layer.innerProduct.weights.floatValue.extend(weights)
layer.innerProduct.bias.floatValue.extend(biases)
spec.neuralNetworkClassifier.stringClassLabels.vector.extend(labels)
# Make this model trainable.
builder = NeuralNetworkBuilder(spec=model._spec)
builder.make_updatable(["fullyconnected0"])
builder.set_categorical_cross_entropy_loss(name="lossLayer", input="labelProbability")
builder.set_epochs(10, [1, 10, 50])
# Using the SDG optimizer:
sgd_params = SgdParams(lr=0.001, batch=8, momentum=0)
sgd_params.set_batch(8, [1, 2, 8, 16])
builder.set_sgd_optimizer(sgd_params)
# Using the Adam optimizer:
# adam_params = AdamParams(lr=0.001, batch=8, beta1=0.9, beta2=0.999, eps=1e-8)
# adam_params.set_batch(8, [1, 2, 8, 16])
# builder.set_adam_optimizer(adam_params)
builder.spec.description.trainingInput[0].shortDescription = "Example image"
builder.spec.description.trainingInput[1].shortDescription = "True label"
coremltools.utils.save_spec(builder.spec, "../Models/HandsTuri.mlmodel")
# Replace the weights of the last layer with random weights.
model = coremltools.models.MLModel("../Models/HandsTuri.mlmodel")
model.short_description = ""
input_features = [('data', datatypes.Array(*input_dim))]
output_features = [('result', datatypes.Array(*output_dim))]
weights = np.random.rand(1, input_max_size)
bias = np.random.rand(1)
builder = NeuralNetworkBuilder(input_features, output_features)
builder.add_inner_product(name='ip_layer',
W=weights,
b=bias,
input_channels=input_max_size,
output_channels=1,
has_bias=True,
input_name='data',
output_name='result')
builder.make_updatable(['ip_layer'])
builder.set_mean_squared_error_loss(name='lossLayer',
input_feature=output_features[0])
optimizerParams = SgdParams(lr=0.01, batch=1)
optimizerParams.set_batch(1, allowed_set=[1, 2, 4, 8, 16, 32])
builder.set_sgd_optimizer(optimizerParams)
builder.set_epochs(16, allowed_set=[2, 4, 8, 16, 32, 64, 128, 256])
#builder.spec = convert_neural_network_spec_weights_to_fp16(builder.spec)
save_spec(builder.spec, '../core/LinearRegressionModel.mlmodel')
