deepvisiontools 0.2.0__tar.gz → 1.0.0__tar.gz

{deepvisiontools-0.2.0 → deepvisiontools-1.0.0}/PKG-INFO

@@ -1,8 +1,8 @@
 Metadata-Version: 2.4
 Name: deepvisiontools
-Version: 0.2.0
-Summary: Overlay of PyTorch to generalize trainning and inference processes for detection, instance segmentation and semantic segmentation tasks.
-Author-email: Lucas Bernigaud Samatan <lucas.bernigaud-samatan@inrae.fr>, Jordan Bernigaud Samatan <jordan.bernigaud-samatan@inrae.fr>, Paul Faucher <paul.faucher@inrae.fr>
+Version: 1.0.0
+Summary: deepvisiontools is a library built on torch for training and inference processes for deep learning computer vision tasks.
+Author-email: Lucas Bernigaud Samatan <lucas.bernigaud-samatan@inrae.fr>, Jordan Bernigaud Samatan <jordan.bernigaud-samatan@inrae.fr>
 Maintainer-email: Jordan Bernigaud Samatan <jordan.bernigaud-samatan@inrae.fr>, Lucas Bernigaud Samatan <lucas.bernigaud-samatan@inrae.fr>
 License:                     GNU AFFERO GENERAL PUBLIC LICENSE
                                Version 3, 19 November 2007
@@ -666,65 +666,250 @@ License:                     GNU AFFERO GENERAL PUBLIC LICENSE
         For more information on this, and how to apply and follow the GNU AGPL, see
         <https://www.gnu.org/licenses/>.
         
-Project-URL: Homepage, https://forgemia.inra.fr/ue-apc/librairies/python/deepvisiontools_v1
-Project-URL: Issues, https://forgemia.inra.fr/ue-apc/librairies/python/deepvisiontools_v1/-/issues
-Project-URL: Documentation, https://deepvisiontools.readthedocs.io/en/latest/
-Project-URL: Repository, https://forgemia.inra.fr/ue-apc/librairies/python/deepvisiontools_v1
-Classifier: Development Status :: 4 - Beta
+Project-URL: Homepage, https://forge.inrae.fr/ue-apc/librairies/python/deepvisiontools
+Project-URL: Issues, https://forge.inrae.fr/ue-apc/librairies/python/deepvisiontools/-/issues
+Project-URL: Repository, https://forge.inrae.fr/ue-apc/librairies/python/deepvisiontools
+Project-URL: Documentation, https://deepvisiontools-1233bd.pages-forge.inrae.fr/
+Classifier: Development Status :: 5 - Production/Stable
+Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
-Requires-Python: <=3.12.12,>=3.11.9
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: 3.14
+Requires-Python: <3.15,>=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE.txt
-Requires-Dist: torch<=2.6.0,>=2.4.0
-Requires-Dist: torchvision<=0.21.0,>=0.19.0
-Requires-Dist: pycocotools==2.0.8
-Requires-Dist: torchmetrics<=1.7.0,>=1.4.0
-Requires-Dist: tensorboard<=2.19.0,>=2.18.0
-Requires-Dist: ultralytics<=8.3.95,>=8.3.28
-Requires-Dist: transformers<=4.50.0,>=4.44.2
-Requires-Dist: colorcet==3.1.0
-Requires-Dist: segmentation-models-pytorch<=0.5.0,>0.4.0
-Requires-Dist: openpyxl==3.1.5
+Requires-Dist: torch<3.0,>=2.4.0
+Requires-Dist: torchvision<1.0,>=0.19.0
+Requires-Dist: pycocotools<3.0,>=2.0.8
+Requires-Dist: torchmetrics<2.0,>=1.4.0
+Requires-Dist: tensorboard<3.0,>=2.4.1
+Requires-Dist: torch-tb-profiler<1.0,>=0.4.0
+Requires-Dist: mlflow<4.0,>=3.1.0
+Requires-Dist: fastapi!=0.136.3,<1.0
+Requires-Dist: starlette<2.0,>=1.0.1
+Requires-Dist: ultralytics<9.0,>=8.3.28
+Requires-Dist: transformers<6.0,>=5.0.0
+Requires-Dist: colorcet<4.0,>=3.1.0
+Requires-Dist: segmentation-models-pytorch<1.0,>=0.4.1
+Requires-Dist: openpyxl<4.0,>=3.1.5
+Requires-Dist: optuna<5.0,>=4.0.0
+Requires-Dist: optuna-dashboard<1.0,>=0.20.0
+Requires-Dist: huggingface_hub<2.0,>=1.5.0
+Requires-Dist: pillow<13.0,>=12.2.0
+Requires-Dist: setuptools<82,>=78.1.1
+Provides-Extra: dev
+Requires-Dist: pytest<10.0,>=9.0; extra == "dev"
+Requires-Dist: pytest-cov<8.0,>=7.0; extra == "dev"
+Requires-Dist: coverage[toml]<8.0,>=7.0; extra == "dev"
+Requires-Dist: hypothesis<7.0,>=6.0; extra == "dev"
+Requires-Dist: ruff<1.0,>=0.13.0; extra == "dev"
+Requires-Dist: mypy<2.0,>=1.0; extra == "dev"
+Requires-Dist: bandit[toml]<2.0,>=1.7; extra == "dev"
+Requires-Dist: pip-audit<3.0,>=2.7; extra == "dev"
+Requires-Dist: nox[pbs]<2027.0.0,>=2025.0.0; extra == "dev"
+Requires-Dist: build<2.0,>=1.2; extra == "dev"
+Requires-Dist: twine<7.0,>=6.0; extra == "dev"
+Requires-Dist: types-PyYAML<7.0,>=6.0.12; extra == "dev"
+Provides-Extra: internimage-build
+Requires-Dist: huggingface_hub<1.0,>=0.24.0; extra == "internimage-build"
 Dynamic: license-file
 
