<img src="https://github.com/user-attachments/assets/f8c9586f-6220-4a53-9669-2aee3300b492" alt="TerraTorch" width="400"/>
## Overview
TerraTorch is a library based on [PyTorch Lightning](https://lightning.ai/docs/pytorch/stable/) and the [TorchGeo](https://github.com/microsoft/torchgeo) domain library
for geospatial data.
TerraTorch’s main purpose is to provide a flexible fine-tuning framework for Geospatial Foundation Models, which can be interacted with at different abstraction levels. The library provides:
- Convenient modelling tools:
- Flexible trainers for Image Segmentation, Classification and Pixel Wise Regression fine-tuning tasks
- Model factories that allow to easily combine backbones and decoders for different tasks
- Ready-to-go datasets and datamodules that require only to point to your data with no need of creating new custom classes
- Launching of fine-tuning tasks through CLI and flexible configuration files, or via jupyter notebooks
- Easy access to:
- Open source pre-trained Geospatial Foundation Model backbones:
* [Prithvi](https://huggingface.co/ibm-nasa-geospatial/Prithvi-100M)
* [SatMAE](https://sustainlab-group.github.io/SatMAE/)
* [ScaleMAE](https://github.com/bair-climate-initiative/scale-mae)
* Satlas (as implemented in [TorchGeo](https://github.com/microsoft/torchgeo))
* DOFA (as implemented in [TorchGeo](https://github.com/microsoft/torchgeo))
* SSL4EO-L and SSL4EO-S12 models (as implemented in [TorchGeo](https://github.com/microsoft/torchgeo))
* [Clay](https://github.com/Clay-foundation/model)
- Backbones available in the [timm](https://github.com/huggingface/pytorch-image-models) (Pytorch image models)
- Decoders available in [SMP](https://github.com/qubvel/segmentation_models.pytorch) (Pytorch Segmentation models with pre-training backbones) and [mmsegmentation](https://github.com/open-mmlab/mmsegmentation) packages
- Fine-tuned models such as [granite-geospatial-biomass](https://huggingface.co/ibm-granite/granite-geospatial-biomass)
- All GEO-Bench datasets and datamodules
- All [TorchGeo](https://github.com/microsoft/torchgeo) datasets and datamodules
## Install
### Pip
In order to use th file `pyproject.toml` it is necessary to guarantee `pip>=21.8`. If necessary upgrade `pip` using `python -m pip install --upgrade pip`.
[comment]: <For a stable point-release, use `pip install terratorch`.>
[comment]: <If you prefer to get the most recent version of the main branch, install the library with `pip install git+https://github.com/IBM/terratorch.git`.>
To get the most recent version of the main branch, install the library with `pip install git+https://github.com/IBM/terratorch.git`.
[comment]: <Another alternative is to install using [pipx](https://github.com/pypa/pipx) via `pipx install terratorch`, which creates an isolated environment and allows the user to run the application as a common CLI tool, with no need of installing dependencies or activating environments.>
TerraTorch requires gdal to be installed, which can be quite a complex process. If you don't have GDAL set up on your system, we reccomend using a conda environment and installing it with `conda install -c conda-forge gdal`.
To install as a developer (e.g. to extend the library):
```
git clone https://github.com/IBM/terratorch.git
cd terratorch
pip install -r requirements/required.txt -r requirements/dev.txt
conda install -c conda-forge gdal
pip install -e .
```
To install terratorch with partial (work in development) support for Weather Foundation Models, `pip install -e .[wxc]`, which currently works just for `Python >= 3.11`.
## Documentation
To get started, check out the [quick start guide](https://ibm.github.io/terratorch/quick_start).
Developers, check out the [architecture overview](https://ibm.github.io/terratorch/architecture).
## Contributing
This project welcomes contributions and suggestions.
A simple hint for any contributor. If you want to meet the GitHub DCO checks, just do your commits as below:
```
git commit -s -m <message>
```
It will sign the commit with your ID and the check will be met.
