Skip to content

probcomp/b3d

Folders and files

NameName
Last commit message
Last commit date
Jun 27, 2024
Jul 23, 2024
Jul 19, 2024
Jul 23, 2024
Jul 23, 2024
Jul 23, 2024
Jul 23, 2024
Jul 19, 2024
Jul 23, 2024
Jul 23, 2024
Jul 19, 2024
Jul 23, 2024
Jul 23, 2024
Jul 19, 2024
Jul 23, 2024
Jul 19, 2024

Repository files navigation

b3d

animated

b3d

This repository contains code for Bayesian 3D inverse graphics.

The b3d/bayes3d/ subdirectory contains code for the bayes3d project, and the b3d/chisight/ subdirectory contains code for post-bayes3D ChiSight systems (currently, SAMA4D).

GCloud Machine

Run the following command to launch your instance:

export ZONE="us-west1-a"

# Make sure to replace the value with a unique name!
export INSTANCE_NAME="your-instance-name-here"

gcloud compute instances create $INSTANCE_NAME \
  --zone=$ZONE \
  --image-family="common-cu123-ubuntu-2204-py310" \
  --image-project=deeplearning-platform-release \
  --maintenance-policy=TERMINATE \
  --boot-disk-size=400GB \
  --machine-type g2-standard-32 \
  --accelerator="type=nvidia-l4,count=1" \
  --metadata="install-nvidia-driver=True"

to ssh into your new machine, run:

gcloud compute config-ssh
ssh $INSTANCE_NAME.$ZONE.$PROJECT_ID

Environment Variables

Add the following to your bash_rc:

export XLA_PYTHON_CLIENT_PREALLOCATE="false"
export XLA_PYTHON_CLIENT_ALLOCATOR="platform"

Installing b3d

b3d's GenJAX dependency is hosted in Google Artifact Registry. To configure your machine to access the package:

Next, on the machine, close the b3d repo:

git clone https://github.com/probcomp/b3d.git
cd b3d

Create and activate b3d conda environment:

conda create -n b3d python=3.12

When that completes, run the install script:

bash -i install.sh

Then activate the conda environment:

conda activate b3d

Visualizer

Tunnel port 8812 for Rerun visualization by adding the RemoteForwardline to your ssh config:

Host xx.xx.xxx.xxx
    HostName xx.xx.xxx.xxx
    IdentityFile ~/.ssh/id_rsa
    User thomasbayes
    RemoteForward 8812 127.0.0.1:8812

Install rerun on local machine pip install rerun-sdk and open viewer:

rerun --port 8812