-<!--
 <p align="center">
-  <img src="https://github.com///raw/main/docs/source/logo.png" height="150">
+  <img align="center" src="supplementary/logo-dvt.png" height="100">
 </p>
--->
 
-<h1 align="center">
-  deepvisiontools
-</h1>
+# Welcome to deepvisiontools!
 
+**deepvisiontools** is a library built on torch for training and inference processes for deep learning computer vision tasks.
 
-Torch overlay for trainning and inference processes for object detection & instance segmentation tasks. 
+<p align="center">
+  <img align="center" src="supplementary/coco_visu.png" height="130"> &nbsp; &nbsp; &nbsp; <img align="center" src="supplementary/veg_visu.png" height="130"> &nbsp; &nbsp; &nbsp; <img align="center" src="supplementary/dice_visu.png" height="130">
+</p>
+
+## News
+
+- *release 1.0.0: detector / detector-seg / tabdet / tabdet-seg model wrappers are heavy but powerful models for detection and instance segmentation. However, they are not fully pretrained (only backbone on imagenet). Therefore, on small datasets (< 1000 images ish, but depend on the dataset ...) they can be slightly less competitive than yolo models (also included in the library) that were fully pretrained on coco. We are currently pretraining some of the models but it takes time. We are goind to update the weights once done, so keep tuned for new releases. We suggest you always train a yolo model as well for a baseline comparison. However, on larger datasets they can significantly outperform yolo11x, but again it depends on the dataset ...*
+
+## Scope
+
+deepvisiontools is suitable for supervised learning computer vision tasks. Tasks currently available are:
+
+- detection (bounding boxes)
+- instance segmentation (object instance masks)
+- semantic segmentation (semantic masks)
+
+Future releases will include:
+
+- classification
+- regression
+- keypoints/pose detection
+- crowd detection (point detection)
+
+deepvisiontools provides integrated pipelines that include dataset handlers, model wrappers, training and hyper-parameter fine-tuning helpers, metrics, inference and evaluation tools.
 
