TypeError: iteration over a 0-d tensor

[duguiming111]

TypeError: iteration over a 0-d tensor，请问你在运行得时候遇到过这个问题吗

[Machine-Tom]

I have met the same question. Have you solved it?

[ZhijianChan]

修改两个地方，在pytorch1.0下面可以运行：

1）repackage_hidden函数

def repackage_hidden(h):
	"""用新的变量重新包装隐藏层，将它们从历史中分离。"""
    if type(h) == tuple:  # lstm
    	return tuple(repackage_hidden(v) for v in h)
    else:  # rnn/gru
    	return Variable(h.data)

2）train函数内

total_loss += loss.tolist()  # loss累计

if ibatch % config.log_interval == 0 and ibatch > 0:  # 每隔多少个批次输出一次状态
	cur_loss = total_loss / config.log_interval

[purpletpig]

I have met the same question too. Have you solved it?