Set up a local model

Before you start

Set up LiquidAI d1 on a desktop Station. Installation verifies the downloads and runs a video test before the model is ready.

Pair a desktop Station and keep it online during setup. The model files total about 2.05 GB. Allow extra disk space for the runtime and temporary files; setup checks the space needed for your computer.

Set it up

1. Allow setup in the Windows installer or during pairing on macOS or Linux. To allow it later, use the Station menu on Windows or macOS, or the Linux command below.
2. Station then installs the model by itself. It uses the files it already found, for example in LM Studio, or downloads the verified files, and runs on an NVIDIA GPU if it has one. You can add cameras meanwhile: they start analysing when the model is ready.
3. To follow the progress, open Settings → Stations and choose your Station’s local model. Before setup starts, Change lets you choose the files and whether it runs on the GPU or the processor.
4. Station verifies every file and runs a video test. When it passes, the model is ready and selected. Cameras on that Station use it, unless a camera has its own AI provider setting.

Linux

The Linux Station runs as a service without a menu. Pairing with sudo agentinosaurus-station pair asks whether to set up the local model. To allow it later, run this once:

sudo agentinosaurus-station allow-local-model

Use files you already have

Station looks in Downloads, LM Studio and the Hugging Face cache. If your files are elsewhere, choose Open local models folder in the Station menu, copy both files there, then choose Look again in setup.

Each model needs its exact model file and vision file, as named in setup. Other versions or quantizations are not accepted.

If something goes wrong

Setup didn’t start by itself: Station waits after you pause it, and on computers with less than 6 GB of memory or too little free disk space. Setup says which, and Install starts it anyway.

Not enough disk space: free some space on the Station computer and choose Resume setup. Parts already downloaded are kept.

The download stopped: check the internet connection on that computer and resume.

The video test failed: check available memory and update Station to include FFmpeg and ffprobe. Try the processor if the GPU cannot load the model.

The runtime couldn’t start on the GPU: update the NVIDIA driver, or choose the processor.

The model didn’t start after a restart: Station retries automatically. If the files were moved or changed, set the model up again.

Switch back or remove it

Choose another provider in the Station’s AI provider menu at any time. Disconnect model stops using it but keeps its files, so you can delete them yourself or set the model up again without downloading.