-## 💪 Context
+deepvisiontools is designed for both **non-expert and expert** users by providing default operational frameworks that are easy to use, while keeping maximal flexibility and control for experts.
 
-deepvisiontools is developped by INRAE (french National Research Institute for Agriculture, Food and the Environment) and in [PHENOME-EMPHASIS project](https://www.phenome-emphasis.fr/).
+
+## Context
+
+deepvisiontools is developed by INRAE, the French National Research Institute for Agriculture, Food and the Environment. It was partly supported by the [PHENOME-EMPHASIS project](https://www.phenome-emphasis.fr/). It is associated with INRAE CATI CONNECT.
 
 ## Citations
 
-deepvisiontools provides model wrappers that allow you to use differents models with all machinery available. It is the responsability of the user to cite properly the actual model developpers.
-In particular, if using one of these models : Yolo, SMP please cite the actual models developpers (which are not deepvisiontools). For instance for Yolo, please cite ultralytics (https://www.ultralytics.com/) or for SMP cite https://github.com/qubvel-org/segmentation_models.pytorch  
-Similar stands for other models.
-TimmYolo model is a new model that combines timm encoders with a Yolo detection head. The strategy is based on patchification for the backbone (to avoid loosing resolution in resizing) then merge the features before running a Yolo detection heads. If you use this wrapper please cite both timm library, ultralytics library as well as deepvisiontools as model developper.
+To cite deepvisiontools when used, please cite the official repository: https://forge.inrae.fr/ue-apc/librairies/python/deepvisiontools
+
+**Important**: deepvisiontools provides model wrappers and does not reimplement all models. It is the responsibility of the user to also cite the model developers.
+
+If you are using one of these models, please cite the corresponding library as well:
+
+- `"yolo"`: cite Ultralytics (https://www.ultralytics.com/)
+- `"smp"`: cite segmentation-models-pytorch (https://github.com/qubvel-org/segmentation_models.pytorch)
 
-## Documentation
+Note that detector and tabdet models (VFNet, Cascade R-CNN/Mask R-CNN) are deepvisiontools reimplementations using PyTorch Image Models as backbones (https://timm.fast.ai/). They were reimplemented following MMDetection (https://github.com/open-mmlab/mmdetection) guidelines.
 
-All documentation about deepvisiontools, including tutorials, can be found at : https://deepvisiontools.readthedocs.io/en/latest/
+Finally, hyper-parameter optimisations were built on Optuna. If you use the OptunaWrapper helper, it is worth mentioning Optuna in your work (https://optuna.org/).
 
 ## Installation
-We have tested deepvisiontools for python versions between 3.11.9 and 3.12.11, we recommend to use python 3.12.11 to minimize the risk of bugs.
 
