Python 3层从头开始的神经网络（调试）

我遵循并建立了一个神经网络，但每次运行它时，列表都会出错。请帮我调试一下。我的代码在google colab链接上-

显示的错误如下所示-

---

欢迎访问该网站。你应该为问题使用更具描述性的标题，并在问题正文中显示你遇到的错误。我添加了错误。

"""ml_nolibs.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/1Skfq3A1u7Mwdo72YBRWOm4x0SCp8mIFn
"""

from random import seed,random
import numpy as np

def make_nn(n_inps,n_midd,n_outs):
    network_weights = []
    weights_input_hidden = [{'weights':[random() for i in range(n_inps+1)]} for i in range(n_midd) ]
    weights_hidden_output = [{'weights':[random() for i in range(n_midd+1)]} for i in range(n_outs) ]
    network_weights.append(weights_input_hidden)
    network_weights.append(weights_hidden_output)
    return network_weights
print(make_nn(1,2,2))

def sigmoid(n):
    return 1/(1+np.exp(-n))

def activate(weights,inputs):
    if (len(weights)-1) != len(inputs):
        return "length error ____ activat"
    ret = weights[-1]
    for i in range(len(inputs)):
        ret+= inputs[i] * weights[i]
        
            
    return sigmoid(ret)

def forward_propagate(network,inputs):
    print(inputs)
    outs = []
    for layer_of_connections in network:
        outs = []
        for neuron in layer_of_connections:
            neuron['output']=activate(neuron['weights'],inputs)
            outs.append(neuron['output'])
        inputs = outs
    return outs
print(forward_propagate([[{'weights': [0.7160471173051909, 0.5215147566159989]}, {'weights': [0.604197405116923, 0.4628263091169783]}], [{'weights': [0.4638546941280478, 0.5191037591740162, 0.8253877642443779]}, {'weights': [0.4635745950402146, 0.6078498529022434, 0.0074536694308950935]}]]
,[1]))

"""testing"""

network = make_nn(2,4,3)
print(forward_propagate(network,[6.9,9.6]))

"""assigning blame"""

def transfer_d(output):
    return output*(1-output)
def backpropagate(network,expected):
    for i in reversed(range(len(network))):
        layer = network[i]
        errors = []
        if i == (len(network) -1):
            for j in range(len(layer)):
                neuron = layer[j]
                errors.append(expected - neuron['output'])
        else:
            for j in range(len(layer)):
                err = 0
                for neuron in network[i+1]:
                    err+=neuron['weights'][j]*neuron['delta']
                errors.append(err)
        for j in range(len(layer)):
            neuron = layer[j]
            neuron['delta'] = errors[j]*transfer_d(neuron['output'])

"""# TRAINING TIME!"""

def update_weights(network,inps,l_rate = .1):
    processed_inps = inps[:-1]
    for i in range(len(network)):
        if i!=0:
            processed_inps = [neuron['output'] for neuron in network[i-1]]
        for neuron in network[i]:
            for j in range(len(processed_inps)):
                neuron['weights'][j]+=l_rate*neuron['delta']*inputs[j]
            neuron['weights'][-1]+=l_rate*neuron['delta ']

def choose_ele(l):
    return l[int(random()*(len(l)-1))]

def train_netw(network,data,n_outputs,l_rate = .1,n_epoch = 10000):#n_outputs is used for onr hot encoding using binary vector
    for epoch in range(n_epoch):
        sum_error = 0
        row = choose_ele(data)
        nn_outs = forward_propagate(network,row[:-1])
        expected = [0 for i in range(n_outputs)]
        expected[row[-1]] = 1
        sum_error = sum([(expected[i]-nn_outs[i])**2 for i in range(len(expected))])
        backpropagate(network,expected)
        update_weights(network,row[:-1])#possible mistake
        if epoch%100 == 0:
            print('epoch = %d err = %.3f'%(epoch,sum_error))

data = [[2.7810836,2.550537003,0],
    [1.465489372,2.362125076,0],
    [3.396561688,4.400293529,0],
    [1.38807019,1.850220317,0],
    [3.06407232,3.005305973,0],
    [7.627531214,2.759262235,1],
    [5.332441248,2.088626775,1],
    [6.922596716,1.77106367,1],
    [8.675418651,-0.242068655,1],
    [7.673756466,3.508563011,1]]
n_inps = data.__len__()-1
n_outs = len(set([d[-1] for d in data]))
netw = make_nn(n_inps,2,n_outs)
train_netw(netw,data,n_outs,.5,200)
print(network)