惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

酷 壳 – CoolShell
酷 壳 – CoolShell
P
Privacy & Cybersecurity Law Blog
G
GRAHAM CLULEY
T
The Exploit Database - CXSecurity.com
www.infosecurity-magazine.com
www.infosecurity-magazine.com
Threat Intelligence Blog | Flashpoint
Threat Intelligence Blog | Flashpoint
Project Zero
Project Zero
S
Security @ Cisco Blogs
TaoSecurity Blog
TaoSecurity Blog
A
Arctic Wolf
Webroot Blog
Webroot Blog
Hacker News - Newest:
Hacker News - Newest: "LLM"
Security Latest
Security Latest
H
Heimdal Security Blog
N
News and Events Feed by Topic
N
News | PayPal Newsroom
T
Tor Project blog
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
GbyAI
GbyAI
The Last Watchdog
The Last Watchdog
Y
Y Combinator Blog
宝玉的分享
宝玉的分享
Scott Helme
Scott Helme
A
About on SuperTechFans
M
MIT News - Artificial intelligence
V
V2EX
V
Visual Studio Blog
Recorded Future
Recorded Future
博客园 - 叶小钗
F
Fortinet All Blogs
L
Lohrmann on Cybersecurity
The GitHub Blog
The GitHub Blog
博客园 - Franky
P
Proofpoint News Feed
MyScale Blog
MyScale Blog
D
Darknet – Hacking Tools, Hacker News & Cyber Security
S
Secure Thoughts
D
DataBreaches.Net
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
博客园 - 三生石上(FineUI控件)
I
InfoQ
SecWiki News
SecWiki News
Blog — PlanetScale
Blog — PlanetScale
Engineering at Meta
Engineering at Meta
J
Java Code Geeks
B
Blog RSS Feed
AWS News Blog
AWS News Blog
Know Your Adversary
Know Your Adversary
V
Vulnerabilities – Threatpost
H
Help Net Security

