LSTM打印渐变色

Question

我正在试着打印渐变，到目前为止我有这个。你知道吗

my_model=load_model('word2vec_lstm.h5')

weights = my_model.weights # weight tensors

gradients = my_model.optimizer.get_gradients(my_model.total_loss, weights) # gradient tensors

input_tensors = [my_model.inputs[0], # input data
                 my_model.sample_weights[0], # sample weights
                 my_model.targets[0], # labels
                 K.learning_phase(), # train or test mode
]

get_gradients = K.function(inputs=input_tensors, outputs=gradients)

inputs = [[trainDataVecs], # X input data
          [1], # sample weights
          [Y_train], # y labels
          0.005 # learning phase in TEST mode
]

print [a for a in zip(weights, get_gradients(inputs))]

但是，这给了我以下错误InvalidArgumentError: transpose expects a vector of size 4. But input(1) is a vector of size 3