Module monk.gluon.finetune.level_13_updates_main

Classes

class prototype_updates (verbose=1)

Main class for all parametric update functions

Args

verbose : int

Set verbosity levels 0 - Print Nothing 1 - Print desired details

Expand source code

class prototype_updates(prototype_losses):
    '''
    Main class for all parametric update functions

    Args:
        verbose (int): Set verbosity levels
                        0 - Print Nothing
                        1 - Print desired details
    '''

    def __init__(self, verbose=1):
        super().__init__(verbose=verbose);


    ##########################################################################################################################################################
    def update_input_size(self, input_size):
        '''
        Update input size.

        Args:
            input_size (int): New input size

        Returns:
            None
        '''
        self.system_dict = set_input_size(input_size, self.system_dict);
        
        self.custom_print("Update: Input size - {}".format(self.system_dict["dataset"]["params"]["input_size"]));
        self.custom_print("");
        


    def update_batch_size(self, batch_size):
        '''
        Update batch size.

        Args:
            batch_size (int): New batch size

        Returns:
            None
        '''
        self.system_dict = set_batch_size(batch_size, self.system_dict);
        
        self.custom_print("Update: Batch size - {}".format(self.system_dict["dataset"]["params"]["batch_size"]));
        self.custom_print("");
        


    def update_shuffle_data(self, shuffle):
        '''
        Update to shuffle data or not.

        Args:
            shuffle (bool): If True, will shuffle data

        Returns:
            None
        '''
        self.system_dict = set_data_shuffle(shuffle, self.system_dict);
        
        self.custom_print("Update: Data shuffle - {}".format(self.system_dict["dataset"]["params"]["train_shuffle"]));
        self.custom_print("");
        
    

    def update_num_processors(self, num_processors):
        '''
        Update num processors for data loader.

        Args:
            num_processors (int): Max CPUs for data sampling

        Returns:
            None
        '''
        self.system_dict = set_num_processors(num_processors, self.system_dict);
        
        self.custom_print("Update: Num processors - {}".format(self.system_dict["dataset"]["params"]["num_workers"]));
        self.custom_print("");
        


    def update_weighted_sampling(self, sample):
        '''
        Function inactive
        '''
        self.system_dict = set_weighted_sampling(sample, self.system_dict);
        
        self.custom_print("Update: Weighted Sampling - {}".format(self.system_dict["dataset"]["params"]["weighted_sample"]));
        self.custom_print("");




    def update_trainval_split(self, value):
        '''
        Update training-validation split
        Args:
            value (float): Indicating train validation split
                            Division happens as follows:
                                train - total dataset * split * 100
                                val   - total dataset * (1-split) * 100 

        Returns:
            None
        '''
        if(self.system_dict["dataset"]["dataset_type"] == "train"):
            dataset_path = self.system_dict["dataset"]["train_path"];
            path_to_csv=False;
        elif(self.system_dict["dataset"]["dataset_type"] == "train-val"):
            dataset_path = [self.system_dict["dataset"]["train_path"], self.system_dict["dataset"]["val_path"]];
            path_to_csv=False;
        elif(self.system_dict["dataset"]["dataset_type"] == "csv_train"):
            dataset_path = self.system_dict["dataset"]["train_path"];
            path_to_csv = self.system_dict["dataset"]["csv_train"];
        elif(self.system_dict["dataset"]["dataset_type"] == "csv_train-val"):
            dataset_path = [self.system_dict["dataset"]["train_path"], self.system_dict["dataset"]["val_path"]];
            path_to_csv = [self.system_dict["dataset"]["csv_train"], self.system_dict["dataset"]["csv_val"]];
        else:
            msg = "Dataset Type invalid.\n";
            msg += "Cannot update split"
            ConstraintsWarning(msg)

        self.system_dict = set_dataset_train_path(self.system_dict, dataset_path, value, path_to_csv, self.system_dict["dataset"]["params"]["delimiter"]);


    def update_dataset(self, dataset_path=False, split=0.9, path_to_csv=False, delimiter=","):
        '''
        Update dataset path