-```shell
+It is recommended to install deepvisiontools on a GPU-enabled machine/server. We suggest that you use a Python virtual environment (uv, conda, venv, etc.) with **Python version between 3.10 and 3.14**.
+
+deepvisiontools is available on PyPI, so you can simply run in a Python environment:
+
+```bash
 pip install deepvisiontools
 ```
-If you wish to use TimmYolo model, you may encounter an error saying that feature_only is not available for a given encoder name. Please upgrade your timm version (1.0.15 should have most of them). You may need to reinstall deepvisiontools as you could create conflicts with segmentation-models-pytorch.
+
+To install deepvisiontools in development mode, please clone the repository and run:
+
+```bash
+pip install -e deepvisiontools/ # local cloned repo
+```
+
+Finally, to install deepvisiontools with developer extras, move inside the cloned repo and run:
+
+```bash
+pip install -e ".[dev]"
+```
+
+This will include extras like documentation, pytest, nox, etc.
+
+## Documentation and tutorials
+
+Official documentation for deepvisiontools can be found here: https://deepvisiontools-1233bd.pages-forge.inrae.fr/
+
+To get started with deepvisiontools, after installation you can run:
+
+```bash
+deepvisiontools download-tutorials
+```
+
+You will also need to download the demo datasets:
+
+```bash
+deepvisiontools download-demo-datasets
+```
+
+The first tutorial, **getting_started**, takes about 1 hour and explains all the basics of deepvisiontools. From there you have all information to get started with simple tasks.
+
+We then recommend moving to **advanced_usage** to get a broader view of what you can configure in deepvisiontools and all available helpers.
+
+Finally, the **hyper_parametrs_optimization** tutorial will explain the usage of Optuna wrappers available in deepvisiontools to perform extensive parameter space searches.
+
+## Monitoring via MLflow
+
+deepvisiontools includes MLflow for monitoring, checkpointing and artifacts during training.
+
+If you use deepvisiontools as intended, at some point you will end up with an `mlruns` folder.
+
+You can use the following command to start MLflow and then open the local page in your browser to display the MLflow UI:
+
+```bash
+mlflow server --backend-store-uri sqlite:////$(pwd)/mlruns/mlflow.db --default-artifact-root file:$(pwd)/mlruns --host 127.0.0.1 --port 5000
+```
+
+An alternative way for Jupyter notebooks is presented in the tutorial.
+
 ## License
 
 GNU Affero General Public License v3 or later (AGPLv3+) (AGPL-3.0).
 
+## Contributor specifics
+
+If you wish to contribute as a developer to deepvisiontools, you need to be part of INRAE or PHENOME-EMPHASIS. You can contact jordan.bernigaud-samatan@inrae.fr to discuss your involvement.
+
+### Organization and CI
+
+The deepvisiontools repository has two protected branches: *main* and *prerelease*. *main* is reserved for official releases, and only maintainers are allowed to merge or push. *prerelease* is used for pre-release validation: the full CI stages are run, but they are non-blocking, which allows checking potential problems or errors before merging into *main*.
+
+The deepvisiontools repository has a CI pipeline that includes:
+
+- linting / formatting checks with ruff
+- security checks (high-risk issues are blocking only on protected release paths)
+- package build
+- tests (Python 3.10, 3.11, 3.12, 3.13, 3.14)
+- documentation build
+
+### Unit and functional tests
+
+For contributors, you can run the local series of unit/functional tests. First, install the additional test-related dependencies.
+
+We strongly recommend that you run local tests before creating a merge request.
+
+#### Run with coverage
+
+```bash
+coverage run -m pytest      # run tests under coverage
+coverage report -m          # text coverage summary