Show HN

GitHub - steveking-gh/firmion: Firmion is DSL and engine for firmware image generation. GitHub - villagesql/villagesql-skills: Agent skills for VillageSQL - gemini-cli-extension; claude-code-plugin GitHub - flightdeckhq/flightdeck: Observability and control plane for AI agents. CSP Radar GitHub - Light-Heart-Labs/DreamServer: Turn your PC, Mac, or Linux box into an AI server. LLM inference, chat UI, voice, agents, workflows, RAG, and image generation. GitHub - Diplomat-ai/diplomat-agent-ts: What can your TypeScript AI agent do to the real world? Scan your code. See which tool calls have zero checks Code Block Selector - Visual Studio Marketplace Prometheus dependency graph — interactive showcase | Riftmap Show HN: I made a vi-like modal keyboard plugin for Figma GitHub - run-llama/liteparse: A fast, helpful, and open-source document parser GitHub - dalemyers/Roar: A macOS CLI tool for notifications GitHub - district-solutions/open-agent-tools-coder: Enables small-to-large self-hosted ai models to use local source code when running tool-calling agentic workloads. We actively data mine 20,900+ (2+ TB) popular github repos using large and small ai models to create reuseable: json, markdown and parquet files for local-first tool-calling models. GitHub - progapandist/stripeek: A local TUI proxy for real-time Stripe API debugging, built for navigating complex payloads fast. GitHub - sir1st/hermes-desktop: All-in-one cross-platform desktop app for Hermes Agent — bundles Python + hermes-agent + hermes-web-ui GitHub - astefanutti/shaderbang: Shebang for Shaders Show HN: Generate Claude Code Workflows using Spec Driven Development approach GitHub - nixys/nxs-universal-chart: The Helm chart you can use to install any of your applications into Kubernetes/OpenShift Show HN: AI agents for UK GDAD PCF roles and their skills The Two Pillars: Mixer Mode and Meta-Software in the Reorganization of Software Work After AI GitHub - JaiCode08/teleport-env What 1,000+ Harness Experiments Taught Me About Self-Improving Agents Show HN: Liiists, a Markdown-first, iOS and CLI list app SwiperTab – Get this Extension for 🦊 Firefox (en-US) GitHub - kouhxp/fftext: Summarize, explain, fact-check, or translate any text, URL, or file. No GPU. No cloud. One command GitHub - sweetpad-dev/sweetpad: Develop Swift/iOS projects using VSCode GitHub - dogmaticdev/IRON: IRON a.k.a. Intermediate Representation Object Notation is a Interpreter/Database that is used to create Programming Languages. GitHub - sjhalani7/vaen: Package your AI coding harness into a portable .agent file, and share it across repos, teams, & the community without ever having to copy-paste instructions, skills, MCP config, or secrets. Show HN: Gandalf the Grader Show HN: Citadeld – replay any CI failure locally from a single file GitHub - tdortman/cuSBF: High-Performance GPU Super Bloom Filter coral-ai/claude-code-token-xray at main · Coral-Bricks-AI/coral-ai GitHub - ulyssestenn/funes: Funes is a Git-based framework for LLM-managed knowledge work: an AI Librarian ingests raw sources, builds an interlinked Markdown knowledge base, and uses it to produce cited reports, analyses, and other outputs. GitHub - ThatXliner/gah: Git Add Hunk, built for agents to use GitHub - harmont-dev/harmont-cli: Command-line client for the Harmont CI platform GitHub - brooksmcmillin/mcp-authflow: OAuth 2.0 Authorization Server framework for MCP servers GitHub - javaid-codes/audit-supply-chain-agents GitHub - amorey/gochan: A small library of common channel architectures for Go, inspired by Rust GitHub - arifozgun/OpenGem: Free, Open-Source AI API Gateway with Gemini, OpenAI & Anthropic Compatibility in 1 file GitHub - Pranesh950/BioPetals: 🌸 Run BIOxAI models at home, BitTorrent-style. Fine-tuning and inference up to 10x faster than offloading GitHub - cnguyen14/bounty-doctor: Diagnose a GitHub bounty issue before you waste hours: detects honeypot scam repos, AI-bot attempt swarms, and stale contests. Show HN: CoreMCP – MCP Server for On-Prem DBs Show HN: KittyHTML – Render HTML/CSS as an inline image in your terminal GitHub - bingud/filemat: Web-based file manager Show HN: TruthLens – Free multi-signal deepfake image detector GitHub - apexlocal-jz/claude-usage-tray: Windows system-tray app showing your Claude Code rate-limit usage at a glance. Zero deps, ~300 lines of PowerShell. Cross-IDE (works regardless of VS Code, Cursor, plain terminal). Release v0.1.2.1 · kouhxp/yapsnap GitHub - noopolis/moltnet: Self-hostable chat network for AI agents. Pre-built bridges for Claude Code, Codex, and the Claws. Rooms, DMs, history. No Slack bots, no Matrix, no glue code. GitHub - tamerh/enju: Coordinating Humans, AI Agents, and Compute as Peers on a Shared Workflow Graph Show HN: Continuity-auth – Respect-weighted rate limits for the open web GitHub - luml-ai/luml: AI lifecycle platform where engineers and agents track experiments, train models, and ship to production. GitHub - mrdanielcasper/CoreTex: A UNIX-inspired, biomimetic, flat-file AI harness and knowledge engine. GitHub - clemg/pierre-github: Pierre's diffs.com and trees.software for Github GitHub - lyriks-io/unspaghettit: Behavior-driven AI development without prompt spaghetti. GitHub - sofumel/claude-handoff-revive: Resume Claude Code work after rate/usage/context limits without replaying the prior transcript. Auto-saves at 90%/95% usage. Plugin-installable, 10 languages. GitHub - dotexorg/saferpc: Typed, end-to-end encrypted RPC over any bidirectional channel. GitHub - BeeZeeAgent/beezee: Agent harness orchestration Legato Next.js Boilerplate for Internal Tools · CoreUI GitHub - clark-labs-inc/clark-hash: Clark Hash, 32x smaller searchable sketches for embeddings GitHub - ZeroPointRepo/youtube-mcp: The fastest YouTube transcript + YouTube search MCP for AI agents. Try for free. Typing Mastery — climb toward 100+ WPM, deliberately GitHub - Andebugulin/Awareen GitHub - fayzan123/claude-workflow-composer: Visual desktop app for composing multi-agent coding workflows. Drag agents, attach skills and MCPs, wire handoffs, export to .claude/ GitHub - harshaneel/humanize: Best static AI text humanizer. Two research-grounded skills that work in any LLM (Claude, ChatGPT, Gemini, Codex): humanize beats perplexity-based detectors, ai-check produces forensic scoring with evidence-quoted flags. Nine levers, 50+ peer-reviewed sources, 2024-2026 detection literature. GitHub - StackOneHQ/stack-nudge GitHub - nodes-app/swift-markdown-engine: A native AppKit Markdown editor for macOS, built on TextKit 2 and bridged to SwiftUI. We hardened an LLM agent. Each defense we added made it more exploitable. GitHub - alkait/WhatsKept: Agent-queryable WhatsApp history from an iOS backup — a single Go binary. GitHub - octelium/cordium: Open-source, general-purpose sandbox platform for devs and AI agents that provides identity-based secure access to infrastructure without credentials. WAR.GOV/UFO Microfilm5 GitHub - scosman/videowright: Build animated explainer videos with your coding agent GitHub - dipankar/dscode: The code editor you can take apart. GitHub - zoharbabin/web-researcher-mcp: MCP server (Go) for AI assistants: web search, content extraction, academic/patent/news research. Multi-provider routing, 4-tier scraping, search lenses. Works with Claude, Cursor, and any MCP client. GitHub - ruvnet/RuView: π RuView turns commodity WiFi signals into real-time spatial intelligence, vital sign monitoring, and presence detection — all without a single pixel of video. GitHub - scanaislop/aislop: Catch the slop AI coding agents leave in your code: narrative comments, swallowed exceptions, as-any casts, dead code, oversized functions. 50+ rules across 7 languages (TypeScript, JavaScript, Python, Go, Rust, Ruby, PHP). Sub-second, deterministic, no LLM at runtime. MIT-licensed. GitHub - kouhxp/cheap-im: CPU-only voice agent approximating Thinking Machines' Interaction Models demo GitHub - unprovable/OrchidMantis: Orchid Mantis — standalone framework for Zero-Knowledge Proofs of eXploit (ZKPoX). GitHub - MarcellM01/TinySearch: Shrink the web for your local LLMs! GitHub - TangibleResearch/Halgorithem: A Algo designed to detect AI Hallucitions GitHub - DO-SAY-GO/freelang: I love freelang GitHub - CarpseDeam/Aura-IDE: An AI coding harness that shaped itself - Planner/Worker agents, repo awareness, surgical edits, validation, recovery, and safe diff approvals. GitHub - chojs23/concord: A feature-rich TUI client for Discord GitHub - tommyjepsen/awesome-ux-skills: UX & AI Product designs skills you can use today in Claude Code GitHub - aerf-spec/aerf: Agent Evidence Receipt Format (AERF) — an open specification for tamper-evident, independently verifiable records of AI agent actions. GitHub - kklimuk/docx-cli: CLI for AI agents (Claude, Codex) to read, edit, and comment on .docx files with full format fidelity. GitHub - Jwrede/tokentoll: Catch LLM cost changes in code review. Infracost for LLM spend. GitHub - samchon/ttsc: A `typescript-go` toolchain for compiler-powered plugins and type-safe execution + 500x faster lint integrated into compiler GitHub - Higangssh/homebutler: 🏠 Manage your homelab from chat. Single binary, zero dependencies. GitHub - olalie/tapmap: See where your computer connects and what stands out on a live world map. GitHub - Diplomat-ai/diplomat-agent: What can your AI agent do to the real world? Scan your code. See which tool calls have zero checks GitHub - Bajusz15/beacon: Open-source agent for secure remote access, monitoring, and deploys across home-lab and self-hosted machines like Raspberry Pi, N100, or any Linux server. Open web based TTY or tunnel Home Assistant and other local services securely without opening ports. BigTech AI News - Chrome 应用商店 GitHub - vinhnx/VTCode: VT Code is an open-source coding agent with LLM-native code understanding and robust shell safety. Supports multiple LLM providers with automatic failover and efficient context management. GitHub - michaelaz774/decision-engine: A decision operating system for startup founders, powered by Claude Code. Synthesizes wisdom from 25+ legendary founders and investors into interactive AI-driven decision frameworks. GitHub - Chrilleweb/dotenv-diff: Validate environment variable usage in your codebase GitHub - Lumen-Labs/brainapi2: BrainAPI is a knowledge graph–powered AI memory layer that transforms unstructured data into structured knowledge, enabling intelligent search, recommendations, and contextual memory for AI agents and applications. GitHub - familiar-software/familiar: Let AI watch you work. Familiar lets your AI update its memory, skills, and knowledge by watching your screen. GitHub - skorotkiewicz/rudo: A small, elegant dock for Wayland GitHub - muxshed/shed: One stream in, or many. Every destination, simultaneously. No cloud middleman, no per-channel fees, no limits. make sidebar/address bar rounded corner toggleable
Project Huginn — Distributed GPU Sharing for AI Training
SMware ApS · 2026-06-17 · via Show HN