        Args:
            dataset_path (str, list): Path to Dataset folder
                                      1) Single string if validation data does not exist
                                      2) List [train_path, val_path] in case of separate train and val data
            path_to_csv (str, list): Path to csv file pointing towards images
                                     1) Single string if validation data does not exist
                                     2) List [train_path, val_path] in case of separate train and val data
            value (float): Indicating train validation split
                            Division happens as follows:
                                train - total dataset * split * 100
                                val   - total dataset * (1-split) * 100 
            delimiter (str): Delimiter for csv file

        Returns:
            None
        '''
        self.system_dict = set_dataset_train_path(self.system_dict, dataset_path, split, path_to_csv, delimiter);
    ##########################################################################################################################################################




    ##########################################################################################################################################################
    def update_model_name(self, model_name, force=False):
        '''
        Update model name

        Args:
            model_name (str): Select from available models. Check via List_Models() function
            force (bool): Dummy function

        Returns:
            None
        '''
        if(not force):
            if(self.system_dict["training"]["status"]):
                ConstraintWarning("Model trained using {}\n".format(self.system_dict["model"]["params"]["model_name"]));
                ConstraintWarning("Changing the model will overwrite previously trained models if training is executed.\n");
                inp = input("Do you wish to continue further (y/n):");

                if(inp == "y"):
                    self.system_dict = set_model_name(model_name, self.system_dict);
                    self.custom_print("Update: Model name - {}".format(self.system_dict["model"]["params"]["model_name"]));
                    self.custom_print("");
                else:
                    self.custom_print("Model not updated.");
                    self.custom_print("");
            else:
                self.system_dict = set_model_name(model_name, self.system_dict);
                self.custom_print("Update: Model name - {}".format(self.system_dict["model"]["params"]["model_name"]));
                self.custom_print("");
        else:
            self.system_dict = set_model_name(model_name, self.system_dict);
            self.custom_print("Update: Model name - {}".format(self.system_dict["model"]["params"]["model_name"]));
            self.custom_print("");
    ##########################################################################################################################################################
            

            

    ##########################################################################################################################################################
    def update_model_path(self, model_path, force=False):
        '''
        Update model path for inferencing

        Args:
            model_path (str): Path to model weights.
            force (bool): Dummy function

        Returns:
            None
        '''
        if(not force):
            if(self.system_dict["training"]["status"]):
                ConstraintWarning("Model trained using {}\n".format(self.system_dict["model"]["params"]["model_name"]));
                ConstraintWarning("Changing the model will overwrite previously trained models if training is executed.\n");
                inp = input("Do you wish to continue further (y/n):");

                if(inp == "y"):
                    self.system_dict = set_model_path(model_path, self.system_dict);
                    self.custom_print("Update: Model path - {}".format(self.system_dict["model"]["params"]["model_path"]));
                    self.custom_print("");
                else:
                    self.custom_print("Model not updated.");
                    self.custom_print("");
            else:
                self.system_dict = set_model_path(model_path, self.system_dict);
                self.custom_print("Update: Model path - {}".format(self.system_dict["model"]["params"]["model_path"]));
                self.custom_print("");
        else:
            self.system_dict = set_model_path(model_path, self.system_dict);
            self.custom_print("Update: Model path - {}".format(self.system_dict["model"]["params"]["model_path"]));
            self.custom_print("");               
    ##########################################################################################################################################################



    ##########################################################################################################################################################
    def update_use_gpu(self, gpu):
        '''
        Update to use gpu or cpu

        Args:
            gpu (bool): If True, then use GPU

        Returns:
            None
        '''
        self.system_dict = set_device(gpu, self.system_dict);
        
        self.custom_print("Update: Use Gpu - {}".format(self.system_dict["model"]["params"]["use_gpu"]));
        self.custom_print("");
    ##########################################################################################################################################################


        
    ##########################################################################################################################################################
    def update_use_pretrained(self, pretrained):
        '''
        Update to use pretrained wights or randomly initialized weights

        Args:
            pretrained (bool): If True, use pretrained weights
                                else, use randomly initialized weights

        Returns:
            None
        '''
        self.system_dict = set_pretrained(pretrained, self.system_dict);
        
        self.custom_print("Update: Use pretrained - {}".format(self.system_dict["model"]["params"]["use_pretrained"]));
        self.custom_print("");
    ##########################################################################################################################################################
        



