experiment-queue

Name: experiment-queue
Availability: InStock
Author: dogekiki

Orchestrate machine learning experiment batches on remote SSH servers with state tracking.

1|Updated Jul 21, 2026

Search Tags:#automation #machine learning #batch processing #ssh #experiment tracking #gpu scheduling

Ready to run

npx skills add https://github.com/dogekiki/SP-test --skill experiment-queue-dogekiki

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: experiment-queue
Source: https://github.com/dogekiki/SP-test/tree/main/.trae/skills/experiment-queue
Command: npx skills add https://github.com/dogekiki/SP-test --skill experiment-queue-dogekiki

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires pyyaml, and includes scripts (resource) components.

What problem does it solve?

This skill eliminates the engineering friction of managing large-scale machine learning experiments, preventing common issues like stale screen sessions, GPU memory exhaustion, and race conditions during wave transitions.

Core Features & Use Cases

Automated Orchestration: Manages multi-seed sweeps and teacher-student training chains with built-in dependency tracking.
Robust Execution: Features OOM-aware retries, automatic stale screen cleanup, and state persistence for crash recovery.
Use Case: When running a grid of 50+ experiments across multiple GPUs, this skill handles the scheduling, monitoring, and result aggregation, ensuring that failed jobs are retried and subsequent phases only trigger once prerequisites are met.

Quick Start

Use the experiment-queue skill to launch the batch of experiments defined in the grid_spec.yaml file on the remote gpu-server.

Frequently Asked Questions about experiment-queue

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

FAQPage Schema

How do I automate large-scale machine learning experiment batches on remote SSH servers?▼

To automate large-scale machine learning experiment batches on remote SSH servers, you can orchestrate multi-seed sweeps and manage GPU scheduling with automated state tracking. This handles wave-based job dependencies and OOM-aware retries for complex training pipelines.

What is the best way to manage GPU scheduling and prevent OOM errors during multi-seed sweeps?▼

Managing GPU scheduling and preventing OOM errors during multi-seed sweeps requires robust execution with OOM-aware retries and automatic stale screen cleanup. This ensures persistent state monitoring and crash recovery across distributed GPU environments.

Can I run wave-based job dependencies for teacher-student training chains without race conditions?▼

Yes, you can run wave-based job dependencies for teacher-student training chains without race conditions by using automated orchestration with built-in dependency tracking. Subsequent phases only trigger once prerequisites are met.

Do I need pyyaml to orchestrate experiment tracking across distributed GPU environments?▼

Yes, you need pyyaml to define batch configurations for orchestrating experiment tracking across distributed GPU environments. It serves as the input format for specifying your experiment grids and scheduling parameters.

Why does my machine learning experiment pipeline fail during wave transitions on remote servers?▼

Machine learning experiment pipelines often fail during wave transitions on remote servers due to race conditions, stale screen sessions, and GPU memory exhaustion. Automated state tracking and OOM-aware retries prevent these common engineering friction points.