cannot import name ‘deserialize‘ from ‘tensorflow.keras.models‘ 的解决方案

星马梦缘2026-04-07 10:29

关于版本不匹配的处理

嗯，报错了，这种版本不匹配的问题就是很恶心，。。。

或者你报错：

回到vscode终端。

先把tensorflow的相关包清理干净

python 复制代码

pip uninstall tensorflow tensorflow-estimator tensorboard keras -y

重装tensorflow和keras，要严格按照这个版本安装：

python 复制代码

pip install tensorflow==2.10.0 keras-rl2==1.0.5

进到你的keras默认目录，维度在这里"C:\Users\HP\miniconda3\envs\brain\Lib\site-packages\rl"

进入文件夹，要修改两个文件，一个是callbacks.py

对于callbacks.py文件：

找到 第 8 行 左右的代码：python

复制代码

1from tensorflow.keras import __version__ as KERAS_VERSION

将其删除 ，并替换为下面这行代码：python
复制代码
```
1import tensorflow as tf; KERAS_VERSION = tf.__version__
```

头部如下：

python 复制代码

import warnings
import timeit
import json
from tempfile import mkdtemp

import numpy as np
import tensorflow as tf
from tensorflow.keras import __version__ as KERAS_VERSION
from tensorflow.python.keras.callbacks import Callback as KerasCallback, CallbackList as KerasCallbackList
from tensorflow.python.keras.utils.generic_utils import Progbar

再打开utils.py文件：

直接复制下面内容覆盖掉前面两个clone_函数。

python 复制代码

import numpy as np

from tensorflow.keras.models import model_from_config, Sequential, Model, model_from_config
import tensorflow.keras.optimizers as optimizers
import tensorflow.keras.backend as K
import tensorflow as tf


def clone_model(model, custom_objects={}):
    # Requires Keras 1.0.7 since get_config has breaking changes.
    config = {
        'class_name': model.__class__.__name__,
        'config': model.get_config(),
    }
    clone = model_from_config(config, custom_objects=custom_objects)
    clone.set_weights(model.get_weights())
    return clone


def clone_optimizer(optimizer):
    if type(optimizer) is str:
        print(optimizer)
        return optimizers.get(optimizer)
    # Requires Keras 1.0.7 since get_config has breaking changes.
    params = {k: v for k, v in optimizer.get_config().items()}
    config = {
        'class_name': optimizer.__class__.__name__,
        'config': params,
    }
    if hasattr(optimizers, 'optimizer_from_config'):
        # COMPATIBILITY: Keras < 2.0
        clone = optimizers.optimizer_from_config(config)
    else:
        clone = optimizers.deserialize(config)
    return clone

然后关闭jupyter重启，

噫，我好了！