Unify heterogeneous GPUs (2GB–32GB+) into a single compute pool. Train AI models at lower cost with measurable sustainability impact.

Why Choose Hugin?

One network to train your AI — cheaper, private, and built for the physical world.

🔒

Private by design

You choose the protection level on every job: Shield+ keeps your data encrypted and split so no machine ever sees the whole — or Vault runs on sealed hardware no one else can read.

Built for Physical AI

More than fine-tuning. Hugin Learning trains robot, drone and automation control from scratch — our own breakthrough, for the problems others can’t train.

Verified & reliable

Every result is independently re-checked, and a slow or dropped machine never stalls your job. You get a model you can trust — completely hands-off.

What you can build

Built for real AI projects

Bring your data, train production-ready models across the network, then download and use them anywhere.

Custom language models

Fine-tune open models (LLaMA, Mistral, Phi, Gemma, Qwen) on your own data with LoRA & QLoRA.

Chatbots · Assistants · Code · Multilingual

Computer vision models

Train image classification & object-detection models on your own images, then download the trained model.

Quality control · Drones · Cameras · Inspection

Robotics & automation

Build perception models and prepare data — augmentation, validation and benchmarking for robotics teams.

Perception · Dataset prep · Model evaluation

Hugin Learning

Train control policies for robots and agents — the system discovers behaviour from scratch by trial-and-error, no labelled data needed.

Robots · Control · Agents · Automation

How Hugin Works

Six-step pipeline from job definition to verified delivery with transparent billing.

Job Definition

Data classification, target metric, and pool selection.

Cost Estimate

HU & GPU-seconds estimate with upper bound and SLA.

Sharding & Scheduling

Capability-aware distribution, node selection, redundant execution.

Execution

Sandboxed micro-shard execution with telemetry, health checks, auto-retry.

Aggregation & Verification

Quality threshold, spot-check, and anomaly detection validation.

Billing & Payment

Verified job billing, automatic GPU owner payout.

The journey

From your data to your trained model

You upload, we take care of everything, and your finished model is ready to download.

Upload your data

Secure & private

We prepare it

Fully automatic

Trained across our network

Powered at scale

Turned into your model

Brought together

Download the result

Ready to use

Our breakthrough

Hugin Learning

Our own innovation: control intelligence that learns from scratch — by trial-and-error, with no labelled data — trained and verified across the network. Built for robots, drones and automation.

Learns from zero

Discovers the right behaviour by trial-and-error — no demonstrations or labelled data required.

Built for Physical AI

Balancing, positioning, navigation and your own custom tasks — the brain that runs robots, drones and automated lines.

Distributed & verified

Trained across the network and independently re-checked, so the result can be trusted.

⬡ Engineered by Hugin

Your data, protected by design

Most GPU networks ship your raw data to strangers' machines. Hugin doesn't. Every job runs under Hugin Shield+ — a multi-layer protection stack we built — or, for the most sensitive work, inside a hardware-isolated Vault.

🛡️ Hugin Shield+

Included on every job

Protection is built in on every job — at no extra cost. Your data stays private, safe, and completely yours.

  • 🔒Private by default — protected automatically the moment you upload
  • 🙈Yours alone — no one ever sees your full data
  • 🤝Trusted partners only — your work runs on verified, reliable machines
  • 😌Total peace of mind — no setup, no risk, no surprises

🔒 Hugin Vault

Maximum confidentiality

Maximum confidentiality. Your data stays invisible — even to the computer that trains it.

  • 🧠Sealed private space — your data is only opened inside a protected area
  • 🙈Invisible to the owner — the machine’s owner can never see it
  • 📜Proven genuine — every machine is verified before it touches your data
  • 🏥For your most sensitive data — health, finance & personal records

You choose the protection level per job — Shield+ is included at the standard rate; Vault adds hardware isolation for the most sensitive data.

Transparent HU-Based Pricing

HU (Hugin Unit) is your simple billing unit. 1 HU ≈ 3,600 normalized GPU-seconds. Pre-estimate and upper bound before every job.

0.21EUR / HU

Transparent Pricing for Training

  • More iterations with same budget
  • Transparent HU + GPU-seconds billing
  • Pre-estimate & upper bound guaranteed
  • Community, Verified, or Dedicated pools
  • ESG reporting: kWh/HU & CO₂e/HU

Start Training Now

Share Your Compute

Turn idle devices into passive value. Hugin network harnesses compute power from any connected device globally.

