GitHub·· 7 天前AI 评分12
NEAT-AI-Snapshot：NEAT-AI 项目的生物/基因组快照数据仓库
stSoftwareAU/NEAT-AI-Snapshot
AI 导读
NEAT-AI-Snapshot 为 NEAT-AI 项目提供默认的生物/基因组快照数据，通过 GitHub Pages 从 docs/ 目录提供 snapshot.json.gz，供 NEAT-AI-Explore 等下游工具直接获取已知可用的示例，无需重新运行训练。NEAT-AI 项目拆分为七个公开仓库，该仓库负责快照格式与测试数据，由 NEAT-AI 生成、供下游工具消费。
正文
Default creature/genome snapshot data for the NEAT-AI project. The repository serves snapshot.json.gz from docs/ via GitHub Pages so NEAT-AI-Explore and other downstream tools can fetch a known-good example without re-running training.
Related Repositories
The NEAT-AI project is split across seven public repositories. Each focuses on one concern and composes with the others as shown below.
| Repository | Role |
|---|---|
| NEAT-AI | Primary Deno/TypeScript neural-network engine (evolution, training, WASM activation). |
| NEAT-AI-core | Shared native Rust library (neat-core) with numerics, topology helpers, and the chunked .bin training stream. |
| NEAT-AI-Discovery | Rust discovery module invoked by NEAT-AI via Deno FFI to search architectures and hyper-parameters. |
| NEAT-AI-Snapshot | Creature/genome snapshot format and fixtures produced by NEAT-AI and consumed by downstream tools. |
| NEAT-AI-scorer | Production forward-only scoring application built on neat-core via a path dependency. |
| NEAT-AI-Explore | Visualiser for creatures that reads NEAT-AI-Snapshot data. |
| NEAT-AI-Examples | Worked examples and tutorials that depend on NEAT-AI. |
Dependency graph
graph TD
Core[NEAT-AI-core<br/>Rust shared lib]
Main[NEAT-AI<br/>Deno/TypeScript engine]
Discovery[NEAT-AI-Discovery<br/>Rust, via Deno FFI]
Snapshot[NEAT-AI-Snapshot<br/>creature data]
Scorer[NEAT-AI-scorer<br/>Rust scorer app]
Explore[NEAT-AI-Explore<br/>visualiser]
Examples[NEAT-AI-Examples<br/>tutorials]
Main -->|Deno FFI| Discovery
Main -->|produces| Snapshot
Scorer -->|path dependency| Core
Explore -->|reads| Snapshot
Examples -->|depends on| Main
Loading
来源：GitHub · github.com