Neural Network Classifier¶
This classifier is based on Neural Network Model
-
class
sentivi.classifier.nn_clf.NeuralNetworkClassifier(num_labels: int = 3, embedding_size: Optional[int] = None, max_length: Optional[int] = None, device: Optional[str] = 'cpu', num_epochs: Optional[int] = 10, learning_rate: Optional[float] = 0.001, batch_size: Optional[int] = 2, shuffle: Optional[bool] = True, random_state: Optional[int] = 101, hidden_size: Optional[int] = 512, num_workers: Optional[int] = 2, *args, **kwargs)¶ Neural Network Classifier
-
__init__(num_labels: int = 3, embedding_size: Optional[int] = None, max_length: Optional[int] = None, device: Optional[str] = 'cpu', num_epochs: Optional[int] = 10, learning_rate: Optional[float] = 0.001, batch_size: Optional[int] = 2, shuffle: Optional[bool] = True, random_state: Optional[int] = 101, hidden_size: Optional[int] = 512, num_workers: Optional[int] = 2, *args, **kwargs)¶ Neural Network Classifier
- Parameters
num_labels – number of polarities
embedding_size – input embedding size
max_length – maximum number of input text
device – training device
num_epochs – maximum number of epochs
learning_rate – training learning rate
batch_size – training batch size
shuffle – whether DataLoader is shuffle or not
random_state – random.seed
hidden_size – hidden size
num_workers – number of DataLoader workers
args – arbitrary arguments
kwargs – arbitrary keyword arguments
-
static
compute_metrics(preds, targets, eval=False)¶ Compute accuracy and F1
- Parameters
preds – prediction output
targets – ground-truth value
eval – whether is eval or not
- Returns
-
fit(*args, **kwargs)¶ Feed-forward network
- Parameters
args – arbitrary arguments
kwargs – arbitrary keyword arguments
- Returns
-
forward(data, *args, **kwargs)¶ Training and evaluating NeuralNetworkClassifier
- Parameters
data – TextEncoder output
args – arbitrary arguments
kwargs – arbitrary keyword arguments
- Returns
training and evaluating results
- Return type
str
-
get_overall_result(loader)¶ Get overall result
- Parameters
loader – DataLoader
- Returns
overall result
- Return type
str
-
load(model_path, *args, **kwargs)¶ Load model from disk
- Parameters
model_path – path to model path
args – arbitrary arguments
kwargs – arbitrary keyword arguments
- Returns
-
save(save_path, *args, **kwargs)¶ Save model to disk
- Parameters
save_path – path to saved model
args – arbitrary arguments
kwargs – arbitrary keyword arguments
- Returns
-