该问题似乎在Tensorflow 2.0中得以解决（至少在简单模型上如此）！这是一个似乎可以产生可重复结果的代码段。
import os####*IMPORANT*: Have to do this line *before* importing tensorflowos.environ['PYTHONHASHSEED']=str(1)import tensorflow as tfimport tensorflow.keras as kerasimport tensorflow.keras.layers import randomimport pandas as pdimport numpy as npdef reset_random_seeds(): os.environ['PYTHONHASHSEED']=str(1) tf.random.set_seed(1) np.random.seed(1) random.seed(1)#make some random datareset_random_seeds()NUM_ROWS = 1000NUM_FEATURES = 10random_data = np.random.normal(size=(NUM_ROWS, NUM_FEATURES))df = pd.Dataframe(data=random_data, columns=['x_' + str(ii) for ii in range(NUM_FEATURES)])y = df.sum(axis=1) + np.random.normal(size=(NUM_ROWS))def run(x, y): reset_random_seeds() model = keras.Sequential([ keras.layers.Dense(40, input_dim=df.shape[1], activation='relu'), keras.layers.Dense(20, activation='relu'), keras.layers.Dense(10, activation='relu'), keras.layers.Dense(1, activation='linear') ]) NUM_EPOCHS = 500 model.compile(optimizer='adam', loss='mean_squared_error') model.fit(x, y, epochs=NUM_EPOCHS, verbose=0) predictions = model.predict(x).flatten() loss = model.evaluate(x, y) #This prints out the loss by side-effect#With Tensorflow 2.0 this is now reproducible! run(df, y)run(df, y)run(df, y)
