class ClassifierPredictor:
def __init__(self, model: Classifier) -> None:
self.model = model
self.reader = ClassificationReader(skip_start_end=True)
self.predictor = TextClassifierPredictor(self.model, self.reader)
def predict(self, sequences: List[str]) -> np.ndarray:
probs = [self.predictor.predict(seq)['probs'] for seq in sequences]
probs = np.array(probs)
return probs
trainer = Trainer(
model=model,
optimizer=optimizer,
iterator=iterator,
train_dataset=train_data,
validation_dataset=validation_data,
patience=2, # early stopping if it is stuck for 2 epochs
num_epochs=
10, # we should increase the number of epoch later, it's just for early try
cuda_device=cuda_device)
trainer.train()
# predictor
pre_example = 'Good morning'
predictor = TextClassifierPredictor()
pre = predictor.predict(pre_example)
# save the model
with open('./tmp/classifier_biattention_model.th', 'wb') as f:
torch.save(model.state_dict(), f)
# save the vocabulary
vocab.save_to_files('./tmp/vocabulary')
# reload the model
# vocab2 = Vocabulary.from_files('./tmp/vocabulary')
# model2 = BiattentiveClassificationNetwork(word_embeddings, encoder, vocab2)
# with open('./tmp/classifier_biattention_model.th', 'rb') as f:
# model2.load_state_dict(torch.load(f))
# if cuda_device > -1:
# model2.cuda(cuda_device)