skills.rest
ruiyangruiyi
ruiyangruiyiCOMMUNITY
Explore all skills by this author →
GitHub

weights-and-biases

Log ML experiment metrics and configurations to Weights & Biases.

Updated May 26, 2026
Search Tags:#collaboration#model registry#weights and biases#artifact versioning#hyperparameter sweeps#ml experiment tracking
One-click install
npx skills add https://github.com/ruiyangruiyi/hermes-agent --skill weights-and-biases-ruiyangruiyi
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: weights-and-biases
Source: https://github.com/ruiyangruiyi/hermes-agent/tree/main/skills/mlops/evaluation/weights-and-biases
Command: npx skills add https://github.com/ruiyangruiyi/hermes-agent --skill weights-and-biases-ruiyangruiyi

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill helps you reliably log ML experiments so you can compare results across runs, reproduce configurations, and manage models and artifacts over time.

Core Features & Use Cases

  • Experiment tracking: Log metrics, hyperparameters, and run metadata to W&B for fast comparison and review.
  • Hyperparameter sweeps: Run automated searches (grid/random/bayesian) to find better settings with consistent evaluation metrics.
  • Artifacts & model registry: Store versioned datasets, checkpoints, and models with lineage, aliases (e.g., best/production), and promotion workflows.
  • Integrations: Works across common training stacks (e.g., PyTorch, Hugging Face Trainer, PyTorch Lightning, Keras/TensorFlow) via native logging or callbacks.

Example use case: you train an image classifier across many learning rates and optimizers, track validation accuracy for each run, then register the best checkpoint to a model registry stage like staging/production.

Quick Start

Use the weights-and-biases skill to log a training run’s hyperparameters and metrics to W&B for instant dashboards and run comparison.

Frequently Asked Questions about weights-and-biases

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

FAQPage Schema
How do I log ML experiment metrics and hyperparameters for reproducible run history?▼

To log ML experiment metrics, initialize a run with wandb.init, track hyperparameters and metrics via wandb.log, and finalize with wandb.finish to capture reproducible run history for real-time visualization and comparison across PyTorch, Transformers, Lightning, or Keras workflows.

How do I run hyperparameter sweeps to find the best model training settings?▼

Run hyperparameter sweeps by defining a configuration with wandb.sweep and executing automated searches via wandb.agent, supporting grid, random, and bayesian methods to automatically select best-performing settings using consistent evaluation metrics.

How does artifact versioning work for managing datasets and model checkpoints?▼

Artifact versioning stores versioned datasets, checkpoints, and models with lineage tracking, utilizing artifacts and model registry APIs to assign aliases like best or production and manage promotion workflows across training stages.

Can I use W&B experiment tracking with PyTorch Lightning and Hugging Face Trainer?▼

Yes, W&B experiment tracking works with PyTorch Lightning and Hugging Face Trainer through native logging or callbacks, enabling you to log training metrics, configuration metadata, and validation accuracy directly from your existing training stack.

What do I need to set up before tracking ML experiments with Weights and Biases?▼

You need the wandb package installed and authenticated in your environment before calling wandb.init to start logging, ensuring your training configuration and metrics are correctly captured for dashboards and run comparison.

How do I register the best model checkpoint to a staging or production stage?▼

Register the best model checkpoint by using the model registry APIs to store the artifact with lineage, then assign promotion aliases like staging or production to manage the versioned model lifecycle and track its origin training run.

Related Skills

Explore More Data & Analytics Skills→
ovachieverovachiever
52

weights-and-biases

Automate ML experiment tracking and model registry management via the wandb API.

Community
Advanced
carterwayneskhizeinecarterwayneskhizeine
19

weights-and-biases

Log metrics, configs, and artifacts to W&B for experiment tracking.

Community
Intermediate
twjarviszyra-webtwjarviszyra-web
3

weights-and-biases

Track ML experiments, metrics, and artifacts across training runs.

Community
Advanced
DarkArty07DarkArty07
3

weights-and-biases

Track ML experiment metrics, hyperparameters, and model versions with W&B.

Community
Intermediate