[英]what use tensorflow estimator create multi-input
對不起,我的英語不好=。=
我創建了一個keras模型,並使用tf.keras.estimator.model_to_estimator
轉換為estimator,但是該模型是多輸入的,我該如何創建數據集來提供數據?
這是我的型號代碼:
model = VGG19(include_top=False, input_shape=(182, 182 , 3))
y = model.output
y = keras.layers.Flatten()(y)
y = keras.layers.Dense(512, activation='relu')(y)
y = keras.layers.Dense(256, activation='relu')(y)
y = keras.layers.Dense(128, activation='relu')(y)
model = keras.Model(inputs=model.input, outputs=y)
input_image = keras.layers.Input(shape=(182, 182, 3))
input_anchor = keras.layers.Input(shape=(182, 182, 3))
out_image = model(input_image)
out_anchor = model(input_anchor)
out = keras.layers.concatenate([out_image, out_anchor])
out = keras.layers.Dense(1, activation='sigmoid')(out)
img_model = keras.Model([input_image, input_anchor], out)
face_model.compile(optimizer=tf.train.AdamOptimizer(1e-4, loss='binary_crossentropy', metrics=['accuracy'])
distribution = tf.contrib.distribute.CollectiveAllReduceStrategy(num_gpus_per_worker=0)
config = tf.estimator.RunConfig(model_dir='/home/zjq/test/image_model.h5', train_distribute=distribution)
est_model = tf.keras.estimator.model_to_estimator(keras_model=image_model, config=config)
現在,我有一個輸入列表,形狀為[(100000,182,182,3),(100000,182,182,3),(100000,1)],如何定義輸入函數返回tf.data。數據集?
首先,輸入名稱占位符:
input_image = keras.layers.Input(shape=(182, 182, 3),name='image')
input_anchor = keras.layers.Input(shape=(182, 182, 3),name='anchor')
如果您的輸入數據是train_data
並且形狀為[(100000, 182, 182, 3), (100000, 182, 182, 3), (100000, 1)]
train_data
[(100000, 182, 182, 3), (100000, 182, 182, 3), (100000, 1)]
train_data
[(100000, 182, 182, 3), (100000, 182, 182, 3), (100000, 1)]
train_data
[(100000, 182, 182, 3), (100000, 182, 182, 3), (100000, 1)]
,則執行以下操作:
BATCH_SIZE = 512
EPOCHS = 4
def input_fn(data, epochs, batch_size):
# Convert the inputs to a Dataset.
dataset = tf.data.Dataset.from_tensor_slices(({'image':data[0],'anchor':data[1]}, data[2]))
# Shuffle, repeat, and batch the examples.
SHUFFLE_SIZE = 1000
dataset = dataset.shuffle(SHUFFLE_SIZE).repeat(epochs).batch(batch_size)
dataset = dataset.prefetch(2)
# Return the dataset.
return dataset
est_model.train(lambda :input_fn(train_data,EPOCHS,BATCH_SIZE))
可以根據您的需要調整參數BATCH_SIZE
, EPOCHS
和SHUFFLE_SIZE
。
聲明:本站的技術帖子網頁,遵循CC BY-SA 4.0協議,如果您需要轉載,請注明本站網址或者原文地址。任何問題請咨詢:yoyou2525@163.com.