Skills
.Work. You
Rest

reference-generation

Community

Create PyTorch reference code from operator JSON.

Software Engineering#pytorch#code-generation#reference-implementation#nn.Module#operator-description#get_inputs#get_init_inputs
AuthorJust-it
Version1.0.0
Installs0

System Documentation

What problem does it solve?

Generate a reference PyTorch implementation from operator descriptions in JSON to establish a golden standard for testing and documentation.

Core Features & Use Cases

  • Generate a self-contained PyTorch module (nn.Module) along with get_inputs() and get_init_inputs() helpers based on the operator description.
  • Create a portable reference that can be used for regression testing, benchmarking, and as a design guide for downstream implementations.
  • Support exporting to a predictable output path like output/{op_name}_reference.py for easy integration into workflows.

Quick Start

Generate a reference PyTorch module from op_desc.json and save to output/{op_name}_reference.py.

Dependency Matrix

Required Modules

None required

Components

references

💻 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: reference-generation
Download link: https://github.com/Just-it/AscendOpGenAgent/archive/main.zip#reference-generation

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 471,000+ vetted skills library on demand.