+coverage html               # HTML coverage report in htmlcov/index.html
+```
+
+#### Running subsets
+
+```bash
+pytest tests/unit -q         # only unit tests
+pytest tests/functional -q   # only functional tests
+```
+
+#### Run with nox
+
+```bash
+nox -s tests-3.xx   # run CPU-only, slow-only tests on a specific Python version: 3.10, 3.11, 3.12, 3.13 or 3.14
+```
+
+For example:
+
+```bash
+nox -s tests-3.12
+```
+
+### Check other CI stages locally
+
+Use nox, which is installed if `".[dev]"` was used during installation:
+
+```bash
+nox -s ruff        # formatting / linting
+nox -s typecheck   # type checking with mypy
+nox -s security    # security checks with bandit
+nox -s tests-3.xx  # CPU-only, non-slow tests on a specific Python version
+nox -s docs        # build the docs
+```
+**note that the nox docs session produces a _nox_doc/ folder, so you can inspect the doc locally**
+
+### Linting and formatting
+
+We suggest that you use ruff during development to maintain code compatibility with the tests. You can use ruff to check linting and formatting:
+
+```bash
+ruff check src tests          # check linting on src and tests folders
+ruff check --fix src tests    # apply safe lint fixes
+ruff format --check src tests # check formatting
+ruff format src tests         # apply safe formatting on src and tests folders
+```
+
+Even better is to use ruff in your IDE.
+
+### release on PyPi
+*Only maintainer can create release with propper tags*
 
+To release on pypi you must tag the version from the main branch according to this pattern `"vX.Y.Z"`. The tag version must match the `X.Y.Z` package version of `deepvisiontools.__version__`. Once the tagged is done, you can push on main and the CI release stage will upload the dist to pypi.

deepvisiontools-1.0.0/README.md

@@ -0,0 +1,198 @@
+<p align="center">
+  <img align="center" src="supplementary/logo-dvt.png" height="100">
+</p>
+
+# Welcome to deepvisiontools!
+
+**deepvisiontools** is a library built on torch for training and inference processes for deep learning computer vision tasks.
+
+<p align="center">
+  <img align="center" src="supplementary/coco_visu.png" height="130"> &nbsp; &nbsp; &nbsp; <img align="center" src="supplementary/veg_visu.png" height="130"> &nbsp; &nbsp; &nbsp; <img align="center" src="supplementary/dice_visu.png" height="130">
+</p>
+
+## News
+
+- *release 1.0.0: detector / detector-seg / tabdet / tabdet-seg model wrappers are heavy but powerful models for detection and instance segmentation. However, they are not fully pretrained (only backbone on imagenet). Therefore, on small datasets (< 1000 images ish, but depend on the dataset ...) they can be slightly less competitive than yolo models (also included in the library) that were fully pretrained on coco. We are currently pretraining some of the models but it takes time. We are goind to update the weights once done, so keep tuned for new releases. We suggest you always train a yolo model as well for a baseline comparison. However, on larger datasets they can significantly outperform yolo11x, but again it depends on the dataset ...*
+
+## Scope
+
+deepvisiontools is suitable for supervised learning computer vision tasks. Tasks currently available are:
+
+- detection (bounding boxes)
+- instance segmentation (object instance masks)
+- semantic segmentation (semantic masks)
+
+Future releases will include:
+
+- classification
+- regression
+- keypoints/pose detection
+- crowd detection (point detection)
+
+deepvisiontools provides integrated pipelines that include dataset handlers, model wrappers, training and hyper-parameter fine-tuning helpers, metrics, inference and evaluation tools.
+
+deepvisiontools is designed for both **non-expert and expert** users by providing default operational frameworks that are easy to use, while keeping maximal flexibility and control for experts.