Raw data
{
"_id": null,
"home_page": null,
"name": "terratorch",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.10",
"maintainer_email": null,
"keywords": "fine-tuning, geospatial foundation models, artificial intelligence",
"author": "Romeo Kienzler, Benedikt Blumenstiel",
"author_email": "Carlos Gomes <carlos.gomes@ibm.com>, Joao Lucas de Sousa Almeida <joao.lucas.sousa.almeida@ibm.com>",
"download_url": "https://files.pythonhosted.org/packages/e6/c5/5e57c0642a33d7c7401c0584edbc71dc652825f94e91b810213944b642b2/terratorch-0.99.8.tar.gz",
"platform": null,
"description": "\n<img src=\"https://github.com/user-attachments/assets/f8c9586f-6220-4a53-9669-2aee3300b492\" alt=\"TerraTorch\" width=\"400\"/>\n\n## Overview\nTerraTorch is a library based on [PyTorch Lightning](https://lightning.ai/docs/pytorch/stable/) and the [TorchGeo](https://github.com/microsoft/torchgeo) domain library\nfor geospatial data. \n\nTerraTorch\u2019s main purpose is to provide a flexible fine-tuning framework for Geospatial Foundation Models, which can be interacted with at different abstraction levels. The library provides:\n\n- Convenient modelling tools:\n - Flexible trainers for Image Segmentation, Classification and Pixel Wise Regression fine-tuning tasks\n - Model factories that allow to easily combine backbones and decoders for different tasks\n - Ready-to-go datasets and datamodules that require only to point to your data with no need of creating new custom classes\n - Launching of fine-tuning tasks through CLI and flexible configuration files, or via jupyter notebooks\n- Easy access to:\n - Open source pre-trained Geospatial Foundation Model backbones:\n * [Prithvi](https://huggingface.co/ibm-nasa-geospatial/Prithvi-100M)\n * [SatMAE](https://sustainlab-group.github.io/SatMAE/)\n * [ScaleMAE](https://github.com/bair-climate-initiative/scale-mae)\n * Satlas (as implemented in [TorchGeo](https://github.com/microsoft/torchgeo))\n * DOFA (as implemented in [TorchGeo](https://github.com/microsoft/torchgeo))\n * SSL4EO-L and SSL4EO-S12 models (as implemented in [TorchGeo](https://github.com/microsoft/torchgeo))\n * [Clay](https://github.com/Clay-foundation/model)\n - Backbones available in the [timm](https://github.com/huggingface/pytorch-image-models)\u00a0(Pytorch image models)\n - Decoders available in [SMP](https://github.com/qubvel/segmentation_models.pytorch) (Pytorch Segmentation models with pre-training backbones)\u00a0and [mmsegmentation](https://github.com/open-mmlab/mmsegmentation) packages\n - Fine-tuned models such as [granite-geospatial-biomass](https://huggingface.co/ibm-granite/granite-geospatial-biomass)\n - All GEO-Bench datasets and datamodules\n - All [TorchGeo](https://github.com/microsoft/torchgeo) datasets and datamodules \n\n## Install\n### Pip\nIn order to use th file `pyproject.toml` it is necessary to guarantee `pip>=21.8`. If necessary upgrade `pip` using `python -m pip install --upgrade pip`. \n\n[comment]: <For a stable point-release, use `pip install terratorch`.>\n[comment]: <If you prefer to get the most recent version of the main branch, install the library with `pip install git+https://github.com/IBM/terratorch.git`.>\nTo get the most recent version of the main branch, install the library with `pip install git+https://github.com/IBM/terratorch.git`.\n\n[comment]: <Another alternative is to install using [pipx](https://github.com/pypa/pipx) via `pipx install terratorch`, which creates an isolated environment and allows the user to run the application as a common CLI tool, with no need of installing dependencies or activating environments.>\n\nTerraTorch requires gdal to be installed, which can be quite a complex process. If you don't have GDAL set up on your system, we reccomend using a conda environment and installing it with `conda install -c conda-forge gdal`.\n\nTo install as a developer (e.g. to extend the library):\n```\ngit clone https://github.com/IBM/terratorch.git\ncd terratorch\npip install -r requirements/required.txt -r requirements/dev.txt\nconda install -c conda-forge gdal\npip install -e .\n```\n\nTo install terratorch with partial (work in development) support for Weather Foundation Models, `pip install -e .[wxc]`, which currently works just for `Python >= 3.11`. \n\n## Documentation\n\nTo get started, check out the [quick start guide](https://ibm.github.io/terratorch/quick_start).\n\nDevelopers, check out the [architecture overview](https://ibm.github.io/terratorch/architecture).\n\n## Contributing\n\nThis project welcomes contributions and suggestions.\n\nA simple hint for any contributor. If you want to meet the GitHub DCO checks, just do your commits as below:\n```\ngit commit -s -m <message>\n```\nIt will sign the commit with your ID and the check will be met. \n",
"bugtrack_url": null,
"license": "Apache License, Version 2.0",
"summary": "TerraTorch - A model training toolkit for geospatial tasks",
"version": "0.99.8",
"project_urls": {
"Documentation": "https://github.com/IBM/terratorch#readme",
"Issues": "https://github.com/IBM/terratorch/issues",
"Source": "https://github.com/IBM/terratorch"
},
"split_keywords": [
"fine-tuning",
" geospatial foundation models",
" artificial intelligence"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "f3fcfee3e1230d49346caea45de03fd82f303d82f757b93d1678fa833e40bb08",
"md5": "76255412b66d9639b0a06842d87875c4",
"sha256": "a22ce4db9ee2714a8497cfefe8b8660a31ab28ab7543f06f290dc7b7ea995620"
},
"downloads": -1,
"filename": "terratorch-0.99.8-py3-none-any.whl",
"has_sig": false,
"md5_digest": "76255412b66d9639b0a06842d87875c4",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.10",
"size": 317649,
"upload_time": "2025-02-05T18:12:24",
"upload_time_iso_8601": "2025-02-05T18:12:24.928952Z",
"url": "https://files.pythonhosted.org/packages/f3/fc/fee3e1230d49346caea45de03fd82f303d82f757b93d1678fa833e40bb08/terratorch-0.99.8-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "e6c55e57c0642a33d7c7401c0584edbc71dc652825f94e91b810213944b642b2",
"md5": "33df50e56f35081b3093d2e9864e7f12",
"sha256": "93651df7cbc2df5e0deb54dc9330ae9c4778432b5d44291f16e4799540553c38"
},
"downloads": -1,
"filename": "terratorch-0.99.8.tar.gz",
"has_sig": false,
"md5_digest": "33df50e56f35081b3093d2e9864e7f12",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.10",
"size": 221656,
"upload_time": "2025-02-05T18:12:27",
"upload_time_iso_8601": "2025-02-05T18:12:27.466065Z",
"url": "https://files.pythonhosted.org/packages/e6/c5/5e57c0642a33d7c7401c0584edbc71dc652825f94e91b810213944b642b2/terratorch-0.99.8.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-02-05 18:12:27",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "IBM",
"github_project": "terratorch#readme",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "terratorch"
}
JSON
