weights-and-biases

Track machine learning experiments, sweeps, artifacts, and model registry state.

Updated May 24, 2026
One-click install
npx skills add https://github.com/anxiety135790/hermes-skills --skill weights-and-biases-anxiety135790
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: weights-and-biases
Source: https://github.com/anxiety135790/hermes-skills/tree/main/mlops/evaluation/weights-and-biases
Command: npx skills add https://github.com/anxiety135790/hermes-skills --skill weights-and-biases-anxiety135790

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires wandb, and includes references (resource) components.

What problem does it solve?

This Skill helps you manage the full machine learning experiment workflow so you can stop losing track of runs, hyperparameters, artifacts, and model versions across training iterations and teammates.

Core Features & Use Cases

  • Experiment Tracking: Log metrics, configs, media, and system stats in real time for clearer training visibility.
  • Sweeps & Tuning: Run grid, random, or Bayesian hyperparameter searches to find stronger model settings faster.
  • Artifacts & Registry: Version datasets, checkpoints, and production models with lineage, aliases, and registry promotion.
  • Framework Integrations: Connect W&B to PyTorch, TensorFlow/Keras, HuggingFace, Lightning, Fast.ai, XGBoost, and LightGBM for automatic logging.
  • Use Case: A research team can compare dozens of image-classification runs, promote the best checkpoint to a model registry, and share dashboard results with collaborators.

Quick Start

Ask the skill to set up Weights & Biases for your training project, log your metrics and artifacts, and prepare a sweep configuration for hyperparameter tuning.

Frequently Asked Questions about weights-and-biases

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

FAQPage Schema
How do I track machine learning experiments across PyTorch and HuggingFace training runs?

Track machine learning experiments by logging metrics, configs, media, and system stats in real time. The Skill connects W&B to PyTorch, TensorFlow, HuggingFace, Lightning, Fast.ai, XGBoost, and LightGBM for automatic framework-specific integration and clearer training visibility.

How do I run hyperparameter sweeps for model tuning?

Run hyperparameter sweeps by configuring grid, random, or Bayesian searches to find stronger model settings faster. The Skill handles sweep orchestration for your training project to streamline the tuning process and compare multiple runs.

Can I version datasets and checkpoints using W&B artifacts?

Yes, you can version datasets and checkpoints using W&B artifacts with full lineage tracking. The Skill supports artifact versioning and allows you to promote the best checkpoints to a model registry using aliases for production readiness.

How do I manage model registry state and collaborate with a research team?

Manage model registry state by promoting your best checkpoints with aliases and sharing dashboard results with collaborators. The Skill tracks registry state in a single workflow, allowing research teams to compare dozens of runs and coordinate production model promotion.

Does this Skill support experiment tracking for XGBoost and LightGBM projects?

Yes, experiment tracking supports XGBoost and LightGBM projects alongside PyTorch and TensorFlow. The Skill provides framework-specific integration to automatically log metrics, configs, and system stats during training and evaluation scenarios.