c5504174948be1855a9636a416524cc7cb9dabca,examples/text_classification/tutorial_imdb_fasttext_dynamic_eager.py,,train_test_and_save_model,#,119

Before Change

                // forward and define the loss function
                with tf.GradientTape() as tape:
                    y_pred = model(tl.prepro.pad_sequences(X_batch)).outputs
                    cost = tl.cost.cross_entropy(y_pred, y_batch, name="cost")

                // backward, calculate gradients and update the weights
                grad = tape.gradient(cost, model.weights)
                optimizer.apply_gradients(zip(grad, model.weights))

                // calculate the accuracy
                predictions = tf.argmax(y_pred, axis=1, output_type=tf.int32)
                are_predictions_correct = tf.equal(predictions, y_batch)
                accuracy = tf.reduce_mean(tf.cast(are_predictions_correct, tf.float32))

                train_accuracy.append(accuracy)
                if len(train_accuracy) % N_STEPS_TO_PRINT == 0:

After Change

                // forward and define the loss function
                with tf.GradientTape() as tape:
                    y_pred = model(tl.prepro.pad_sequences(X_batch))
                    cost = tl.cost.cross_entropy(y_pred, y_batch, name="cost")

                // backward, calculate gradients and update the weights
                grad = tape.gradient(cost, model.weights)
                optimizer.apply_gradients(zip(grad, model.weights))

                // calculate the accuracy
                predictions = tf.argmax(y_pred, axis=1, output_type=tf.int32)
                are_predictions_correct = tf.equal(predictions, y_batch)
                accuracy = tf.reduce_mean(tf.cast(are_predictions_correct, tf.float32))

                train_accuracy.append(accuracy)
                if len(train_accuracy) % N_STEPS_TO_PRINT == 0:

Instances