there is a problem if the program is running on the python 3.X #2
Description
follow is the trace
Traceback (most recent call last):
File "C:/Users/tians/Desktop/Language-Model-RNN-master/task.py", line 144, in
train(config)
File "C:/Users/tians/Desktop/Language-Model-RNN-master/task.py", line 81, in train
grads = tape.gradient(loss, variables_list)
File "C:\Users\tians\AppData\Roaming\Python\Python36\site-packages\tensorflow\python\eager\backprop.py", line 869, in gradient
output_gradients=output_gradients)
File "C:\Users\tians\AppData\Roaming\Python\Python36\site-packages\tensorflow\python\eager\imperative_grad.py", line 63, in imperative_grad
tape._tape, vspace, target, sources, output_gradients) # pylint: disable=protected-access
AttributeError: 'dict_values' object has no attribute '_id'