---
id: edge-mdt-cl
version: "1.1.1"
license: unclear
license_treatment: permissive
maintenance: aging
---

# edge-mdt-cl — Edge MDT Custom Layers package

License: permissive · Maintenance: aging · Downloads: 108.4K/mo

## What it is and what it does

Edge MDT Custom Layers fills a gap in standard deep learning frameworks by providing detection post-processing layers that TensorFlow Keras and PyTorch do not natively support. It allows integration of these custom layers into pretrained models for edge deployment scenarios. The package supports both TensorFlow (with .keras serialization) and PyTorch (with ONNX export and OnnxRuntime support), making it useful for workflows that need specialized detection post-processing without building custom operations from scratch.

The package depends on numpy and packaging, and installs as a pure Python wheel with optional framework-specific dependencies. It is designed for Python >=3.10, with tested support for TensorFlow 2.14–2.15 and PyTorch 2.3–2.6. The codebase is permissively licensed under Apache License 2.0, but maintenance activity has slowed—the last release was 200 days ago—so adoption should account for potential delays in bug fixes or compatibility updates.

Use it for:

- Add detection post-processing layers to TensorFlow models for edge inference without writing custom operations.
- Export PyTorch detection models to ONNX with custom layers preserved for OnnxRuntime deployment.
- Integrate specialized detection post-processing into pretrained models when standard framework layers are insufficient.
- Serialize and load detection models with custom layers using TensorFlow's .keras format or ONNX.
- Support edge ML pipelines that require detection post-processing not available in mainstream framework APIs.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Provides custom neural network layers for detection post-processing that are not available in TensorFlow Keras or PyTorch's standard APIs, enabling integration into pretrained models.

Yes, if you need detection post-processing layers not available in standard TensorFlow or PyTorch APIs and are comfortable with a package in aging maintenance status. The low install friction, permissive license, and zero known vulnerabilities make it safe to try. However, verify that the specific layers you need are supported and plan for potential delays in updates or bug fixes.

## Install

pip install edge-mdt-cl
uv add edge-mdt-cl
poetry add edge-mdt-cl

## Installing edge-mdt-cl

Before you install: Low friction installation with optional framework dependencies. Maintenance status is aging—last release was 200 days ago—so expect slower response to issues or updates.

License in practice: Apache License 2.0 (permissive) allows commercial and private use with minimal restrictions; you must include a copy of the license and state significant changes.

Quickstart:

pip install edge-mdt-cl
# For TensorFlow support:
pip install edge-mdt-cl[tf]
# For PyTorch support:
pip install edge-mdt-cl[torch]

Requires Python >=3.10; TensorFlow tested on 2.14–2.15 with Python 3.10–3.11; PyTorch tested on torch 2.3–2.6 with Python 3.10–3.12.

Verify before relying:

- Specific custom layer names and their detection post-processing capabilities.
- Whether the package is actively maintained or in maintenance-only mode given the 200-day gap.
- Real-world performance or accuracy impact of using these layers versus alternatives.

## Package facts

- License: not declared (permissive)
- Python support: supports_current
- Install friction: low
- Maintenance: aging
- Downloads: 108.4K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

custom neural network layers detection, tensorflow keras custom layers, pytorch detection post-processing, edge ml model layers, onnx custom operations, detection-layers, edge-ml, onnx-export

[View on SkillFed](https://skillfed.io/packages/edge-mdt-cl) · [View on PyPI](https://pypi.org/project/edge-mdt-cl/)