eval

Run inference on neural network checkpoints and generate PSNR, SSIM, and LPIPS metrics.

Updated Apr 25, 2026
One-click install
npx skills add https://github.com/Jonah-1/DDSR --skill eval-jonah-1
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: eval
Source: https://github.com/Jonah-1/DDSR/tree/main/.claude/skills/eval
Command: npx skills add https://github.com/Jonah-1/DDSR --skill eval-jonah-1

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill simplifies the process of evaluating trained neural network models by automating inference and metrics analysis, reducing manual effort.

Core Features & Use Cases

  • Automated Inference: Runs model evaluation on specified checkpoints to generate performance metrics.
  • Results Visualization: Displays PSNR, SSIM, LPIPS scores in a clear, formatted table for each scene and overall.
  • Use Case: A researcher completes training and wants to quickly assess model performance across multiple scenes by running a single evaluation command.

Quick Start

Run evaluation on your trained model checkpoint and review the generated metrics and output images with a simple command.

Frequently Asked Questions about eval

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

FAQPage Schema
How do I automate neural network model evaluation and generate performance metrics?

Automating model evaluation requires running inference scripts on your trained neural network checkpoints to generate quantitative performance metrics. This Skill automates that inference process and parses the results into standardized metrics reporting.

What metrics are used for scene reconstruction quality benchmarking?

Scene reconstruction quality benchmarking commonly uses PSNR, SSIM, and LPIPS scores to quantify performance. This Skill evaluates your model checkpoints and displays these specific metrics in a clear, formatted table for each scene and overall.

How do I validate trained deep learning model checkpoints across multiple scenes?

Validating trained deep learning model checkpoints involves running inference across multiple scenes using an evaluation script. You can run a single evaluation command to process the checkpoints and generate standardized quantitative results for each scene.

Can I benchmark neural network inference results without writing custom metric parsing tools?

Benchmarking neural network inference results typically requires metric parsing tools to process checkpoints. This Skill includes the necessary scripts and references to automate the inference and metrics parsing, eliminating the need for custom reporting code.

What is the best way to report PSNR, SSIM, and LPIPS scores for model validation?

The best way to report PSNR, SSIM, and LPIPS scores during model validation is to automate the inference and formatting process. This Skill generates a clear table displaying these scores for each scene and overall, streamlining deep learning benchmarking workflows.