Integrations / AI Toolkit

RunExperimental

Run AI Toolkit fine-tuning jobs

Run an AI Toolkit training config as a finite job and download the trained weights.

A setup guide only. Badgr has not tested this path, so expect to adjust it.

Why use Badgr for AI Toolkit

Badgr interface
Run
Where AI Toolkit fits
Training runtime
Why Badgr
Model fine-tuning
  • Run a training config as a finite job with a maximum cost.
  • Keep the config in git and rerun it exactly.
  • Checkpoints and logs come back with a receipt. This path is a guide, not a tested template.

What you can do

  • Train a LoRA for a diffusion model from a YAML config
  • Repeat a training run with a changed dataset
  • Keep a config in git and rerun it exactly

Quickstart: Run a training config

badgr run . --cmd "python run.py config/my_lora.yaml" --max-cost 10

How it works

  1. 1.You run one command
  2. 2.Badgr checks it and picks a GPU
  3. 3.Your code runs and logs stream back
  4. 4.Outputs are saved and billing stops

What you get

  • Logs while it runs
  • Output files, downloaded with badgr artifacts <id>
  • A receipt with cost and duration
  • Checkpoints and sample images the config writes to its output folder

Requirements

  • The Badgr CLI installed and logged in (npm install -g badgr-cli)
  • A funded Badgr balance
  • A project folder containing the AI Toolkit code or a requirements file, your config and your dataset
  • A GPU with enough memory for the base model

What it costs

  • A job is billed for the time it runs and stops billing when your command exits.
  • Pass --max-cost to set the most a job can spend. Badgr stops it if it reaches the cap.
  • 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

  • Different results on each run

    Set a fixed seed in the config and pin package versions so runs can be compared.

  • Checkpoints not saved

    Point the config's output folder at the job's output folder.

  • Out of memory

    Lower the resolution or batch size, or quantise the base model if the config allows.

  • Not sure it will run

    Run badgr check on your project first. It validates the setup without starting a paid job.

Developer reference

Ready to start with AI Toolkit?

Every run or service has a spending limit you set before it starts.