Q&A
TypeError: __init__() got multiple values for argument 'strides'とエラーが出てしまいます。
前提
プログラミングを勉強しているものです。
https://github.com/IDEALLab/bezier-gan
上記のプログラムを実装したいと思っています。
バージョンが違い現在は使われていないコードを書き直したりしてきました（調べながら）
その過程で以下のエラーが発生しました。
発生している問題・エラーメッセージ
エラーメッセージ Traceback (most recent call last): File "C:\Users\phanton\Desktop\?\pycharm_env\python3.7\lib\site-packages\tensorflow_core\python\keras\utils\conv_utils.py", line 80, in normalize_tuple int(single_value) TypeError: int() argument must be a string, a bytes-like object or a number, not 'tuple' During handling of the above exception, another exception occurred: Traceback (most recent call last): File "C:/Users/phanton/Desktop/?/bezier-gan/bezier-gan-master/beziergan/train_gan.py", line 60, in <module> model.train(X_train, batch_size=batch_size, train_steps=train_steps, save_interval=args.save_interval, directory=directory) File "C:\Users\phanton\Desktop\?\bezier-gan\bezier-gan-master\beziergan\gan.py", line 262, in train d_real, _ = self.discriminator(self.x) File "C:\Users\phanton\Desktop\?\bezier-gan\bezier-gan-master\beziergan\gan.py", line 173, in discriminator strides=(2, 1), padding='same') File "C:\Users\phanton\Desktop\?\pycharm_env\python3.7\lib\site-packages\tensorflow_core\python\keras\layers\convolutional.py", line 498, in __init__ **kwargs) File "C:\Users\phanton\Desktop\?\pycharm_env\python3.7\lib\site-packages\tensorflow_core\python\keras\layers\convolutional.py", line 126, in __init__ kernel_size, rank, 'kernel_size') File "C:\Users\phanton\Desktop\?\pycharm_env\python3.7\lib\site-packages\tensorflow_core\python\keras\utils\conv_utils.py", line 85, in normalize_tuple ' ' + str(type(single_value))) ValueError: The `kernel_size` argument must be a tuple of 2 integers. Received: (64, (4, 2)) including element (4, 2) of type <class 'tuple'>
該当のソースコード
python
1ソースコード 2 def discriminator(self, x, reuse=tf.compat.v1.AUTO_REUSE, training=True): 3 4 depth = 64 5 dropout = 0.4 6 kernel_size = (4, 2) 7 8 with tf.compat.v1.variable_scope('Discriminator', reuse=reuse): 9 x = tf.keras.layers.Conv2D(x, (depth * 1, kernel_size), 10 strides=(2, 1), padding='same') 11 x = tf.keras.layers.BatchNormalization(x, 12 momentum=0.9) # , training=training) 13 x = tf.nn.leaky_relu(x, alpha=0.2) 14 x = tf.keras.layers.Dropout(x, dropout, training=training) 15 16 x = tf.keras.layers.Conv2D(x, (depth * 2, kernel_size), 17 strides=(2, 1), padding='same') 18 x = tf.keras.layers.BatchNormalization(x, 19 momentum=0.9) # , training=training) 20 x = tf.nn.leaky_relu(x, alpha=0.2) 21 x = tf.keras.layers.dropout(x, dropout, training=training) 22 23 x = tf.keras.layers.Conv2D(x, (depth * 4, kernel_size), 24 strides=(2, 1), padding='same') 25 x = tf.keras.layers.BatchNormalization(x, 26 momentum=0.9) # , training=training) 27 x = tf.nn.leaky_relu(x, alpha=0.2) 28 x = tf.keras.layers.dropout(x, dropout, training=training) 29 30 x = tf.keras.layers.Conv2D(x, (depth * 8, kernel_size), 31 strides=(2, 1), 32 padding='same') 33 x = tf.keras.layers.BatchNormalization(x, 34 momentum=0.9) # , training=training) 35 x = tf.nn.leaky_relu(x, alpha=0.2) 36 x = tf.keras.layers.dropout(x, dropout, training=training) 37 38 x = tf.keras.layers.Conv2D(x, (depth * 16, kernel_size), 39 strides=(2, 1), 40 padding='same') 41 x = tf.keras.layers.BatchNormalization(x, 42 momentum=0.9) # , training=training) 43 x = tf.nn.leaky_relu(x, alpha=0.2) 44 x = tf.keras.layers.dropout(x, dropout, training=training) 45 46 x = tf.keras.layers.Conv2D(x, (depth * 32, kernel_size), 47 strides=(2, 1), 48 padding='same') 49 x = tf.keras.layers.BatchNormalization(x, 50 momentum=0.9) # , training=training) 51 x = tf.nn.leaky_relu(x, alpha=0.2) 52 x = tf.keras.layers.dropout(x, dropout, training=training) 53 54 x = tf.keras.layers.flatten(x) 55 x = tf.keras.layers.dense(x, 1024) 56 x = tf.keras.layers.BatchNormalization(x, 57 momentum=0.9) # , training=training) 58 x = tf.nn.leaky_relu(x, alpha=0.2) 59 60 d = tf.keras.layers.dense(x, 1) 61 62 q = tf.keras.layers.dense(x, 128) 63 q = tf.keras.layers.BatchNormalization(q, 64 momentum=0.9) # , training=training) 65 q = tf.nn.leaky_relu(q, alpha=0.2) 66 q_mean = tf.keras.layers.Dense(q, self.latent_dim) 67 q_logstd = tf.keras.layers.Dense(q, self.latent_dim) 68 q_logstd = tf.maximum(q_logstd, -16) 69 # Reshape to batch_size x 1 x latent_dim 70 q_mean = tf.reshape(q_mean, (-1, 1, self.latent_dim)) 71 q_logstd = tf.reshape(q_logstd, (-1, 1, self.latent_dim)) 72 q = tf.concat([q_mean, q_logstd], axis=1, 73 name='predicted_latent') # batch_size x 2 x latent_dim 74 75 return d, q
どこを変えたら動くようになりますか？
教えていただきたいです
補足情報（FW/ツールのバージョンなど）
python 3.7
tensorflow1.15
keras 2.24
回答2件
0
x = tf.keras.layers.Conv2D(x, (depth * 1, kernel_size),strides=(2, 1), padding='same')
公式ドキュメントより、Conv2Dの引数確認すると
keras.layers.Conv2D(filters, kernel_size, strides=(1, 1), padding='valid', data_format=None, dilation_rate=(1, 1), activation=None, use_bias=True, kernel_initializer='glorot_uniform', bias_initializer='zeros', kernel_regularizer=None, bias_regularizer=None, activity_regularizer=None, kernel_constraint=None, bias_constraint=None)
なんで、 xは不要かな？xになに入ってるか知らないけど
x = tf.keras.layers.Conv2D(depth, kernel_size,strides=(2, 1), padding='same'
これで試してみては。
投稿2022/12/09 02:16
総合スコア808
あなたの回答
tips
プレビュー
バッドをするには、ログインかつ
こちらの条件を満たす必要があります。