal Mediation II
    
    DAS
  - Probing
    Probing
    
    Logit Lens
    
    Diffusion Lens
  - Steering
    Steering
    
    Dictionary Learning
    
    LoRA Tutorial
- Mini Papers
  Mini Papers
Blog
Blog
- Archive
  Archive
  - 2026
- Categories
  Categories
  - Ecosystem
  - Release
Status
About
About

Table of contents

model_runners
Submodules

model_runners¶

Model runner implementations for vLLM.

model_runners ¶

Submodules¶

GPUModelRunner - GPU model runner

Copyright © 2026 nnsight