src.core.rnn package

Submodules

src.core.rnn.dataset module

class src.core.rnn.dataset.RNNDataset(config_dict)

Bases: Word2VecDataset

RNN Dataset

Parameters:

config_dict (dict) – Config Params Dictionary

get_data()

Generates tokens and labels from extracted data

Returns:

Input tokens, Labels

Return type:

tuple (numpy.ndarray [num_samples, seq_len], numpy.ndarray [num_samples, num_classes])

get_test_data()

Generates test tokens and labels from extracted data

Returns:

Input tokens, Labels

Return type:

tuple (numpy.ndarray [num_samples, seq_len], numpy.ndarray [num_samples, num_classes])

pad_slice_text(text_ls)

Pads and slices text to seq_len tokens

Parameters:

text_ls (list) – List of text samples

Returns:

list of preprocessed text

Return type:

list

src.core.rnn.dataset.create_dataloader(X, y, val_split=0.2, batch_size=32, seed=2024)

Creates Train, Validation DataLoader

Parameters:

• X (torch.Tensor (num_samples, seq_len)) – Input tokens

• y (torch.Tensor (num_samples, num_classes)) – Output Labels

• val_split (float, optional) – validation split, defaults to 0.2

• batch_size (int, optional) – Batch size, defaults to 32

• seed (int, optional) – Seed, defaults to 2024

Returns:

Train, Val dataloaders

Return type:

tuple (torch.utils.data.DataLoader, torch.utils.data.DataLoader)

src.core.rnn.model module

class src.core.rnn.model.RNNCell(h_dim, inp_x_dim, out_x_dim)

Bases: Module

RNN Cell

Parameters:

• h_dim (int) – Hidden state vector dimension

• inp_x_dim (int) – Input vector dimension

• out_x_dim (int) – Output vector dimension

forward(ht_1, xt)

Forward propogation

Parameters:

• ht_1 (torch.Tensor (batch_size, h_dim)) – Hidden state vector

• xt (torch.Tensor (batch_size, embed_dim)) – Input vector

Returns:

New hidden states, output

Return type:

tuple (torch.Tensor [batch_size, h_dim], torch.Tensor [batch_size, out_dim])

class src.core.rnn.model.RNNModel(config_dict)

Bases: Module

RNN Architecture

Parameters:

config_dict (dict) – Config Params Dictionary

forward(X)

Forward propogation

Parameters:

X (torch.Tensor (batch_size, seq_len)) – Input tokens

Returns:

Prediction labels

Return type:

torch.Tensor (batch_size, num_classes)

init_hidden()

Initialized hidden states

Returns:

List of hidden states

Return type:

list

class src.core.rnn.model.RNNTrainer(model, optimizer, config_dict)

Bases: Module

RNN Trainer

Parameters:

• model (torch.nn.Module) – RNN model

• optimizer (torch.optim) – Optimizer

• config_dict (dict) – Config Params Dictionary

fit(train_loader, val_loader)

Fits the model on dataset. Runs training and Validation steps for given epochs and saves best model based on the evaluation metric

Parameters:

• train_loader (torch.utils.data.DataLoader) – Train Data loader

• val_loader (torch.utils.data.DataLoader) – Validaion Data Loader

Returns:

Training History

Return type:

dict

predict(data_loader)

Runs inference to predict a translation of soruce sentence

Parameters:

data_loader (torch.utils.data.DataLoader) – Infer Data loader

Returns:

Predicted tokens

Return type:

numpy.ndarray (num_samples, num_classes)

train_one_epoch(data_loader, epoch)

Train step

Parameters:

• data_loader (torch.utils.data.Dataloader) – Train Data Loader

• epoch (int) – Epoch number

Returns:

Train Losse, Train Metrics

Return type:

tuple (torch.float32, dict)

val_one_epoch(data_loader)

Validation step

Parameters:

data_loader (torch.utils.data.Dataloader) – Validation Data Loader

Returns:

Validation Losse, Validation Metrics

Return type:

tuple (torch.float32, dict)

src.core.rnn.rnn module

class src.core.rnn.rnn.RNN(config_dict)

Bases: object

A class to run RNN data preprocessing, training and inference

Parameters:

config_dict (dict) – Config Params Dictionary

run()

Runs Seq2Seq Training and saves output

run_infer()

Runs inference

Returns:

True and predicted captions

Return type:

tuple (torch.Tensor [batch_size, seq_len, num_classes], torch.Tensor [batch_size, seq_len, num_classes])

save_output()

Saves Training and Inference results

Module contents