    ##########################################################################################################################################################
    def update_freeze_base_network(self, freeze):
        '''
        Update whether freeze base network or not

        Args:
            freeze (bool): If True, then base network is non-trainable, works as a feature extractor

        Returns:
            None
        '''
        self.system_dict = set_freeze_base_network(freeze, self.system_dict);
        
        self.custom_print("Update: Freeze Base Network - {}".format(self.system_dict["model"]["params"]["freeze_base_network"]));
        self.custom_print("");
    ##########################################################################################################################################################



    ##########################################################################################################################################################
    def update_freeze_layers(self, num_freeze):
        '''
        Update to freeze certain layers in the network

        Args:
            num_freeze (int): Number of layers to freeze in network starting from top

        Returns:
            None
        '''
        self.system_dict["model"]["params"]["num_freeze"] = num_freeze;
        
        self.custom_print("Update: Freeze layers - {}".format(self.system_dict["model"]["params"]["num_freeze"]));
        self.custom_print("");
    ##########################################################################################################################################################





    ##########################################################################################################################################################
    def update_num_epochs(self, num_epochs):
        '''
        Update number of epochs to train the network

        Args:
            num_epochs (int): New number of epochs

        Returns:
            None
        '''
        self.system_dict = set_num_epochs(num_epochs, self.system_dict);
        
        self.custom_print("Update: Num Epochs - {}".format(self.system_dict["hyper-parameters"]["num_epochs"]));
        self.custom_print("");
    ##########################################################################################################################################################



    ##########################################################################################################################################################
    def update_learning_rate(self, learning_rate):
        '''
        Update base learning rate for training

        Args:
            learning_rate (float): New base learning rate

        Returns:
            None
        '''
        self.system_dict["hyper-parameters"]["learning_rate"] = learning_rate;
        self.system_dict["hyper-parameters"]["optimizer"]["params"]["lr"] = learning_rate;
        
        self.custom_print("Update: Learning Rate - {}".format(self.system_dict["hyper-parameters"]["learning_rate"]));
        self.custom_print("");
    ##########################################################################################################################################################



    ##########################################################################################################################################################
    def update_display_progress_realtime(self, value):  
        '''
        Update display progress param

        Args:
            value (bool): If True, then real time progress is displayed
            
        Returns:
            None
        '''  
        self.system_dict = set_display_progress_realtime(value, self.system_dict);
        
        self.custom_print("Update: Display progress realtime - {}".format(self.system_dict["training"]["settings"]["display_progress_realtime"]));
        self.custom_print("");
    ##########################################################################################################################################################
        

    ##########################################################################################################################################################
    def update_display_progress(self, value): 
        '''
        Update display progress param

        Args:
            value (bool): If True, then per epoch progress is displayed
            
        Returns:
            None
        ''' 
        self.system_dict = set_display_progress(value, self.system_dict);
        
        self.custom_print("Update: Display progress  - {}".format(self.system_dict["training"]["settings"]["display_progress"]));
        self.custom_print("");
    ##########################################################################################################################################################


        
    ##########################################################################################################################################################
    def update_save_intermediate_models(self, value, prefix="intermediate_model_"): 
        '''
        Update whether to save intermediate models or not

        Args:
            value (bool): If True, saves model weight post every epoch
            prefix (str): Appends a prefix to intermediate weights
            
        Returns:
            None
        ''' 
        if(value):
            if(not os.access(self.system_dict["model_dir"], os.W_OK)):
                msg = "Folder \"{}\" has no read access".format(self.system_dict["model_dir"])
                msg += "Cannot save Intermediate models";
                raise ConstraintError(msg);

        self.system_dict = set_save_intermediate_models(value, self.system_dict);
        self.system_dict = set_intermediate_model_prefix(prefix, self.system_dict);
        
        self.custom_print("Update: Save Intermediate models - {}".format(self.system_dict["training"]["settings"]["save_intermediate_models"]));
        if(self.system_dict["training"]["settings"]["save_intermediate_models"]):
            self.custom_print("Update: Intermediate model prefix - {}".format(self.system_dict["training"]["settings"]["intermediate_model_prefix"]));
        self.custom_print("");
    ##########################################################################################################################################################



    ##########################################################################################################################################################
    def update_save_training_logs(self, value):
        '''
        Update whether to save training logs or not

        Args:
            value (bool): If True, saves all training and validation metrics. Required for comparison.
            
