# microsoft/MMdnn

MMdnn is a set of tools to help users inter-operate among different deep learning frameworks. E.g. model conversion and visualization. Convert models between Caffe, Keras, MXNet, Tensorflow, CNTK, PyTorch Onnx and CoreML.

Repository: https://github.com/microsoft/MMdnn
Canonical: https://ross.abutalabs.com/products/mmdnn
Language: Python
License: MIT
License Family: permissive
Topics: cntk, visualization, tensorflow, model-converter, pytorch, caffe, keras, mxnet, coreml, onnx, darknet
Last push: 2025-08-07T06:18:09+00:00

## Health v2 (maintenance only)
Score: 39/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 35, release rhythm 8, longevity 100
- inputs: {"age_days": 3304, "days_push": 391, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 5805, forks 955 (observed 2026-08-28T04:09:29.859360+00:00)

## What it is
MMdnn is a Microsoft toolkit for converting, visualizing, and diagnosing deep learning models across frameworks such as TensorFlow, PyTorch, Keras, Caffe, MXNet, CNTK, ONNX, and CoreML. It also generates code snippets for retraining and provides a model collection and network architecture visualizer.

## Use cases
- convert a pytorch model to tensorflow
- convert keras model to onnx
- visualize deep learning model architecture
- train a model in one framework and deploy in another
- convert caffe model to pytorch
- prepare a model for tensorrt inference
- find popular pretrained deep learning models

## When to choose
- you need to move a trained model between deep learning frameworks
- you want a visual graph of a model's network architecture
- you work with legacy frameworks like Caffe, CNTK, or MXNet
- you need code snippets to retrain a converted model

## When to avoid
- you only need modern framework-to-ONNX conversion, which native exporters handle better
- you need actively maintained tooling with support for the latest framework versions
- you need quantization or model compression features

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, developer-tools, data-visualization
- domain: deep-learning, machine-learning, developer-tools
- platform: python, cross-platform
- tags: model-conversion, model-converter, onnx, tensorflow, pytorch, keras, caffe, mxnet, coreml, cntk, darknet, model-visualization, interoperability

## Member repositories
- microsoft/MMdnn (main) score 39

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:29.859360+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:52:44.414536+00:00, confidence not recorded.
  - readme: https://github.com/microsoft/MMdnn (fetched 2026-08-28T04:09:29.859360+00:00, sha 07c449f99bf3)
  - registry_pypi: https://pypi.org/pypi/mmdnn/json (fetched 2026-08-29T08:47:47.753742+00:00, sha 42ab7cfd30ba)
- Data as of 2026-08-30T08:39:29.467469+00:00.