huggingface-trackio

Track and visualize ML training experiments via Python API and CLI.

2|Updated Mar 29, 2026

Search Tags:#monitoring #mlops #machine learning #training #experiment tracking #hugging face

Ready to run

npx skills add https://github.com/NagyVikt/codex-plugins --skill huggingface-trackio-nagyvikt

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: huggingface-trackio
Source: https://github.com/NagyVikt/codex-plugins/tree/main/plugins/hugging-face/skills/trackio
Command: npx skills add https://github.com/NagyVikt/codex-plugins --skill huggingface-trackio-nagyvikt

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This skill addresses the lack of visibility into long-running machine learning experiments by providing structured logging, real-time diagnostic alerts, and persistent dashboarding.

Core Features & Use Cases

Experiment Tracking: Log training metrics like loss and accuracy with a simple Python API that is compatible with existing frameworks like TRL.
Diagnostic Alerts: Fire automated alerts for training failures such as loss divergence, NaN gradients, or training stalls, which can be sent to webhooks or monitored by LLM agents.
Remote Monitoring: Sync local training runs to Hugging Face Spaces to ensure experiment data persists even after cloud instances terminate.

Quick Start

Use the trackio skill to initialize a new project and log the training loss from your current machine learning script.

Frequently Asked Questions about huggingface-trackio

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I monitor machine learning training in real-time?▼

Track machine learning training in real-time by logging metrics like loss and accuracy through a unified Python API. This provides live visualization of your experiment progress and ensures continuous visibility into long-running training runs.

Can I log training metrics with existing frameworks like TRL?▼

Yes, you can log training metrics with frameworks like TRL. The experiment tracking functionality uses a simple Python API that integrates directly with your existing machine learning workflows to capture loss, accuracy, and other metrics.

How do I detect training anomalies like NaN gradients or loss divergence?▼

Detect training anomalies by configuring diagnostic alerts that automatically fire for issues like loss divergence, NaN gradients, or training stalls. These automated alerts can be sent directly to webhooks or monitored by LLM agents.

How do I persist experiment dashboards after cloud instances terminate?▼

Persist experiment dashboards by synchronizing local training runs to Hugging Face Spaces. This remote monitoring capability ensures your experiment data remains accessible and intact even after your underlying cloud instances terminate.

Can LLM agents poll metrics and alerts autonomously?▼

Yes, LLM agents can poll metrics and alerts autonomously. The system provides structured JSON output specifically designed for LLM agents to read, enabling autonomous experiment iteration and automated monitoring of training runs.

What is the best way to debug machine learning training stalls?▼

The best way to debug training stalls is to use diagnostic alert triggering alongside real-time metric logging. This combination provides immediate notification of training failures and continuous visibility into experiment metrics for quick resolution.