triton-inference-server/server

How to free multiple gpu memory

Open

#7,825 opened on Nov 22, 2024

View on GitHub
 (1 comment) (0 reactions) (0 assignees)Python (1,304 forks)batch import
help wantedonnxquestion

Repository metrics

Stars
 (6,593 stars)
PR merge metrics
 (Avg merge 2d 16h) (34 merged PRs in 30d)

Description

The question is how do you free memory

https://github.com/triton-inference-server/onnxruntime_backend/issues/103

When the model is deployed to a single card, I can specify real-time release of gpu memory, but if the model is deployed to multiple cards, I don't know what the format looks like

parameters { key: "memory.enable_memory_arena_shrinkage" value: { string_value: "gpu:3" }  }

instance_group [
    {
        count: 1
        kind: KIND_GPU
        gpus: [ 3 ]
    }
]

Contributor guide

Research direction
Investigate how to specify multiple GPU devices in the `memory.enable memory arena shrinkage` parameter. Check the Triton Inference Server documentation for multi GPU memory configuration, and look for examples or issues related to multiple GPU IDs in that parameter.
Tech stack
python
Domain
backendinfrastructure
Issue type
Research
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
3
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
Half day
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Fresh
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
GPUCUDATriton Inference Server
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
40