bench-run

Community

Automate benchmark runs of LLM models

Authorpshirshov
Version1.0.0
Installs0

System Documentation

What problem does it solve?

This Skill automates the benchmarking process for Language Learning Models (LLMs) by executing benchmark prompts across a set of predefined models in parallel, monitoring progress, and reporting on completion.

Core Features & Use Cases

  • Parallel Execution: Run multiple LLM benchmark prompts simultaneously across different models.
  • Progress Monitoring: Track the progress of each run and report completion.
  • Customization: Allow users to specify prompt IDs, model overrides, and parallelism level.
  • Use Case: Imagine you want to benchmark a set of LLMs on a specific task. This Skill can run all necessary prompts in parallel, saving time and effort.

Quick Start

Use the bench-run skill to execute a benchmark run for the prompt with ID '001-rts-wfc' using the default models and parallelism level.

Dependency Matrix

Required Modules

scripts/run_bench.py

Components

scripts

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: bench-run
Download link: https://github.com/pshirshov/llm-bench-pi-oneshot/archive/main.zip#bench-run

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 620,000+ vetted skills library on demand.