TensorFlow 服务与 contrib 操作答案

【问题标题】：Tensorflow serving with contrib operationsTensorFlow 服务与 contrib 操作
【发布时间】：2019-04-16 12:04:05
【问题描述】：

如果有 tf.contrib 操作，我如何使用 tensorflow-serving 服务模型。我通过 Docker（最新）（tf 1.11 版本）使用 Tensorflow Serving，当我提供模型时，会出现下一条消息：

“Failed to start server. Error: Unknown: 1 servable(s) did not become available: {{{name: slider_universal version: 1} due to error: Not found: Op type not registered ‘ImageProjectiveTransformV2’ in binary running on 254345a5d9f1. Make sure the Op and Kernel are registered in the binary running in this process. Note that if you are loading a saved graph which used ops from tf.contrib, accessing (e.g.) tf.contrib.resampler should be done before importing the graph, as contrib ops are lazily registered when the module is first accessed.}, }”

我也是用 bazel 构建的，但是出现了同样的错误

我用tf.contrib.image.transform

如果我在导出模型期间删除此操作，它可以由 tensorflow serving 提供服务

【问题讨论】：

标签： python tensorflow tensorflow-serving

【解决方案1】：

我在使用同一个操作时遇到了问题，似乎唯一的方法就是构建它。

tensorflow_serving/model_servers/BUILD 定义了变量SUPPORTED_TENSORFLOW_OPS 中将包含哪些Tensorflow ops，我对此有点困惑，因为它指定应该包含contrib ops。但是，由于 tensorflow contrib 构建规则似乎不包括 contrib.image 下的操作，所以我通过将此变量更新为以下内容来明确添加这些操作

SUPPORTED_TENSORFLOW_OPS = [
    "@org_tensorflow//tensorflow/contrib:contrib_kernels",
    "@org_tensorflow//tensorflow/contrib:contrib_ops_op_lib",
    "@org_tensorflow//tensorflow/contrib/image:image_ops_kernels",
    "@org_tensorflow//tensorflow/contrib/image:image_ops_op_lib",
]

【讨论】：