experiment

Scaffold, configure, and launch ML experiments with SLURM submission scripts.

11|1|Updated Feb 10, 2026

Search Tags:#scaffold #config #training #reproducibility #experiment #ml #slurm

Ready to run

npx skills add https://github.com/michaelrizvi/claude-config --skill experiment

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: experiment
Source: https://github.com/michaelrizvi/claude-config/tree/main/skills/experiment
Command: npx skills add https://github.com/michaelrizvi/claude-config --skill experiment

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Reduce setup time and errors by scaffolding, configuring, and launching ML experiments.

Core Features & Use Cases

Scaffold and organize experiments with standardized config, scripts, and run tracking.
Provide guidance on planning, data checks, logging, and reproducibility.
Use Case: planning a new training run and evaluation with SLURM submission.

Quick Start

Ask this Skill to scaffold a new experiment with a minimal config and a SLURM submission script.

Frequently Asked Questions about experiment

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

FAQPage Schema

How do I scaffold and configure ML experiments for reproducible training runs?▼

Scaffold ML experiments by generating standardized configurations, scripts, and run tracking structures to ensure reproducible training. This approach reduces setup time and errors by providing a minimal project configuration with logging hooks for consistent experiment organization.

Can I use this to launch hyperparameter sweeps and evaluations on a SLURM cluster?▼

Yes, you can launch hyperparameter sweeps and evaluations on SLURM-managed clusters. The process provides SLURM submission guidance to ensure reproducible runs across both local machines and managed cluster environments.

What is the best way to organize machine learning experiments to reduce setup errors?▼

The best way to reduce setup errors is scaffolding experiments with standardized configs, scripts, and run tracking. This ensures consistent organization and reproducibility by enforcing minimal project configurations and logging hooks from the start.

Do I need a minimal project configuration to start planning ML training runs?▼

Yes, a minimal project configuration is required to start planning training runs. This setup includes logging hooks and standardized scripts to guarantee reproducible runs when executing evaluations locally or on SLURM clusters.

How does scaffolding an ML experiment improve reproducibility for local and cluster runs?▼

Scaffolding improves reproducibility by enforcing standardized configs and logging hooks across all environments. This ensures that training runs, hyperparameter sweeps, and evaluations execute consistently whether launched on local machines or SLURM-managed clusters.