def train_eval(x_train, x_test, is_peeky): if is_peeky: model = PeekySeq2seq(vocab_size, wordvec_size, hidden_size) else: model = Seq2seq(vocab_size, wordvec_size, hidden_size) optimizer = Adam() trainer = Trainer(model, optimizer) acc_list = [] for epoch in range(max_epoch): trainer.fit(x_train, t_train, max_epoch=1, batch_size=batch_size, max_grad=max_grad) correct_num = 0 for i in range(len(x_test)): question, correct = x_test[[i]], t_test[[i]] verbose = i < 10 correct_num += eval_seq2seq(model, question, correct, id_to_char, verbose) acc = float(correct_num) / len(x_test) acc_list.append(acc) print('val acc %.3f%%' % (acc * 100)) return acc_list
def main(): # データの読み込み (x_train, t_train), (x_test, t_test) = sequence.load_data('date.txt') char_to_id, id_to_char = sequence.get_vocab() # 入力文を反転 x_train, x_test = x_train[:, ::-1], x_test[:, ::-1] # ハイパーパラメータの設定 vocab_size = len(char_to_id) wordvec_size = 16 hidden_size = 256 batch_size = 128 max_epoch = 10 max_grad = 5.0 model = AttentionSeq2seq(vocab_size, wordvec_size, hidden_size) optimizer = Adam() trainer = Trainer(model, optimizer) acc_list = [] for epoch in range(max_epoch): trainer.fit(x_train, t_train, max_epoch=1, batch_size=batch_size, max_grad=max_grad) correct_num = 0 for i in range(len(x_test)): question, correct = x_test[[i]], t_test[[i]] verbose = i < 10 correct_num += eval_seq2seq(model, question, correct, id_to_char, verbose, is_reverse=True) acc = float(correct_num) / len(x_test) acc_list.append(acc) print('val acc %.3f%%' % (acc * 100))
def main(): # データセットの読み込み (x_train, t_train), (x_test, t_test) = sequence.load_data('addition.txt') x_train, x_test = x_train[:, ::-1], x_test[:, ::-1] char_to_id, id_to_char = sequence.get_vocab() # ハイパーパラメータの設定 vocab_size = len(char_to_id) wordvec_size = 16 hidden_size = 128 batch_size = 128 max_epoch = 25 max_grad = 5.0 # モデル / オプティマイザ / トレーナーの生成 # model = Seq2seq(vocab_size, wordvec_size, hidden_size) model = PeekySeq2seq(vocab_size, wordvec_size, hidden_size) optimizer = Adam() trainer = Trainer(model, optimizer) acc_list = [] for epoch in range(max_epoch): trainer.fit(x_train, t_train, max_epoch=1, batch_size=batch_size, max_grad=max_grad) correct_num = 0 for i in range(len(x_test)): question, correct = x_test[[i]], t_test[[i]] verbose = i < 10 correct_num += eval_seq2seq(model, question, correct, id_to_char, verbose) acc = float(correct_num) / len(x_test) acc_list.append(acc) print('val acc %.3f%%' % (acc * 100))
acc_list = [] for epoch in range(max_epoch): trainer.fit(x_train, t_train, max_epoch=1, batch_size=batch_size, max_grad=max_grad) correct_num = 0 for i in range(len(x_test)): question, correct = x_test[[i]], t_test[[i]] verbose = i < 10 correct_num += eval_seq2seq(model, question, correct, id_to_char, verbose, is_reverse=True) acc = float(correct_num) / len(x_test) acc_list.append(acc) print('val acc %.3f%%' % (acc * 100)) model.save_params() # グラフの描画 x = np.arange(len(acc_list)) plt.plot(x, acc_list, marker='o') plt.xlabel('epochs') plt.ylabel('accuracy') plt.ylim(-0.05, 1.05) plt.show()
wordvec_size = 16 hidden_size = 128 batch_size = 128 max_epoch = 25 max_grad = 5.0 # 모델 / 옵티마이저 / 트레이너 생성 # model = Seq2Seq(vocab_size, wordvec_size, hidden_size) model = PeekySeq2seq(vocab_size, wordvec_size, hidden_size) optimizer = Adam() trainer = Trainer(model, optimizer) acc_list = [] for epoch in range(max_epoch): trainer.fit(x_train, t_train, max_epoch=1, batch_size=batch_size, max_grad=max_grad) correct_num = 0 for i in range(len(x_test)): question, correct = x_test[[i]], t_test[[i]] verbose = i < 10 correct_num += eval_seq2seq(model, question, correct, id_to_char, verbose) acc = float(correct_num) / len(x_test) acc_list.append(acc) print('검증 정확도 %.3f%%' % (acc * 100))
for epoch in range(max_epoch): trainer.fit(x_train, t_train, max_epoch=1, batch_size=batch_size, max_grad=max_grad) model.save_params('../pkl/myAttentionSeq2seq3.pkl') correct_num = 0 for i in range(len(x_test)): question, correct = x_test[[i]], t_test[[i]] verbose = i < 10 correct_num += eval_seq2seq(model, question, correct, preprocessing.id_to_word, verbose, is_reverse=False) acc = float(correct_num) / len(x_test) acc_list.append(acc) print('정확도 %.3f%%' % (acc * 100)) # 그래프 그리기 x = np.arange(len(acc_list)) plt.plot(x, acc_list, marker='o') plt.xlabel('에폭') plt.ylabel('정확도') plt.ylim(-0.05, 1.05) plt.show()