        Returns:
            None
        '''
        self.system_dict = set_save_training_logs(value, self.system_dict);
        
        self.custom_print("Update: Save Training logs - {}".format(self.system_dict["training"]["settings"]["save_training_logs"]));
        self.custom_print("");

Ancestors

• gluon.finetune.level_12_losses_main.prototype_losses
• gluon.finetune.level_11_optimizers_main.prototype_optimizers
• gluon.finetune.level_10_schedulers_main.prototype_schedulers
• gluon.finetune.level_9_transforms_main.prototype_transforms
• gluon.finetune.level_8_layers_main.prototype_layers
• gluon.finetune.level_7_aux_main.prototype_aux
• gluon.finetune.level_6_params_main.prototype_params
• gluon.finetune.level_5_state_base.finetune_state
• gluon.finetune.level_4_evaluation_base.finetune_evaluation
• gluon.finetune.level_3_training_base.finetune_training
• gluon.finetune.level_2_model_base.finetune_model
• gluon.finetune.level_1_dataset_base.finetune_dataset
• system.base_class.system

Methods

def update_batch_size(self, batch_size)

Update batch size.

Args

batch_size : int

New batch size

Returns

None

 

Expand source code

def update_batch_size(self, batch_size):
    '''
    Update batch size.

    Args:
        batch_size (int): New batch size

    Returns:
        None
    '''
    self.system_dict = set_batch_size(batch_size, self.system_dict);
    
    self.custom_print("Update: Batch size - {}".format(self.system_dict["dataset"]["params"]["batch_size"]));
    self.custom_print("");

def update_dataset(self, dataset_path=False, split=0.9, path_to_csv=False, delimiter=',')

Update dataset path

Args

dataset_path : str, list

Path to Dataset folder 1) Single string if validation data does not exist 2) List [train_path, val_path] in case of separate train and val data

path_to_csv : str, list

Path to csv file pointing towards images 1) Single string if validation data does not exist 2) List [train_path, val_path] in case of separate train and val data

value : float

Indicating train validation split Division happens as follows: train - total dataset * split * 100 val - total dataset * (1-split) * 100

delimiter : str

Delimiter for csv file

Returns

None

 

Expand source code

def update_dataset(self, dataset_path=False, split=0.9, path_to_csv=False, delimiter=","):
    '''
    Update dataset path

    Args:
        dataset_path (str, list): Path to Dataset folder
                                  1) Single string if validation data does not exist
                                  2) List [train_path, val_path] in case of separate train and val data
        path_to_csv (str, list): Path to csv file pointing towards images
                                 1) Single string if validation data does not exist
                                 2) List [train_path, val_path] in case of separate train and val data
        value (float): Indicating train validation split
                        Division happens as follows:
                            train - total dataset * split * 100
                            val   - total dataset * (1-split) * 100 
        delimiter (str): Delimiter for csv file

    Returns:
        None
    '''
    self.system_dict = set_dataset_train_path(self.system_dict, dataset_path, split, path_to_csv, delimiter);

def update_display_progress(self, value)

Update display progress param

Args

value : bool

If True, then per epoch progress is displayed

Returns

None

 

Expand source code

def update_display_progress(self, value): 
    '''
    Update display progress param

    Args:
        value (bool): If True, then per epoch progress is displayed
        
    Returns:
        None
    ''' 
    self.system_dict = set_display_progress(value, self.system_dict);
    
    self.custom_print("Update: Display progress  - {}".format(self.system_dict["training"]["settings"]["display_progress"]));
    self.custom_print("");

def update_display_progress_realtime(self, value)

Update display progress param

Args

value : bool

If True, then real time progress is displayed

Returns

None

 

Expand source code

def update_display_progress_realtime(self, value):  
    '''
    Update display progress param

