Skip to content
Get packs

Trackio Experiment Tracking

Log ML training metrics, fire alerts, and query results from a live dashboard

Source Download skill
Repository
huggingface/skills
Stars
11,253
Licence
Apache-2.0
Files
4 files

What it does

This skill teaches an AI agent to use Trackio, a library for recording machine learning training runs. Once installed, the agent can add logging calls to training code, insert alerts for problems like loss spikes or NaN values, and use a command-line tool to read back metrics and alerts in JSON form. Results can be synced to a Hugging Face Space to view as a live dashboard.

Great for

  • ML engineers who want a lightweight way to monitor training runs from a shared dashboard
  • Researchers who let an AI agent run and adjust experiments autonomously and need machine-readable status
  • Teams using TRL that want its training reports sent to Trackio

Skip it if

  • Users who need private experiment data but do not set private=True: auto-created Spaces are public by default, so metrics may be exposed publicly
  • Non-technical users with no Python training code, since the skill is about instrumenting training scripts
  • Production monitoring of non-ML systems, which this skill does not cover

Try asking

  • “Add Trackio logging to my PyTorch training script and sync the dashboard to a private Space”
  • “Insert alerts that fire when loss exceeds 5.0 after 100 steps or goes NaN”
  • “Check the Trackio alerts for my project since yesterday and tell me whether to restart the run”
  • “Show me the final loss and accuracy for run my-run in project my-project as JSON”
Step-by-step guide: how to use Trackio Experiment TrackingInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 4 original files, the Apache-2.0 licence and an attribution note.

    huggingface-trackio.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip huggingface-trackio.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip huggingface-trackio.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Pass private=True to trackio.init() whenever metrics should not be public, since Spaces default to public
  • Review generated alert and webhook settings before running, especially if sending data to Slack or Discord
  • Use --json output and --since timestamps when polling alerts from a background run

What's inside

  • references/alerts.md
  • references/logging_metrics.md
  • references/retrieving_metrics.md
  • SKILL.md

Redistributed unmodified from huggingface/skills under the Apache-2.0 licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.