Heavy work in progress

NNPack contains tools & standards for packaging and distributing Neural Networks. Today's Deep Learning frameworks are very powerful, but they their tools for persisting and distributing Machine Learning models are lacking. The NNPack specification provides a high-level structure for packaging up various types of models and their training data (Model Scaffolds).

• NNPack Specification 1.0-DRAFT

• Coming soon: Easily push/pull models from the open nnpack.org repository
• Developer-friendly packaging of models. Model meta-data lives in /nnpackage.json
• Packaging and distributing training data using Model Scaffolds
• Rich contextual meta-data for model, training data and labels

nnpack pull dominiek/scene_type_vision

In order to package a model, simply create an nnpackage.json in its root directory. All neural network state information needs to be moved into a state/ subfolder. Here's an example:

{
  "id": "58745c9fbd17c82dd4ff7c9c",
  "name": "Scene Types",
  "description": "Classify images into Indoor VS Outdoor scenes",
  "version": 0.1,
  "engines": {
    "tensorflow": ">=0.11"
  },
  "nodes": {
    "softmaxOutput": {
      "type": "tensor",
      "id": "retrained_layer:0"
    },
    "convolutionalRepresentation": {
      "type": "tensor",
      "id": "pool_3/_reshape:0"
    }
  },
  "author": {
    "name": "Dominiek Ter Heide",
    "email": "info@dominiek.com"
  },
  "labelsDefinitionFile": "labels.json",
  "stateDir": "state"
}

For each softmax output of the neural net, a labels.json provides context for what the predictions mean. Each node_id corresponds to a softmax output:

{
  "labels": [
    {"id": "58745e65bd17c82ec1545a64", "name": "Indoor", "node_id": 0},
    {"id": "58745e69bd17c82ec1545a65", "name": "Outdoor", "node_id": 1}
  ]
}

In the above two files a lot of information is contained on how to run the model, what engine is required, what the outputs mean and how transfer learning can occur.

Read Documentation

nnpack init .
nnpack push dominiek/scene_type_vision

• Encapsulate utility functions in PIP module
• Add utility functions for creating Models and Scaffolds
• Create a Spec roadmap
• Create simple website skeleton
• Create NNPack Registry beta