    Args:
        value (bool): If True, then real time progress is displayed
        
    Returns:
        None
    '''  
    self.system_dict = set_display_progress_realtime(value, self.system_dict);
    
    self.custom_print("Update: Display progress realtime - {}".format(self.system_dict["training"]["settings"]["display_progress_realtime"]));
    self.custom_print("");

def update_freeze_base_network(self, freeze)

Update whether freeze base network or not

Args

freeze : bool

If True, then base network is non-trainable, works as a feature extractor

Returns

None

 

Expand source code

def update_freeze_base_network(self, freeze):
    '''
    Update whether freeze base network or not

    Args:
        freeze (bool): If True, then base network is non-trainable, works as a feature extractor

    Returns:
        None
    '''
    self.system_dict = set_freeze_base_network(freeze, self.system_dict);
    
    self.custom_print("Update: Freeze Base Network - {}".format(self.system_dict["model"]["params"]["freeze_base_network"]));
    self.custom_print("");

def update_freeze_layers(self, num_freeze)

Update to freeze certain layers in the network

Args

num_freeze : int

Number of layers to freeze in network starting from top

Returns

None

 

Expand source code

def update_freeze_layers(self, num_freeze):
    '''
    Update to freeze certain layers in the network

    Args:
        num_freeze (int): Number of layers to freeze in network starting from top

    Returns:
        None
    '''
    self.system_dict["model"]["params"]["num_freeze"] = num_freeze;
    
    self.custom_print("Update: Freeze layers - {}".format(self.system_dict["model"]["params"]["num_freeze"]));
    self.custom_print("");

def update_input_size(self, input_size)

Update input size.

Args

input_size : int

New input size

Returns

None

 

Expand source code

def update_input_size(self, input_size):
    '''
    Update input size.

    Args:
        input_size (int): New input size

    Returns:
        None
    '''
    self.system_dict = set_input_size(input_size, self.system_dict);
    
    self.custom_print("Update: Input size - {}".format(self.system_dict["dataset"]["params"]["input_size"]));
    self.custom_print("");

def update_learning_rate(self, learning_rate)

Update base learning rate for training

Args

learning_rate : float

New base learning rate

Returns

None

 

Expand source code

def update_learning_rate(self, learning_rate):
    '''
    Update base learning rate for training

    Args:
        learning_rate (float): New base learning rate

    Returns:
        None
    '''
    self.system_dict["hyper-parameters"]["learning_rate"] = learning_rate;
    self.system_dict["hyper-parameters"]["optimizer"]["params"]["lr"] = learning_rate;
    
    self.custom_print("Update: Learning Rate - {}".format(self.system_dict["hyper-parameters"]["learning_rate"]));
    self.custom_print("");

def update_model_name(self, model_name, force=False)

Update model name

Args

model_name : str

Select from available models. Check via List_Models() function

force : bool

Dummy function

Returns

None

 

Expand source code

def update_model_name(self, model_name, force=False):
    '''
    Update model name

    Args:
        model_name (str): Select from available models. Check via List_Models() function
        force (bool): Dummy function

    Returns:
        None
    '''
    if(not force):
        if(self.system_dict["training"]["status"]):
            ConstraintWarning("Model trained using {}\n".format(self.system_dict["model"]["params"]["model_name"]));
            ConstraintWarning("Changing the model will overwrite previously trained models if training is executed.\n");
            inp = input("Do you wish to continue further (y/n):");

            if(inp == "y"):
                self.system_dict = set_model_name(model_name, self.system_dict);
                self.custom_print("Update: Model name - {}".format(self.system_dict["model"]["params"]["model_name"]));
                self.custom_print("");
            else:
                self.custom_print("Model not updated.");
                self.custom_print("");
        else:
            self.system_dict = set_model_name(model_name, self.system_dict);
            self.custom_print("Update: Model name - {}".format(self.system_dict["model"]["params"]["model_name"]));
            self.custom_print("");
    else:
        self.system_dict = set_model_name(model_name, self.system_dict);
        self.custom_print("Update: Model name - {}".format(self.system_dict["model"]["params"]["model_name"]));
        self.custom_print("");

def update_model_path(self, model_path, force=False)

