To calibrate and validate the LogoNet network. In this example, use the images in the logos_dataset data set Parameters of the convolution layers in the network. Understand the effects of the limited range and precision of the quantized learnable Results, the calibration data must be representative of inputs to the LogoNetĪfter quantization, the app uses the validation data set to test the network to Of the activations in all layers of the LogoNet network. The app also exercises the dynamic ranges Network and collect the dynamic ranges of the weights and biases in the convolutionĪnd fully connected layers of the network. The Deep Network Quantizer app uses calibration data to exercise the If the metric function is not on the path, this step will produce an error.ĭefine calibration and validation data to use for quantization. The custom metric function must be on the path. Select hComputeModelAccuracy as the metric function to use. Select Add to add hComputeModelAccuracy to the list of metric functions available in the app. To revalidate the network using this custom metric function, under Quantization Options, enter the name of the custom metric function, hComputeModelAccuracy. PredictionError = (yActual = groundTruth(idx)) %#ok end % Sum all prediction errors.Īccuracy = sum(predictionError)/numel(predictionError) % Compare with predicted label with actual ground truth %% Computes model-level accuracy statistics % Load ground truth For more information on how to interpret these histograms, see Quantization of Deep Neural Networks.įunction accuracy = hComputeModelAccuracy(predictionScores, net, dataStore) The gray regions of the histograms indicate data that cannot be represented by the quantized representation. To the right of the table, the app displays histograms of the dynamic ranges of the parameters. When the calibration is complete, the app displays a table containing the weights and biases in the convolution and fully connected layers of the network and the dynamic ranges of the activations in all layers of the network and their minimum and maximum values during the calibration. The Deep Network Quantizer uses the calibration data to exercise the network and collect range information for the learnable parameters in the network layers. In the Calibrate section of the toolstrip, under Calibration Data, select the augmentedImageDatastore object from the base workspace containing the calibration data, aug_calData. The app displays the layer graph of the selected network.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |