dlpy.model.Model¶
-
class
dlpy.model.
Model
(conn, inputs=None, outputs=None, model_table=None, model_weights=None)¶ -
__init__
(conn, inputs=None, outputs=None, model_table=None, model_weights=None)¶ Initialize self. See help(type(self)) for accurate signature.
Methods
__init__
(conn[, inputs, outputs, …])Initialize self.
change_labels
(label_file, id_column, …)Overrides the labels already in the model
compile
()parse the network nodes and process CAS Action
count_instances
()Count the total number of parameters in the model
create_layer_id_name_mapping
()Create a dictionary which maps layer id to layer name. One use case is the model creates weights table given a pre-trained weights and a pre-trained model. Example: mapper = model.create_layer_id_name_mapping() pretrained_weights_table = WeightsTable(conn, weights_tbl_name=’my_pretrained_weights_table’, model_tbl_name=’my_pretrained_model_table’) pretrained_weights_table.remap_layer_ids(mapper, casout=’new_weights).
deploy
(path[, output_format, model_weights, …])Deploy the deep learning model to a data file
evaluate
(data[, text_parms, layer_out, …])Evaluate the deep learning model on a specified validation data set
evaluate_object_detection
(ground_truth, …)Evaluate the deep learning model on a specified validation data set.
fit
(data[, inputs, target, data_specs, …])Fitting a deep learning model.
fit_and_visualize
(data[, inputs, target, …])Fitting a deep learning model while visulizing the fit and loss at each iteration.
forecast
([test_table, horizon, train_table, …])Make forecasts based on deep learning models trained on TimeseriesTable.
format_name
([block_num, local_count])Format the name of the layer
from_caffe_model
(conn, input_network_file[, …])Generate a model object from a Caffe model proto file (e.g. *.prototxt), and convert the weights (e.g. *.caffemodel) to a SAS capable file (e.g. *.caffemodel.h5).
from_keras_model
(conn, keras_model[, …])Generate a model object from a Keras model object
from_onnx_model
(conn, onnx_model[, …])Generate a Model object from ONNX model.
from_sashdat
(conn, path[, output_model_table])Generate a model object using the model information in the sashdat file
from_table
(input_model_table[, …])Create a Model object from CAS table that defines a deep learning model
get_feature_maps
(data[, label, idx, image_id])Extract the feature maps for a single image
get_features
(data, dense_layer[, target])Extract linear features for a data table from the layer specified by dense_layer
Return the information about the model table
get_number_of_instances
()heat_map_analysis
([data, mask_width, …])Conduct a heat map analysis on table of images
load
(path[, display_note])Load the deep learning model architecture from existing table
load_weights
(path[, labels, data_spec, …])Load the weights from a data file specified by ‘path’
load_weights_attr
(path)Load the weights attribute form a sashdat file
load_weights_from_caffe
(path[, labels, …])Load the model weights from a HDF5 file
load_weights_from_file
(path[, format_type, …])Load the model weights from a HDF5 file
load_weights_from_file_with_labels
(path[, …])Load the model weights from a HDF5 file
load_weights_from_keras
(path[, labels, …])Load the model weights from a HDF5 file
load_weights_from_table
(path)Load the weights from a file
plot_evaluate_res
([cas_table, img_type, …])Plot the bar chart of the classification predictions
plot_heat_map
([idx, alpha])Display the heat maps analysis results
Display a graph that summarizes the model architecture.
plot_training_history
([items, fig_size, …])Display the training iteration history.
predict
(data[, text_parms, layer_out, …])Evaluate the deep learning model on a specified validation data set
Display a table that summarizes the model architecture
save_to_astore
([path, layers])Save the model to an astore object, and write it into a file.
save_to_onnx
(path[, model_weights])Save to ONNX model
save_to_table
(path)Save the model as SAS dataset
save_to_table_with_caslibify
(path)Save the model as SAS dataset
save_weights_csv
(path)Save model weights table as csv
score
(table[, model, init_weights, …])Inference of input data with the trained deep learning model
set_weights
(weight_tbl)Assign weights to the Model object
set_weights_attr
(attr_tbl[, clear])Attach the weights attribute to the model weights
share_weights
(layers)Share weights between layers
to_functional_model
([stop_layers])Convert a Sequential into a functional model and return the functional model.
to_model_params
()Convert the model configuration to CAS action parameters
train
(table[, attributes, inputs, nominals, …])Trains a deep learning model
tune
(data[, inputs, target])Tunes hyper parameters for the deep learning model.
Attributes
can_be_last_layer
feature_maps
layer_id
model_ever_trained
model_explain_table
n_epochs
name
number_of_instances
rnn_summary
Return a DataFrame containing the layer information for rnn models
score_message_level
src_layers
summary
Return a DataFrame containing the layer information
train_tbl
training_history
type
type_desc
type_label
valid_conf_mat
valid_res
valid_res_tbl
valid_score
valid_tbl
-