翻译模型预测: TypeError：'EagerTensor‘类型的对象不可序列化

Question

我已经按照建议的by Google's tensor2tensor repository遵循了colab笔记本的翻译教程

在导出模型并将其上传到Google的AI平台引擎进行在线预测后，我在向模型发出请求时遇到了问题。

我认为翻译模型的输入是源文本的张量。但是我收到一个错误，TypeError: Object of type 'EagerTensor' is not JSON serializable

def encode(input_str, output_str=None):
  """Input str to features dict, ready for inference"""
  inputs = encoders["inputs"].encode(input_str) + [1]  # add EOS id
  batch_inputs = tf.reshape(inputs, [1, -1, 1])  # Make it 3D.
  return {"inputs": batch_inputs}

enfr_problem = problems.problem(PROBLEM)
encoders = enfr_problem.feature_encoders(DATA_DIR)

encoded_inputs = encode("Some text")
model_output = predict_json('project_name','model_name', encoded_inputs,'version_1')["outputs"]

我试着将张量转换为numpy，但仍然没有成功。有人能给我指个方向吗？

Answer 1

问题是，当您这样做时，TensorFlow会返回一个EagerTensor：

inputs = encoders["inputs"].encode(input_str) + [1]  # add EOS id
batch_inputs = tf.reshape(inputs, [1, -1, 1])

并且不能将EagerTensor转换为JSON。不幸的是，3D numpy数组也不能转换为JSON。但是numpy数组可以很容易地转换为列表。举个例子：

import json
import numpy as np
import tensorflow as tf

a = np.array([1, 2, 3])
b = np.array([1, 2, 3])
c = tf.multiply(a, b)

print(c)  # -> <tf.Tensor: shape=(3,), dtype=int64, numpy=array([1, 4, 9])>
print(c.numpy())  # -> array([1, 4, 9])
print(c.numpy().tolist())  # -> [1, 4, 9]

with open("example.json", "w") as f:
   json.dump(c, f)  # TypeError: Object of type EagerTensor is not JSON serializable
   json.dump(c.numpy(), f)  # TypeError: Object of type ndarray is not JSON serializable
   json.dump(c.numpy().tolist(), f)  # works!

我无法提供与您的情况完全相同的示例，因为您的代码片段不够完整。但

return {"inputs": batch_inputs.numpy().tolist()}

应该能胜任这项工作。

Answer 2

如果您想要将dict中的张量数据保存到JSON文件中，一个简单的解决方案是递归地进入您的字典，并使用正确的函数将您的数据转换为Json中可序列化的数据(例如，如果只是为了保存字符串，则为字符串)。我相信tensorflow一定有办法将你的数据保存为pickle文件，如果这是你真正想做的(即保存你的数据)。

以下代码用于递归地将dict中的内容转换为字符串，但您应该能够根据您的用例轻松地修改和数值化、jsonify等代码。我的用例是以人类可读的格式保存数据(不仅仅是torch.save)：

#%%

def _to_json_dict_with_strings(dictionary):
    """
    Convert dict to dict with leafs only being strings. So it recursively makes keys to strings
    if they are not dictionaries.

    Use case:
        - saving dictionary of tensors (convert the tensors to strins!)
        - saving arguments from script (e.g. argparse) for it to be pretty

    e.g.

    """
    if type(dictionary) != dict:
        return str(dictionary)
    d = {k: _to_json_dict_with_strings(v) for k, v in dictionary.items()}
    return d

def to_json(dic):
    import types
    import argparse

    if type(dic) is dict:
        dic = dict(dic)
    else:
        dic = dic.__dict__
    return _to_json_dict_with_strings(dic)

def save_to_json_pretty(dic, path, mode='w', indent=4, sort_keys=True):
    import json

    with open(path, mode) as f:
        json.dump(to_json(dic), f, indent=indent, sort_keys=sort_keys)

def my_pprint(dic):
    """

    @param dic:
    @return:

    Note: this is not the same as pprint.
    """
    import json

    # make all keys strings recursively with their naitve str function
    dic = to_json(dic)
    # pretty print
    pretty_dic = json.dumps(dic, indent=4, sort_keys=True)
    print(pretty_dic)
    # print(json.dumps(dic, indent=4, sort_keys=True))
    # return pretty_dic

import torch
# import json  # results in non serializabe errors for torch.Tensors
from pprint import pprint

dic = {'x': torch.randn(1, 3), 'rec': {'y': torch.randn(1, 3)}}

my_pprint(dic)
pprint(dic)

输出：

{
    "rec": {
        "y": "tensor([[-0.3137,  0.3138,  1.2894]])"
    },
    "x": "tensor([[-1.5909,  0.0516, -1.5445]])"
}
{'rec': {'y': tensor([[-0.3137,  0.3138,  1.2894]])},
 'x': tensor([[-1.5909,  0.0516, -1.5445]])}

相关链接：

https://discuss.pytorch.org/t/typeerror-tensor-is-not-json-serializable/36065/3或How to prettyprint a JSON file?和https://github.com/fossasia/visdom/issues/554。