Source codes
This page lists the open source codes created during the course of the Embed2Scale project:
NeuCo-Bench
NeuCo-Bench is a benchmarking framework designed to evaluate how effectively compact, fixed-size embeddings preserve information for downstream tasks. The repo contains the code for users to evaluate embeddings, alongside code and notebooks that show examples of how to generate and work with embeddings.
TerraTorch Embedding Extension
This TerraTorch extension is designed to enable Earth Observation (EO) embedding workflows. It adds low‑code and no‑code tools to generate, analyse, and reuse EO embeddings across a wide range of downstream tasks.
TerraCodec
TerraCodec introduces neural compression models tailored for Earth observation workflows.
OpenEO-Slurm
OpenEO components are designed to be portable, but deploying them on an HPC/HPDA system such as terrabyte requires specific adaptations to ensure compatibility with the underlying infrastructure. Terrabyte uses SLURM as its job scheduling system. To integrate openEO workflows with this environment, the standard openeo-argoworkflows component is adapted to support SLURM‑based execution. It modifies the workflow execution layer: tasks are submitted as SLURM jobs instead of relying on default container orchestration mechanisms.
openeo-processes-dask-ml
openeo-processes-dask-ml is a Python package that implements generic machine learning processes for openEO. It is built to integrate with openeo-processes-dask, extending it by machine learning capabilities.
AI for Good Tutorial 2026
This repository builds on the TerraTorch and NeuCo-Bench work, and shows examples of working with and building embedding workflows. Alongside the examples provided in the original repos, this is a good entry point for people who want to start working with our code.