simple_intent_classifier_pytorch

class besser.agent.nlp.intent_classifier.simple_intent_classifier_pytorch.SimpleIntentClassifierTorch(nlp_engine, state)

Bases: IntentClassifier

A Simple Pytorch-based Intent Classifier.

It works using a simple Neural Network (the prediction model) for text classification.

Parameters:

• nlp_engine (NLPEngine) – the NLPEngine that handles the NLP processes of the agent

• state (State) – the state the intent classifier belongs to

_model

The intent classifier language model

Type:

torch.nn.Module

See also

SimpleIntentClassifierConfiguration.

PAD = '<PAD>'

UNK = '<UNK>'

__intent_label_mapping

A mapping of the intent labels and their corresponding intents.

__total_labels

The label (identifying the intent) of all training sentences.

__total_labels_encoded

The encoded label of all training sentences.

__total_training_sentences

All the processed training sentences of all intents of the intent classifier’s state.

__total_training_sequences

All the training sequences of all intents of the intent classifier’s state.

__vocab

The vocabulary of the intent classifier (i.e., all known tokens).

_abc_impl = <_abc._abc_data object>

predict(message)

Predict the intent of a given message.

Instead of returning only the intent with the highest likelihood, return all predictions. Predictions include not only the intent scores but other information extracted from the message.

Parameters:

message (str) – the message to predict the intent

Returns:

the list of predictions made by the intent classifier.

Return type:

list[IntentClassifierPrediction]

train()

Train the intent classifier.

class besser.agent.nlp.intent_classifier.simple_intent_classifier_pytorch.TextClassifier(vocab_size, embed_dim, hidden_dim, output_dim, pad_idx, activation_last_layer)

Bases: Module

forward(x)

Define the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

class besser.agent.nlp.intent_classifier.simple_intent_classifier_pytorch.TextDataset(texts, labels, vocab, max_len, language)

Bases: Dataset