def validate(model,
x,
y_true,
input_len,
label_len,
y_strings,
test=False,
save_file=None):
input_len = np.expand_dims(input_len, axis=1)
label_len = np.expand_dims(label_len, axis=1)
y_pred = model(x)
loss = ctc_batch_cost(y_true, y_pred, input_len, label_len)
input_len = np.squeeze(input_len)
y_decode = ctc_decode(y_pred, input_len)[0][0]
accuracy = 0.0
for i in range(len(y_strings)):
predicted_sentence = indices_to_string(y_decode[i].numpy())
accuracy += wer(predicted_sentence, y_strings[i])
if test:
save_file.write("Correct Sentence:" + str(y_strings[i]) + "\n")
save_file.write("Predicted Sentence:" + predicted_sentence + "\n")
return tf.reduce_mean(loss), accuracy / len(y_strings)
def train_one_step(model, optimizer, x, y_true, input_len, label_len,
y_strings):
input_len = np.expand_dims(input_len, axis=1)
label_len = np.expand_dims(label_len, axis=1)
with tf.GradientTape() as tape:
y_pred = model(x)
loss = ctc_batch_cost(y_true, y_pred, input_len, label_len)
grads = tape.gradient(loss, model.trainable_variables)
optimizer.apply_gradients(zip(grads, model.trainable_variables))
input_len = np.squeeze(input_len)
y_decode = ctc_decode(y_pred, input_len)[0][0]
accuracy = 0.0
for i in range(len(y_strings)):
predicted_sentence = indices_to_string(y_decode[i].numpy())
accuracy += wer(predicted_sentence, y_strings[i])
return tf.reduce_mean(loss), accuracy / len(y_strings)
generate_batches(dataset, batch_size=128)):
y_pred = generater(batch_dict['x_data'])
loss = utils.sequence_loss(y_pred, batch_dict['y_target'],
vectorizer.char_vocab.mask_index)
acc = utils.compute_accuracy(y_pred, batch_dict['y_target'],
vectorizer.char_vocab.mask_index)
loss_epoch += (loss.item() - loss_epoch) / (batch_index + 1)
acc_epoch += (acc - acc_epoch) / (batch_index + 1)
index_view = torch.randint(low=0,
high=len(batch_dict['x_data']),
size=(1, )).item()
print("{}\n{}".format(
utils.indices_to_string(batch_dict['x_data'][index_view].numpy(),
vectorizer.char_vocab),
utils.indices_to_string(
torch.argmax(torch.softmax(y_pred, dim=2)[index_view],
dim=1).numpy(), vectorizer.char_vocab)))
print("epoch {}: train loss {}, acc: {:.2f}\n\n".format(
epoch_index, loss_epoch, acc_epoch))
samples_indices = utils.sample_from_model(model=generater,
vectorizer=vectorizer,
num_samples=10,
sample_size=10,
temperature=0.5)
print("Sample generate: ")
for indices in samples_indices: