nanochat-llm-training 是做什麼的？
使用 nanochat 訓練您自己的 GPT-2 等級 LLM，成本低於 100 美元，包含 Karpathy 極簡可破解框架，涵蓋 tokenization、預訓練、微調、評估、推論及聊天介面。
使用 nanochat 訓練您自己的 GPT-2 等級 LLM，成本低於 100 美元，包含 Karpathy 極簡可破解框架，涵蓋 tokenization、預訓練、微調、評估、推論及聊天介面。
此技能旨在幫助開發者以極低成本（低於 100 美元）訓練自訂的 GPT-2 等級大型語言模型（LLM）。它基於 nanochat 框架，該框架由 Andrej Karpathy 設計，強調極簡且易於修改，非常適合教育用途或實驗。技能完整涵蓋了從資料標記化（tokenization）、預訓練（pretraining）、指令微調（finetuning）到模型評估（evaluation）、推論（inference）以及簡易聊天 UI 的整個流程。它支援多種平台（如 Claude Code、Cursor 等），可透過命令列工具一鍵執行。對於想要深入理解 LLM 訓練機制或需要輕量級自訂模型的開發者來說，這是一個實用且經濟的入門工具。
npx skills add https://github.com/aradotso/trending-skills/tree/main/skills/nanochat-llm-training使用 nanochat 訓練您自己的 GPT-2 等級 LLM，成本低於 100 美元，包含 Karpathy 極簡可破解框架，涵蓋 tokenization、預訓練、微調、評估、推論及聊天介面。
This repo contains 20 individual skills — each has its own dedicated page.
Self-evolving AI agent system with 26 tools, three-layer memory, MCP plugins, and 24/7 self-repair in pure Python.
A collection of specialized AI agent personalities for Claude Code, Cursor, Aider, Windsurf, and other AI coding tools — covering engineering, design, marketing, sales, and more.
Headless browser automation CLI for AI agents using native Rust binary with Chrome DevTools Protocol
AI coding agent skill for Antigravity Manager — a Tauri v2 + Rust desktop app and Docker service that manages multiple Google/Anthropic accounts and proxies them as standard OpenAI/Anthropic/Gemini API endpoints with intelligent account rotation.
Guide to deploying and managing OpenClaw-compatible AI agent systems across cloud, bare metal, and hybrid infrastructure.
Fully autonomous research pipeline that turns a topic idea into a complete academic paper with real citations, experiments, and conference-ready LaTeX.
Structured prompts, vault templates, and autonomous research workflows for AI-assisted genealogy using Claude Code.
Search and retrieve Norwegian company data from Brønnøysundregistrene (the Norwegian Business Registry). Access all ~1.2 million registered companies in Norway.
Give AI agents access to your live Chrome session via CDP — interact with open tabs, logged-in accounts, and current page state
A Claude Code plugin that displays a real-time HUD showing context usage, active tools, running agents, and todo progress in your terminal statusline.
Enable multiple Claude Code instances to discover each other and exchange messages in real-time via a local broker daemon and MCP server.
Command Line User Interface for Claude Code — a floating macOS desktop overlay with multi-tab sessions, permission approval UI, voice input, and skills marketplace.
AI-native terminal multiplexer with programmable socket API, full Playwright-equivalent browser automation, and agent team coordination — built for Claude Code and autonomous agent workflows
Build a persistent knowledge graph of your codebase so Claude reads only what matters — up to 49x fewer tokens on coding tasks.
Self-directed iterative research skill for Codex that continuously cycles through modify, verify, retain or discard, and repeat until a measurable goal is reached.
Personal AI assistant framework supporting multiple chat channels (DingTalk, Feishu, QQ, Discord, etc.) with extensible skills, local/cloud deployment, and cron scheduling.
AI-powered green screen keyer that unmixes foreground colors and generates clean linear alpha channels using neural networks
Personal intelligence agent that aggregates 27 OSINT data sources into a self-hosted Jarvis-style dashboard with Telegram/Discord bots, LLM analysis, and real-time alerts.
LLM-powered A/H/US stock intelligent analysis system with multi-source data, real-time news, AI decision dashboards, and multi-channel push notifications via GitHub Actions.
Expert skill for using DeepSeek-OCR, a vision-language model for optical character recognition with context optical compression supporting documents, PDFs, and images.
ElevenLabs text-to-speech with mac-style say UX.
Oracle CLI second-model review/debug/refactor/design with selected files, dry-run token checks, API or browser engine.
Capture and automate macOS UI with the Peekaboo CLI.
Generate multiple radically different interface designs for a module using parallel sub-agents. Use when user wants to design an API, explore interface options, compare module shapes, or mentions "design it twice".
Build a throwaway prototype to flesh out a design — a runnable terminal app for state/business-logic questions, or several radically different UI variations toggleable from one route.
Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree. Use when user wants to stress-test a plan, get grilled on their design, or mentions "grill me".