neural-training

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Neural Training Skill

神经训练Skill

Purpose

用途

Train and optimize neural patterns using SONA, MoE, and EWC++ systems.

使用SONA、MoE和EWC++系统训练并优化神经模式。

When to Trigger

触发时机

Training new patterns
Optimizing agent routing
Knowledge consolidation
Pattern recognition tasks

训练新模式
优化Agent路由
知识巩固
模式识别任务

Intelligence Pipeline

智能流程

RETRIEVE — Fetch relevant patterns via HNSW (150x-12,500x faster)
JUDGE — Evaluate with verdicts (success$failure)
DISTILL — Extract key learnings via LoRA
CONSOLIDATE — Prevent catastrophic forgetting via EWC++

检索（RETRIEVE） — 通过HNSW获取相关模式（速度提升150倍-12500倍）
判断（JUDGE） — 通过结果（成功/失败）进行评估
提炼（DISTILL） — 通过LoRA提取关键知识
巩固（CONSOLIDATE） — 通过EWC++防止灾难性遗忘

Components

组件

Component	Purpose	Performance
SONA	Self-optimizing adaptation	<0.05ms
MoE	Expert routing	8 experts
HNSW	Pattern search	150x-12,500x
EWC++	Prevent forgetting	Continuous
Flash Attention	Speed	2.49x-7.47x

组件	用途	性能
SONA	自优化适配	<0.05ms
MoE	专家路由	8个专家
HNSW	模式搜索	150倍-12500倍
EWC++	防止遗忘	持续生效
Flash Attention	加速	2.49倍-7.47倍

Commands

命令

Train Patterns

训练模式

bash

npx claude-flow neural train --model-type moe --epochs 10

bash

npx claude-flow neural train --model-type moe --epochs 10

Check Status

检查状态

bash

npx claude-flow neural status

bash

npx claude-flow neural status

View Patterns

查看模式

bash

npx claude-flow neural patterns --type all

bash

npx claude-flow neural patterns --type all

Predict

预测

bash

npx claude-flow neural predict --input "task description"

bash

npx claude-flow neural predict --input "task description"

Optimize

优化

bash

npx claude-flow neural optimize --target latency

bash

npx claude-flow neural optimize --target latency

Best Practices

最佳实践

Use pretrain hook for batch learning
Store successful patterns after completion
Consolidate regularly to prevent forgetting
Route based on task complexity

使用预训练钩子进行批量学习
训练完成后保存成功的模式
定期巩固以防止遗忘
根据任务复杂度进行路由