Optional
convertedWhat library or tool is responsible for converting the original model to this format, applicable only if the model is output by a converter.
Used for debugging purposes. E.g., 'TensorFlow.js Converter v1.0.0'.
A value of null
means the model artifacts are generated without any
conversion process (e.g., saved directly from a TensorFlow.js
tf.LayersModel
instance.)
Optional
formatHard-coded format name for models saved from TensorFlow.js or converted by TensorFlow.js Converter.
Optional
generatedWhat library is responsible for originally generating this artifact.
Used for debugging purposes. E.g., 'TensorFlow.js v1.0.0'.
Optional
initializerInputs and outputs signature for model initializer.
Optional
modelInitializer for the model.
Model topology.
For Keras-style tf.Model
s, this is a JSON object.
For TensorFlow-style models (e.g., SavedModel
), this is the JSON
encoding of the GraphDef
protocol buffer.
Optional
signatureInputs and outputs signature for saved model.
Optional
trainingModel training configuration.
Optional
userUser-defined metadata about the model.
Weights manifest.
The weights manifest consists of an ordered list of weight-manifest
groups. Each weight-manifest group consists of a number of weight values
stored in a number of paths. See the documentation of
WeightsManifestConfig
for more details.
The on-disk format of the
model.json
file.TF.js 1.0 always populates the optional fields when writing model.json. Prior versions did not provide those fields.