README.md

metadata

title: Primate Batch Detection
emoji: 💻
colorFrom: purple
colorTo: gray
sdk: gradio
sdk_version: 5.46.0
app_file: app_batch.py
pinned: false
license: apache-2.0

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

Primate Batch Detection: Installation & Troubleshooting Guide

Installation Guide

Prerequisites

• Computer with GPU, this guide has been tested on an NVIDIA 4090 and 4060.
• 20-50 min of time

General Tips

• Check the troubleshooting guide at the bottom of the page
• If you have already set up WSL, Docker, or NVIDIA Container Toolkit before, feel free to skip those steps.
• Use Google / ChatGPT to look up your errors
• Reread the instructions carefully
• Proceed forward anyways and see if it causes an issue
• Anytime a terminal command with $VARIABLE_NAME or <VARIABLE_NAME>, it means that you should substitute in your appropriate variable or username, without the $ or < decorators.
• Try opening and closing the terminal window you're in, or restart your computer.

---

1. Set Up WSL (Windows Users Only)

If you're using Windows, you'll need to install the Windows Subsystem for Linux (WSL):

We recommend using Ubuntu 22.04 as your Linux distribution.

1. Open Powershell and install the WSL command:

wsl --install Ubuntu

2. Set the default Linux distribution used with WSL:

wsl.exe --set-default Ubuntu

for more information visit: WSL Installation Guide (Microsoft)

Now open the WSL terminal through your Start menu.

Important: Run all following commands in the WSL terminal, not PowerShell.

---

2. Install Docker (Mac and Linux Users Start Here)

Follow these guides to install Docker in your WSL Ubuntu environment:

1. Follow this guide using "Prerequisites" and "Install Docker Desktop":

• Install Docker Desktop

2. Through the previous link, you should have followed step 1 of "Install using the apt repository" as part of your installation process. Now complete steps 2-3 of that section as well:

• Install Docker Engine

---

3. Install the NVIDIA Container Toolkit

This is necessary for GPU support inside Docker containers:

Note: Most likely, unless you built your own computer recently, you will already have installed NVIDIA drivers for your computer. Feel free to ignore the Prerequisites section.

1. Follow the "With apt: Ubuntu, Debian" section, and the "Configuring Docker" section of the "Configuration" section.

NVIDIA Container Toolkit Installation Guide

2. Verify that you've installed correctly using this command:

sudo docker run --rm --gpus all nvidia/cuda:11.0.3-base-ubuntu20.04 nvidia-smi

You should see something that looks like this pop up if successful.

If not, you should restart the NVIDIA Container Toolkit Installation Guide

---

4. Run the Docker Container

Use the pre-configured Docker image provided by the Hugging Face Space:

Run Docker Container

If you run into the "RuntimeError: Found no NVIDIA driver on your system" error, run this command instead.

sudo docker run --rm -it --gpus all -p 7860:7860 registry.hf.space/caltech-animal-tracking-primate-batch-detection:latest python app_batch.py

Note: The first time you've installed the interface on a computer, it will take a while to download, but subsequent startups, just follow steps 4 and 5, and it should be relatively fast to start.

5. Launch the interface

Either click on the local link that is given in your WSL terminal, or go to this link in your web browser.

http:/localhost:7860

---

Troubleshooting Guide

🚫 Error:

The WSL window crashes immediately upon launch.

🛠️ Fix:

Double-check that you've done wsl --set-default Ubuntu in Powershell.

🚫 Error:

docker: Error response from daemon: could not select device driver "" with capabilities: [[gpu]]

🛠️ Fix:

This typically means that the NVIDIA Container Toolkit was not installed or configured correctly. Double-check your installation by revisiting the NVIDIA Container Toolkit Guide.

🚫 Error:

docker: permission denied while trying to connect to the Docker daemon socket at unix:///var/run/docker.sock:
Head "http://%2Fvar%2Frun%2Fdocker.sock/_ping": dial unix /var/run/docker.sock: connect: permission denied

🛠️ Fix:

This error occurs when your user does not have permission to access the Docker socket.

You have two options:

Option 1: Add your user to the Docker group

sudo usermod -aG docker $USER
sudo systemctl restart docker  # May not be needed or available in WSL

Then restart your terminal or log out and back in for the group change to take effect.

Option 2: Use sudo for all your docker commands

sudo docker $YOUR_COMMAND