Integrations / Triton Inference Server
DeployExperimental
Triton Inference Server on Badgr: custom container guide
A guide for running NVIDIA Triton from your own container with your own model repository.
A setup guide only. Badgr has not tested this path, so expect to adjust it.
Why use Badgr for Triton Inference Server
- Badgr interface
- Serve
- Where Triton Inference Server fits
- Inference server
- Why Badgr
- Host model repositories
- Run your own Triton container with your model repository on a rented GPU.
- A cost cap and a stop command bound the spend.
- Badgr has not tested this path. The page explains ports and readiness so you can check it yourself.
What you can do
- Serve several models behind one Triton server
- Host a model in a format Triton supports
- Keep an existing model repository
Quickstart: Run your own container
badgr serve --image nvcr.io/nvidia/tritonserver:<tag>-py3 --cmd "tritonserver --model-repository=/models" --max-cost 5
How it works
- 1.You give Badgr your container image and start command
- 2.Badgr runs it on a GPU
- 3.You check that it answers
- 4.You stop it when done
What you get
- A Triton address if the port is exposed
- Logs and a receipt
Requirements
- The Badgr CLI installed and logged in (npm install -g badgr-cli)
- A funded Badgr balance
- A Triton container image
- A model repository included in the image or downloaded at start
- Knowledge of the ports Triton listens on (8000 HTTP, 8001 gRPC by default)
What it costs
- Badgr has no ready-made setup for this. The command below runs your own container with the generic custom option. It has not been tested by Badgr, so treat this page as a setup guide.
- A service bills for as long as it runs, until you stop it with badgr down <id>.
- Pass --max-cost to bound spend. Badgr shows the rate for the GPU it picked before it starts. This page does not quote a fixed price.
See pricing for how billing works.
Proof
No tested run is recorded for this integration yet. A receipt and example output will be added here after a real run, not before.
Troubleshooting
Model repository not found
Make sure /models exists in the image and holds one folder per model with its config.
Not ready
Triton reports readiness at /v2/health/ready. Check it and read the logs for model load errors.
Port not reachable
Confirm the port your client uses is the port exposed by the service.
Developer reference
Ready to start with Triton Inference Server?
Every run or service has a spending limit you set before it starts.