colaboratory csv 画像処理　エラー

急に以下のようなエラーが出ました。

解決策を教えて頂けるとありがたいです。

python
1import chainer
2import numpy as np
3import chainer.functions as F
4from chainer import serializers,iterators,optimizer
5from chainer.dataset import convert
6
7
8#学習に関する基本情報の定義
9EPOCH = 10 #学習回数
10BATCH = 20 #バッチサイズ
11NUM_SHAPE = 48 #画像一辺の長さ
12LEARN_RATE = 0.001
13WEIGHT_DECAY =1e-4
14GPU = 0
15SAVE_MODEL = "/content/drive/My Drive/saved_model/myresnet.npz"
16
17def learn(csvfile):
18
19    train,publictest,_= dataFromCsv(csvfile)
20  
21    train_iter = iterators.SerialIterator(train,batch_size=BATCH,shuffle=True)
22    publictest_iter = iterators.SerialIterator(publictest,batch_size=BATCH,repeat=False,shuffle=False)
23    
24    
25    model = ResNet(class_labels=9)
26 
27   chainer.cuda.get_device(GPU).use
28    model.to_gpu()
29
30    optimizer = chainer.optimizers.MomentumSGD(LEARN_RATE)
31    optimizer.setup(model)
32    optimizer.add_hook(chainer.optimizer.WeightDecay(WEIGHT_DECAY))
33
34    saved_model = model
35
36    while train_iter.epoch < EPOCH:
37        
38        batch = train_iter.next()
39
40        trainLossList = []
41        
42        x_array, y_array = convert.concat_examples(batch,GPU)
43
44        x = chainer.Variable(x_array)
45        y = chainer.Variable(y_array)
46        m = model(x)
47
48        loss_train = myCrossEntropyError(m,y)
49
50        model.cleargrads()
51
52        loss_train.backward()
53
54        optimizer.update()
55
56        trainLossList.append(chainer.cuda.to_cpu(loss_train.data))
57
58        
59        if  train_iter.is_new_epoch:
60             
61            testLossList = []
62            
63            for batch in publictest_iter:
64                x_array, y_array = convert.concat_examples(batch, GPU)
65                x = chainer.Variable(x_array)
66                y = chainer.Variable(y_array)
67                m = model(x)
68
69                loss_test = myCrossEntropyError(m, y)
70                testLossList.append(chainer.cuda.to_cpu(loss_test.data))
71
72                
73                if loss_test.data == np.min(testLossList):
74                    saved_model = model
75            
76            publictest_iter.reset()
77        
78            print("epo:" + str(train_iter.epoch) + " train_loss:" + str(np.mean(trainLossList)) + " test_loss:" + str(np.mean(testLossList)))
79            
80    
81 
82    chainer.serializers.save_npz(SAVE_MODEL, saved_model)
83
84    return
85
86
87def myCrossEntropyError(m,y):
88    DELTA = 1e-7 # マイナス無限大を発生させないように微小な値を追加する
89    return -F.sum(y*F.log(m+DELTA)+(1-y)*F.log(1-m+DELTA))
90
91if __name__ == "__main__":     
92    learn("/content/drive/My Drive/a.csv")

error
1---------------------------------------------------------------------------
2InvalidType                               Traceback (most recent call last)
3<ipython-input-4-917b77ca19ac> in <module>()
4     94 
5     95 if __name__ == "__main__":
6---> 96     learn("/content/drive/My Drive/a.csv")
7
86 frames
9<ipython-input-4-917b77ca19ac> in learn(csvfile)
10     71                 m = model(x)
11     72 
12---> 73                 loss_test = myCrossEntropyError(m, y)
13     74                 testLossList.append(chainer.cuda.to_cpu(loss_test.data))
14     75 
15
16<ipython-input-4-917b77ca19ac> in myCrossEntropyError(m, y)
17     91 def myCrossEntropyError(m,y):
18     92     DELTA = 1e-7 # マイナス無限大を発生させないように微小な値を追加する
19---> 93     return -F.sum(y*F.log(m+DELTA)+(1-y)*F.log(1-m+DELTA))
20     94 
21     95 if __name__ == "__main__":
22
23/usr/local/lib/python3.6/dist-packages/chainer/functions/math/basic_math.py in mul(self, rhs)
24    391         return MulConstant(rhs).apply((self,))[0]
25    392     rhs = _preprocess_rhs(self, rhs)
26--> 393     return Mul().apply((self, rhs))[0]
27    394 
28    395 
29
30/usr/local/lib/python3.6/dist-packages/chainer/function_node.py in apply(self, inputs)
31    295 
32    296         if configuration.config.type_check:
33--> 297             self._check_data_type_forward(in_data)
34    298 
35    299         hooks = chainer.get_function_hooks()
36
37/usr/local/lib/python3.6/dist-packages/chainer/function_node.py in _check_data_type_forward(self, in_data)
38    398         in_type = type_check.get_types(in_data, 'in_types', False)
39    399         with type_check.get_function_check_context(self):
40--> 400             self.check_type_forward(in_type)
41    401 
42    402     def check_type_forward(self, in_types):
43
44/usr/local/lib/python3.6/dist-packages/chainer/functions/math/basic_math.py in check_type_forward(self, in_types)
45    339         )
46    340         type_check.expect_broadcast_shapes(
47--> 341             in_types[0].shape, in_types[1].shape)
48    342 
49    343     def forward_chainerx(self, x):
50
51/usr/local/lib/python3.6/dist-packages/chainer/utils/type_check.py in expect_broadcast_shapes(*shape_types)
52    643         error = InvalidType('', '', msg='\n'.join(msgs))
53    644     if error is not None:
54--> 645         raise error
55
56InvalidType: cannot broadcast inputs of the following shapes:
57lhs.shape = (20, 8)
58rhs.shape = (20, 9)