+
+
+## Context
+
+deepvisiontools is developed by INRAE, the French National Research Institute for Agriculture, Food and the Environment. It was partly supported by the [PHENOME-EMPHASIS project](https://www.phenome-emphasis.fr/). It is associated with INRAE CATI CONNECT.
+
+## Citations
+
+To cite deepvisiontools when used, please cite the official repository: https://forge.inrae.fr/ue-apc/librairies/python/deepvisiontools
+
+**Important**: deepvisiontools provides model wrappers and does not reimplement all models. It is the responsibility of the user to also cite the model developers.
+
+If you are using one of these models, please cite the corresponding library as well:
+
+- `"yolo"`: cite Ultralytics (https://www.ultralytics.com/)
+- `"smp"`: cite segmentation-models-pytorch (https://github.com/qubvel-org/segmentation_models.pytorch)
+
+Note that detector and tabdet models (VFNet, Cascade R-CNN/Mask R-CNN) are deepvisiontools reimplementations using PyTorch Image Models as backbones (https://timm.fast.ai/). They were reimplemented following MMDetection (https://github.com/open-mmlab/mmdetection) guidelines.
+
+Finally, hyper-parameter optimisations were built on Optuna. If you use the OptunaWrapper helper, it is worth mentioning Optuna in your work (https://optuna.org/).
+
+## Installation
+
+It is recommended to install deepvisiontools on a GPU-enabled machine/server. We suggest that you use a Python virtual environment (uv, conda, venv, etc.) with **Python version between 3.10 and 3.14**.
+
+deepvisiontools is available on PyPI, so you can simply run in a Python environment:
+
+```bash
+pip install deepvisiontools
+```
+
+To install deepvisiontools in development mode, please clone the repository and run:
+
+```bash
+pip install -e deepvisiontools/ # local cloned repo
+```
+
+Finally, to install deepvisiontools with developer extras, move inside the cloned repo and run:
+
+```bash
+pip install -e ".[dev]"
+```
+
+This will include extras like documentation, pytest, nox, etc.
+
+## Documentation and tutorials
+
+Official documentation for deepvisiontools can be found here: https://deepvisiontools-1233bd.pages-forge.inrae.fr/
+
+To get started with deepvisiontools, after installation you can run:
+
+```bash
+deepvisiontools download-tutorials
+```
+
+You will also need to download the demo datasets:
+
+```bash
+deepvisiontools download-demo-datasets
+```
+
+The first tutorial, **getting_started**, takes about 1 hour and explains all the basics of deepvisiontools. From there you have all information to get started with simple tasks.
+
+We then recommend moving to **advanced_usage** to get a broader view of what you can configure in deepvisiontools and all available helpers.
+
+Finally, the **hyper_parametrs_optimization** tutorial will explain the usage of Optuna wrappers available in deepvisiontools to perform extensive parameter space searches.
+
+## Monitoring via MLflow
+
+deepvisiontools includes MLflow for monitoring, checkpointing and artifacts during training.
+
+If you use deepvisiontools as intended, at some point you will end up with an `mlruns` folder.
+
+You can use the following command to start MLflow and then open the local page in your browser to display the MLflow UI:
+
+```bash
+mlflow server --backend-store-uri sqlite:////$(pwd)/mlruns/mlflow.db --default-artifact-root file:$(pwd)/mlruns --host 127.0.0.1 --port 5000
+```
+
+An alternative way for Jupyter notebooks is presented in the tutorial.
+
+## License
+
+GNU Affero General Public License v3 or later (AGPLv3+) (AGPL-3.0).
+
+## Contributor specifics
+
+If you wish to contribute as a developer to deepvisiontools, you need to be part of INRAE or PHENOME-EMPHASIS. You can contact jordan.bernigaud-samatan@inrae.fr to discuss your involvement.
+
+### Organization and CI
+
+The deepvisiontools repository has two protected branches: *main* and *prerelease*. *main* is reserved for official releases, and only maintainers are allowed to merge or push. *prerelease* is used for pre-release validation: the full CI stages are run, but they are non-blocking, which allows checking potential problems or errors before merging into *main*.
+
+The deepvisiontools repository has a CI pipeline that includes:
+
+- linting / formatting checks with ruff
+- security checks (high-risk issues are blocking only on protected release paths)
+- package build
+- tests (Python 3.10, 3.11, 3.12, 3.13, 3.14)
+- documentation build
+
+### Unit and functional tests
+
+For contributors, you can run the local series of unit/functional tests. First, install the additional test-related dependencies.
+
+We strongly recommend that you run local tests before creating a merge request.
+
+#### Run with coverage
+
+```bash
+coverage run -m pytest      # run tests under coverage
+coverage report -m          # text coverage summary
+coverage html               # HTML coverage report in htmlcov/index.html
+```
+
+#### Running subsets
+
+```bash
+pytest tests/unit -q         # only unit tests
+pytest tests/functional -q   # only functional tests
+```
+
+#### Run with nox
+
+```bash
+nox -s tests-3.xx   # run CPU-only, slow-only tests on a specific Python version: 3.10, 3.11, 3.12, 3.13 or 3.14
+```
+
+For example:
+
+```bash
+nox -s tests-3.12
+```
+
+### Check other CI stages locally
+
+Use nox, which is installed if `".[dev]"` was used during installation:
+
+```bash
+nox -s ruff        # formatting / linting
+nox -s typecheck   # type checking with mypy
+nox -s security    # security checks with bandit
+nox -s tests-3.xx  # CPU-only, non-slow tests on a specific Python version
+nox -s docs        # build the docs
+```
+**note that the nox docs session produces a _nox_doc/ folder, so you can inspect the doc locally**
+
+### Linting and formatting
+
+We suggest that you use ruff during development to maintain code compatibility with the tests. You can use ruff to check linting and formatting:
+
+```bash
+ruff check src tests          # check linting on src and tests folders
+ruff check --fix src tests    # apply safe lint fixes
+ruff format --check src tests # check formatting
+ruff format src tests         # apply safe formatting on src and tests folders
+```
+
+Even better is to use ruff in your IDE.
+
+### release on PyPi
+*Only maintainer can create release with propper tags*
+
+To release on pypi you must tag the version from the main branch according to this pattern `"vX.Y.Z"`. The tag version must match the `X.Y.Z` package version of `deepvisiontools.__version__`. Once the tagged is done, you can push on main and the CI release stage will upload the dist to pypi.

deepvisiontools-1.0.0/pyproject.toml

@@ -0,0 +1,177 @@
+[build-system]
+requires = ["setuptools>=78.1.1,<82"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "deepvisiontools"
+description = "deepvisiontools is a library built on torch for training and inference processes for deep learning computer vision tasks."
+readme = "README.md"
+dynamic = ["version"]
+requires-python = ">=3.10,<3.15"
+authors = [
+  { name = "Lucas Bernigaud Samatan", email = "lucas.bernigaud-samatan@inrae.fr" },
+  { name = "Jordan Bernigaud Samatan", email = "jordan.bernigaud-samatan@inrae.fr" },
+]
+maintainers = [
+  { name = "Jordan Bernigaud Samatan", email = "jordan.bernigaud-samatan@inrae.fr" },
+  { name = "Lucas Bernigaud Samatan", email = "lucas.bernigaud-samatan@inrae.fr" },
+]
+classifiers = [
+"Development Status :: 5 - Production/Stable",
+  "Programming Language :: Python :: 3.10",
+  "Programming Language :: Python :: 3.11",
+  "Programming Language :: Python :: 3.12",
+  "Programming Language :: Python :: 3.13",
+  "Programming Language :: Python :: 3.14",
+]
+license = { file = "LICENSE.txt" }
+
+# Runtime dependencies.
+# Policy: allow compatible releases within the same major version.
+# Exception: torch keeps the explicit minimum at 2.4.0.
+dependencies = [
+  "torch>=2.4.0,<3.0",
+  "torchvision>=0.19.0,<1.0",
+  "pycocotools>=2.0.8,<3.0",
+  "torchmetrics>=1.4.0,<2.0",
+  "tensorboard>=2.4.1,<3.0",
+  "torch-tb-profiler>=0.4.0,<1.0",
+  "mlflow>=3.1.0,<4.0",
+
+  # Security constraints for transitive MLflow / dashboard web dependencies.
+  # fastapi 0.136.3 is flagged as MAL-2026-4750.
+  # starlette <1.0.1 is flagged as PYSEC-2026-161.
+  "fastapi!=0.136.3,<1.0",
+  "starlette>=1.0.1,<2.0",
+
+  "ultralytics>=8.3.28,<9.0",
+  "transformers>=5.0.0,<6.0",
+  "colorcet>=3.1.0,<4.0",
+  "segmentation-models-pytorch>=0.4.1,<1.0",
+  "openpyxl>=3.1.5,<4.0",
+  "optuna>=4.0.0,<5.0",
+  "optuna-dashboard>=0.20.0,<1.0",
+  "huggingface_hub>=1.5.0,<2.0",
+  "pillow>=12.2.0,<13.0",
+  "setuptools>=78.1.1,<82",
+]
+
+[project.urls]
+Homepage = "https://forge.inrae.fr/ue-apc/librairies/python/deepvisiontools"
+Issues = "https://forge.inrae.fr/ue-apc/librairies/python/deepvisiontools/-/issues"
+Repository = "https://forge.inrae.fr/ue-apc/librairies/python/deepvisiontools"
+Documentation = "https://deepvisiontools-1233bd.pages-forge.inrae.fr/"
+
+# Development / CI dependencies.
+# Install locally with: python -m pip install -e ".[dev]"
+[project.optional-dependencies]
+dev = [
+  "pytest>=9.0,<10.0",
+  "pytest-cov>=7.0,<8.0",
+  "coverage[toml]>=7.0,<8.0",
+  "hypothesis>=6.0,<7.0",
+  "ruff>=0.13.0,<1.0",
+  "mypy>=1.0,<2.0",
+  "bandit[toml]>=1.7,<2.0",
+  "pip-audit>=2.7,<3.0",
+  "nox[pbs]>=2025.0.0,<2027.0.0",
+  "build>=1.2,<2.0",
+  "twine>=6.0,<7.0",
+  "types-PyYAML>=6.0.12,<7.0",
+]
+
+internimage-build = [
+  "huggingface_hub>=0.24.0,<1.0",
+]
+
+# Setuptools config: discover the package from src/.
+[tool.setuptools.packages.find]
+where = ["src"]
+
+[project.scripts]
+dvt-install-dcnv3 = "deepvisiontools.install_dcnv3:main"
+deepvisiontools = "deepvisiontools.cli:main"
+
+[tool.setuptools.dynamic]
+version = { attr = "deepvisiontools.__version__" }
+
+[tool.setuptools]
+include-package-data = true
+
+[tool.setuptools.package-data]
+deepvisiontools = ["**/*.yaml", "**/*.yml"]
+
+# Pytest config.
+[tool.pytest.ini_options]
+testpaths = ["tests"]
+addopts = "-ra"
+markers = [
+  "cpu: CPU test variant.",
+  "gpu: CUDA test variant, skipped when torch.cuda.is_available() is false.",
+  "slow: Slow runtime tests.",
+]
+
+# Coverage config.
+[tool.coverage.run]
+source = ["src/deepvisiontools"]
+branch = true
+
+[tool.coverage.report]
+skip_empty = true
+show_missing = true
+fail_under = 85
+
+# Ruff config: formatting + linting.
+[tool.ruff]
+target-version = "py310"
+line-length = 88
+src = ["src", "tests"]
+exclude = [
+  ".git",
+  ".mypy_cache",
+  ".nox",
+  ".pytest_cache",
+  ".ruff_cache",
+  ".venv",
+  "build",
+  "dist",
+  "public",
+  "site",
+]
+
+[tool.ruff.lint]
+select = [
+  "E",    # pycodestyle errors
+  "F",    # pyflakes
+  "I",    # import sorting
+  "B",    # bugbear
+  "UP",   # pyupgrade
+  "C4",   # comprehensions
+  "SIM",  # simplifications
+  "RUF",  # Ruff-specific rules
+]
+ignore = [
+  "E501", # formatter handles line length; avoids noisy docstring failures
+]
+
+[tool.ruff.format]
+quote-style = "double"
+indent-style = "space"
+line-ending = "auto"
+
+# Mypy config: gradual typing for the package source.
+[tool.mypy]
+python_version = "3.10"
+mypy_path = "src"
+files = ["src/deepvisiontools"]
+show_error_codes = true
+pretty = true
+ignore_missing_imports = true
+check_untyped_defs = true
+warn_unused_ignores = true
+warn_redundant_casts = true
+no_implicit_optional = true
+
+# Bandit config: scan source code, not tests.
+[tool.bandit]
+exclude_dirs = ["tests", ".nox", "build", "dist", "public", "site"]