Python 在何处使用光线角色导入tensorflow？

使用ray actors并行运行多个tensorflow模型，我问自己在哪里导入tensorflow：

# [1] maybe import tensorflow here?

@ray.remote(num_cpus=1)
class Remote_Runner:
    # [2] maybe import tensorflow here?
    def __init__(self, weights):
        # [3] maybe import tensorflow here?
        self.model=My_model()
        self.model.set_weights(wegihts)

    def do_something_with_model:
        self.model.do_something()

---

文档中给出的示例提到由于“导入tensorflow和设置全局状态的副作用”而在actor中导入tensorflow，但仅给出了光线远程功能的示例。那么我应该在[1]、[2]或[3]甚至其他地方运行“import tensorflow as tf”吗？这里是否有一个最佳实践，在[1 ]、[2 ]和[3 ]之间有什么区别，即在每种情况下如何访问TySoFLUM以及它们何时被执行？

现在可能是固定的，但是考虑的最安全的选项是在[3 ]中引入TysFooSo.

< P>。最好在My_model（）中导入tensorflow

你可以考虑下面的例子

def create_keras_model():
    from tensorflow import keras
    from tensorflow.keras import layers
    model = keras.Sequential()
    # Adds a densely-connected layer with 64 units to the model:
    model.add(layers.Dense(64, activation="relu", input_shape=(32, )))
    # Add another:
    model.add(layers.Dense(64, activation="relu"))
    # Add a softmax layer with 10 output units:
    model.add(layers.Dense(10, activation="softmax"))

    model.compile(
        optimizer=keras.optimizers.RMSprop(0.01),
        loss=keras.losses.categorical_crossentropy,
        metrics=[keras.metrics.categorical_accuracy])
    return model

import ray
import numpy as np

ray.init()

def random_one_hot_labels(shape):
    n, n_class = shape
    classes = np.random.randint(0, n_class, n)
    labels = np.zeros((n, n_class))
    labels[np.arange(n), classes] = 1
    return labels


# Use GPU wth
# @ray.remote(num_gpus=1)
@ray.remote
class Network(object):
    def __init__(self):
        self.model = create_keras_model()
        self.dataset = np.random.random((1000, 32))
        self.labels = random_one_hot_labels((1000, 10))

    def train(self):
        history = self.model.fit(self.dataset, self.labels, verbose=False)
        return history.history

    def get_weights(self):
        return self.model.get_weights()

    def set_weights(self, weights):
        # Note that for simplicity this does not handle the optimizer state.
        self.model.set_weights(weights)