Tried to construct a tensor from a int sequence, but found an item of type NoneType at index (0)

vision
1

When I run the following code that from crnn, I get this error

def encode(self, text):
“”"Support batch or single str.

    Args:
        text (str or list of str): texts to convert.

    Returns:
        torch.IntTensor [length_0 + length_1 + ... length_{n - 1}]: encoded texts.
        torch.IntTensor [n]: length of each text.
    """
    length = []
    result = []
    for item in text:            
        if self.is_chinese(item): 		   
           # item = unicode(item,'utf-8')
           item = item.encode("utf-8")
        length.append(len(item))
        for char in item:				
            index = self.dict.get(char)                
            result.append(index)
        text = result
    return (torch.IntTensor(text), torch.IntTensor(length))
2

You didn’t post the error message. :wink:

3

When I train crnn ,it had an error，how should I solve this problem?

Traceback (most recent call last):
File “crnn/crnn_main.py”, line 212, in
cost = trainBatch(crnn, criterion, optimizer)
File “crnn/crnn_main.py”, line 191, in trainBatch
t, l = converter.encode(gpu_texts)
File “/home/yang/Scence_Recognition/crnn/utils.py”, line 66, in encode
return (torch.IntTensor(text), torch.IntTensor(length))
RuntimeError: tried to construct a tensor from a int sequence, but found an item of type NoneType at index (0)

4

The error message seems to be quite clear.
Could you check, what’s inside text and length?
There seem to be a NoneType at index 0.

5

:grinning: many thank