  • 📱 Smartphones (Browser or App)
  • 🚗 Electric Vehicles (EV Compute)
  • 🌐 Any Web Browser Tab
  • 💻 Desktop & Server GPUs

Join the Network

HU Rate by Device Class

Device Class~HU/HourExamples
Smartphones (Idle)0.05 - 0.15iPhone 15 Pro, Pixel 8
Electric Vehicles (EV)0.2 - 0.4Tesla MCU, Polestar 2
4-6GB Consumer0.4 - 0.6GTX 1650, RTX 3050
8-12GB Consumer0.6 - 0.9RTX 3060, RTX 4070
16-24GB Workstation1.6 - 2.8RTX 3090, RTX 4090
24-48GB+ Datacenter2.8 - 4.5A100, H100

Frequently Asked Questions

Everything you need to know about training AI models on Huginn.

What is Project Huginn?+

Project Huginn is a distributed GPU sharing platform that lets you train and fine-tune AI models at up to 50% lower cost than traditional cloud providers. It pools heterogeneous GPUs (from RTX 3080 to H100) into a unified compute grid. Built by SMware ApS, a Startup Denmark approved company.

How much does it cost to train an AI model on Huginn?+

Huginn uses Hugin Units (HU) for billing. 1 HU = €0.21. A typical LoRA fine-tuning of a 7B model with 1,000 training examples costs approximately 5-8 HU (€1-2). Costs are estimated before you start, with an upper bound guarantee — no surprise charges.

What AI models can I fine-tune?+

Huginn provides 7 base models for free: LLaMA 3.1 (8B & 70B), Mistral 7B, Phi-3 Mini 3.8B, CodeLlama 7B, Gemma 2 9B, and Qwen 2.5 7B. You select a base model and upload your own training data — Huginn handles the rest.

What training methods does Huginn support?+

Three methods: LoRA (Low-Rank Adaptation) for fast, efficient training with ~12GB VRAM; QLoRA (4-bit quantized) for even lower memory usage with ~8GB VRAM; and Full Fine-Tune for highest quality with ~40GB+ VRAM. LoRA with rank 16 is recommended for most use cases.

What data formats are supported for training?+

JSONL (Alpaca format with instruction/input/output fields), CSV (with instruction, input, output columns), and Chat Format JSONL (with messages array). Maximum upload size is 500MB. We recommend at least 1,000 examples for good results.

How is Huginn different from other GPU cloud platforms?+

Unlike traditional GPU rental platforms that only offer raw compute, Huginn provides an all-in-one AI training experience. You get a built-in Model Studio with 7 free base models, an integrated fine-tuning pipeline, automatic cost estimation with upper bounds, and a ChatGPT-style playground to test your model instantly — no setup, no configuration, just upload your data and start training.

Can I earn money by sharing my device?+

Yes! Huginn lets you earn from virtually any device. Share your desktop GPU, laptop, tablet, smartphone, or even your electric vehicle's compute power while you drive. Connect via the Huginn Agent app or simply open a browser tab. Every connected device contributes to the network and earns you HU. Desktop GPUs earn the most (up to €0.60/hour for high-end cards), but even a smartphone or browser tab generates passive income 24/7.

How do I share my GPU and start earning?+

It takes under 5 minutes: 1) Sign in with Google and choose GPU Owner role, 2) Go to GPU Agent page in your dashboard, 3) Download the agent file and generate a token, 4) Run the command in your terminal. The agent auto-detects your GPU and starts processing jobs. You can also use the browser-based Web Engine with zero downloads. Read our full guide at projecthuginn.com/how-to-share for step-by-step instructions.

Is my training data secure?+

Yes. All training jobs run in sandboxed environments with complete data isolation. Each user's models and datasets are private. We use session-based authentication and encrypted connections for all data transfers.

Can I download my trained model?+

Yes. After training completes, you can download your fine-tuned model in .safetensors format. You can run it locally with Ollama, llama.cpp, vLLM, or HuggingFace Transformers. You can also keep it on Huginn and test it via the built-in playground.

What GPU tiers are available?+

Five tiers: T1 (H100/A100 80GB+, 4.50 HU/hr), T2 (A100 40GB/RTX 4090, 2.80 HU/hr), T3 (RTX 4080/3090, 1.60 HU/hr), T4 (RTX 3080/2080 Ti, 0.90 HU/hr), and T5 (entry-level, 0.40 HU/hr). Higher tiers are faster but cost more per hour.