Update model path for inferencing

Args

model_path : str

Path to model weights.

force : bool

Dummy function

Returns

None

 

Expand source code

def update_model_path(self, model_path, force=False):
    '''
    Update model path for inferencing

    Args:
        model_path (str): Path to model weights.
        force (bool): Dummy function

    Returns:
        None
    '''
    if(not force):
        if(self.system_dict["training"]["status"]):
            ConstraintWarning("Model trained using {}\n".format(self.system_dict["model"]["params"]["model_name"]));
            ConstraintWarning("Changing the model will overwrite previously trained models if training is executed.\n");
            inp = input("Do you wish to continue further (y/n):");

            if(inp == "y"):
                self.system_dict = set_model_path(model_path, self.system_dict);
                self.custom_print("Update: Model path - {}".format(self.system_dict["model"]["params"]["model_path"]));
                self.custom_print("");
            else:
                self.custom_print("Model not updated.");
                self.custom_print("");
        else:
            self.system_dict = set_model_path(model_path, self.system_dict);
            self.custom_print("Update: Model path - {}".format(self.system_dict["model"]["params"]["model_path"]));
            self.custom_print("");
    else:
        self.system_dict = set_model_path(model_path, self.system_dict);
        self.custom_print("Update: Model path - {}".format(self.system_dict["model"]["params"]["model_path"]));
        self.custom_print("");               

def update_num_epochs(self, num_epochs)

Update number of epochs to train the network

Args

num_epochs : int

New number of epochs

Returns

None

 

Expand source code

def update_num_epochs(self, num_epochs):
    '''
    Update number of epochs to train the network

    Args:
        num_epochs (int): New number of epochs

    Returns:
        None
    '''
    self.system_dict = set_num_epochs(num_epochs, self.system_dict);
    
    self.custom_print("Update: Num Epochs - {}".format(self.system_dict["hyper-parameters"]["num_epochs"]));
    self.custom_print("");

def update_num_processors(self, num_processors)

Update num processors for data loader.

Args

num_processors : int

Max CPUs for data sampling

Returns

None

 

Expand source code

def update_num_processors(self, num_processors):
    '''
    Update num processors for data loader.

    Args:
        num_processors (int): Max CPUs for data sampling

    Returns:
        None
    '''
    self.system_dict = set_num_processors(num_processors, self.system_dict);
    
    self.custom_print("Update: Num processors - {}".format(self.system_dict["dataset"]["params"]["num_workers"]));
    self.custom_print("");

def update_save_intermediate_models(self, value, prefix='intermediate_model_')

Update whether to save intermediate models or not

Args

value : bool

If True, saves model weight post every epoch

prefix : str

Appends a prefix to intermediate weights

Returns

None

 

Expand source code

def update_save_intermediate_models(self, value, prefix="intermediate_model_"): 
    '''
    Update whether to save intermediate models or not

    Args:
        value (bool): If True, saves model weight post every epoch
        prefix (str): Appends a prefix to intermediate weights
        
    Returns:
        None
    ''' 
    if(value):
        if(not os.access(self.system_dict["model_dir"], os.W_OK)):
            msg = "Folder \"{}\" has no read access".format(self.system_dict["model_dir"])
            msg += "Cannot save Intermediate models";
            raise ConstraintError(msg);

    self.system_dict = set_save_intermediate_models(value, self.system_dict);
    self.system_dict = set_intermediate_model_prefix(prefix, self.system_dict);
    
    self.custom_print("Update: Save Intermediate models - {}".format(self.system_dict["training"]["settings"]["save_intermediate_models"]));
    if(self.system_dict["training"]["settings"]["save_intermediate_models"]):
        self.custom_print("Update: Intermediate model prefix - {}".format(self.system_dict["training"]["settings"]["intermediate_model_prefix"]));
    self.custom_print("");

def update_save_training_logs(self, value)

Update whether to save training logs or not

Args

value : bool

If True, saves all training and validation metrics. Required for comparison.

Returns

None

 

Expand source code

def update_save_training_logs(self, value):
    '''
    Update whether to save training logs or not

    Args:
        value (bool): If True, saves all training and validation metrics. Required for comparison.
        
