GCP自定义预测例程无法包括setup.py中指定的jsonschema依赖项

Question

根据GCP AI平台的文档here，自定义预测例程部署应允许包含PyPI依赖项。我在我的setup.py脚本中包含了我对jsonschema的依赖，如下所示：

from setuptools import setup
from setuptools import find_packages


REQUIRED_PACKAGES = ['jsonschema']

setup(
    name='custom_code',
    version='1.0.2',
    scripts=['predictor.py', 'preprocess.py'],
    install_requires=REQUIRED_PACKAGES,
    packages=find_packages(),
    include_package_data=True
)

但在部署时收到以下错误消息：

ERROR: (gcloud.beta.ai-platform.versions.create) Create Version failed. Bad model detected with error:  "Failed to load model: Unexpected error when loading the model: 'str' object has no attribute 'decode' (Error code: 0)"

当指定类似soREQUIRED_PACKAGES = ['jsonschema==3.2.0']的版本时，相同的错误仍然存在。然后我使用了较低版本：

from setuptools import setup
from setuptools import find_packages


REQUIRED_PACKAGES = ['jsonschema==3.0.0']

setup(
    name='custom_code',
    version='1.0.2',
    scripts=['predictor.py', 'preprocess.py'],
    install_requires=REQUIRED_PACKAGES,
    packages=find_packages(),
    include_package_data=True
)

但现在出现了这个错误：

ERROR: (gcloud.beta.ai-platform.versions.create) Create Version failed. Bad model detected with error:  "Failed to load model: Unexpected error when loading the model: problem in predictor - DistributionNotFound: The 'jsonschema' distribution was not found and is required by the application (Error code: 0)"

这里会出什么问题

Answer 1

原来最初的错误Bad model detected with error: "Failed to load model: Unexpected error when loading the model: 'str' object has no attribute 'decode' (Error code: 0)"实际上是由模型格式问题引起的。这似乎是TensorFlow Keras的a known issue（尽管我的TF版本是1.15，引用的TF版本是2.1.0）。一旦我使用了TensorFlow SavedModel format，错误立即消失了，我还能够将jsonchema依赖项按原样包含在setup.py文件中