skills.rest
ProgMastermind
ProgMastermindCOMMUNITY
Explore all skills by this author →
Website𝕏@PrakashCollymo1GitHub

run-atom-workload

Automate ATOM workload execution for accuracy evaluation, performance benchmarking, and fault reproduction.

Updated Jul 15, 2026
Search Tags:#performance benchmarking#fault reproduction#ATOM#workload execution#accuracy evaluation
One-click install
npx skills add https://github.com/ProgMastermind/ATOM --skill run-atom-workload
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: run-atom-workload
Source: https://github.com/ProgMastermind/ATOM/tree/main/.claude/skills/run-atom-workload
Command: npx skills add https://github.com/ProgMastermind/ATOM --skill run-atom-workload

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill automates the execution of ATOM workloads, streamlining tasks such as accuracy evaluation, performance benchmarking, concurrency sweeps, offline simple inference, and fault reproduction under rocm-debug-agent.

Core Features & Use Cases

  • ATOM Workload Execution: Automates running various ATOM workloads including accuracy eval, performance benchmark, concurrency sweep, offline simple inference, and fault repro.
  • Server-Based Workloads: Handles server-based tasks like GSM8K accuracy eval and benchmarking.
  • Offline Workloads: Supports offline simple inference without a server.
  • Fault Reproduction: Assists in reproducing faults under rocm-debug-agent for debugging.

Quick Start

Run the ATOM workload with the 'run-atom-workload' skill by specifying the desired operation, e.g., 'test accuracy', 'run benchmark', or 'repro the fault'.

Frequently Asked Questions about run-atom-workload

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

FAQPage Schema
How do I automate ATOM workload execution for accuracy evaluation and performance benchmarking?▼

Automating ATOM workload execution is handled by orchestrating script runs to streamline accuracy evaluation and performance benchmarking. You can specify operations like 'test accuracy' or 'run benchmark' to monitor workload completion.

Can I reproduce faults under rocm-debug-agent for AMD ROCm debugging?▼

Yes, you can reproduce faults under rocm-debug-agent by specifying a fault reproduction operation. This automates the execution of ATOM workloads to assist in debugging AMD ROCm-based systems.

Does this approach support offline simple inference without a server?▼

Offline simple inference is supported without requiring a server. The automation executes offline ATOM workloads directly, handling server-less tasks alongside server-based GSM8K accuracy eval and benchmarking.

What is the best way to run concurrency sweeps for AMD ROCm-based workloads?▼

The best way to run concurrency sweeps is by automating ATOM workload execution, which orchestrates and monitors the sweep operations natively for AMD ROCm-based environments.

How do I start running GSM8K accuracy evaluation tasks?▼

Start GSM8K accuracy evaluation by triggering the server-based workload operation. The automation orchestrates the script execution and monitors the workload completion for the benchmarking process.

Related Skills

Explore More Software Engineering Skills→
ProgMastermindProgMastermind

debug-agent-locate-kernel

Diagnoses faulty GPU kernels and silent hangs on ATOM platforms using rocm-debug-agent and rocgdb.

Community
Advanced
sentientspritesentientsprite

trading-analysis

Parse trading bot logs to extract performance metrics and generate Markdown reports.

Community
Intermediate
gdurandvadasgdurandvadas
2

standards-go-performance

Profile Go services with pprof and trace to resolve performance bottlenecks.

Community
Advanced
Alex1980AlexAlex1980Alex

hook-debugging

Diagnose Claude Code hook startup, stdin, and import failures.

Community
Intermediate