我想在我的python 3.5代码中使用TensorFlow和Teflearn库。我已经使用pip
安装了两者。我有以下版本:
tensorflow 0.12.0rc0
tflearn 0.2.1
python 3.5.2
conda 4.2.13
但是,运行一些代码时,我会收到以下错误:
TypeError: Using a `tf.Tensor` as a Python `bool` is not allowed.
从我可以从GitHub上出发的问题来看,应该已经解决此问题。
我该如何修复?
编辑:这是完整的堆栈跟踪:
Traceback (most recent call last):
File "test.py", line 35, in <module>
agent = DQNAgent(create_mlp, n_actions, env.observation_space.shape, min_replay_size=10000, batch_size=64)
File "agents.py", line 86, in __init__
self.q_net = network_fn(self.s, self._n_actions)
File "test.py", line 23, in create_mlp
net = nn.input_data(placeholder=inputs)
File "//anaconda/lib/python3.5/site-packages/tflearn/layers/core.py", line 55, in input_data
if not shape and not placeholder:
File "//anaconda/lib/python3.5/site-packages/tensorflow/python/framework/ops.py", line 533, in __bool__
raise TypeError("Using a `tf.Tensor` as a Python `bool` is not allowed. "
此错误是由tflearn v0.2.1
库引起的。解决此错误的方法是通过git
安装最新版本:
pip install git+https://github.com/tflearn/tflearn.git
这将安装tflearn v0.2.2
,此错误不再发生。