    Returns:
        None
    '''
    self.system_dict = set_save_training_logs(value, self.system_dict);
    
    self.custom_print("Update: Save Training logs - {}".format(self.system_dict["training"]["settings"]["save_training_logs"]));
    self.custom_print("");

def update_shuffle_data(self, shuffle)

Update to shuffle data or not.

Args

shuffle : bool

If True, will shuffle data

Returns

None

 

Expand source code

def update_shuffle_data(self, shuffle):
    '''
    Update to shuffle data or not.

    Args:
        shuffle (bool): If True, will shuffle data

    Returns:
        None
    '''
    self.system_dict = set_data_shuffle(shuffle, self.system_dict);
    
    self.custom_print("Update: Data shuffle - {}".format(self.system_dict["dataset"]["params"]["train_shuffle"]));
    self.custom_print("");

def update_trainval_split(self, value)

Update training-validation split

Args

value : float

Indicating train validation split Division happens as follows: train - total dataset * split * 100 val - total dataset * (1-split) * 100

Returns

None

 

Expand source code

def update_trainval_split(self, value):
    '''
    Update training-validation split
    Args:
        value (float): Indicating train validation split
                        Division happens as follows:
                            train - total dataset * split * 100
                            val   - total dataset * (1-split) * 100 

    Returns:
        None
    '''
    if(self.system_dict["dataset"]["dataset_type"] == "train"):
        dataset_path = self.system_dict["dataset"]["train_path"];
        path_to_csv=False;
    elif(self.system_dict["dataset"]["dataset_type"] == "train-val"):
        dataset_path = [self.system_dict["dataset"]["train_path"], self.system_dict["dataset"]["val_path"]];
        path_to_csv=False;
    elif(self.system_dict["dataset"]["dataset_type"] == "csv_train"):
        dataset_path = self.system_dict["dataset"]["train_path"];
        path_to_csv = self.system_dict["dataset"]["csv_train"];
    elif(self.system_dict["dataset"]["dataset_type"] == "csv_train-val"):
        dataset_path = [self.system_dict["dataset"]["train_path"], self.system_dict["dataset"]["val_path"]];
        path_to_csv = [self.system_dict["dataset"]["csv_train"], self.system_dict["dataset"]["csv_val"]];
    else:
        msg = "Dataset Type invalid.\n";
        msg += "Cannot update split"
        ConstraintsWarning(msg)

    self.system_dict = set_dataset_train_path(self.system_dict, dataset_path, value, path_to_csv, self.system_dict["dataset"]["params"]["delimiter"]);

def update_use_gpu(self, gpu)

Update to use gpu or cpu

Args

gpu : bool

If True, then use GPU

Returns

None

 

Expand source code

def update_use_gpu(self, gpu):
    '''
    Update to use gpu or cpu

    Args:
        gpu (bool): If True, then use GPU

    Returns:
        None
    '''
    self.system_dict = set_device(gpu, self.system_dict);
    
    self.custom_print("Update: Use Gpu - {}".format(self.system_dict["model"]["params"]["use_gpu"]));
    self.custom_print("");

def update_use_pretrained(self, pretrained)

Update to use pretrained wights or randomly initialized weights

Args

pretrained : bool

If True, use pretrained weights else, use randomly initialized weights

Returns

None

 

Expand source code

def update_use_pretrained(self, pretrained):
    '''
    Update to use pretrained wights or randomly initialized weights

    Args:
        pretrained (bool): If True, use pretrained weights
                            else, use randomly initialized weights

    Returns:
        None
    '''
    self.system_dict = set_pretrained(pretrained, self.system_dict);
    
    self.custom_print("Update: Use pretrained - {}".format(self.system_dict["model"]["params"]["use_pretrained"]));
    self.custom_print("");

def update_weighted_sampling(self, sample)

Function inactive

Expand source code

def update_weighted_sampling(self, sample):
    '''
    Function inactive
    '''
    self.system_dict = set_weighted_sampling(sample, self.system_dict);
    
    self.custom_print("Update: Weighted Sampling - {}".format(self.system_dict["dataset"]["params"]["weighted_sample"]));
    self.custom_print("");