名称错误：名称“inp_en_idx”未定义

Question

我尝试运行以下代码（翻译）：

import torch
import torch.nn as nn
lstm = nn.LSTM(input_size=300, hidden_size=256, num_layers=2) 
attn = nn.Linear(256, 10) 
fc = nn.Linear(10 * 2, 300)

input_en = torch.tensor(inp_en_idx) 
hidden = lstm(input_en)[0]
attn_weights = attn(hidden[-1]) 
context = attn_weights.softmax(dim=1).unsqueeze(1).expand_as(hidden)
context = context * hidden 
context = context.sum(dim=1)
out = fc(context)

但出现此错误：

NameError                                 Traceback (most recent call last)
<ipython-input-2-d484f124dd49> in <cell line: 7>()
      5 fc = nn.Linear(10 * 2, 300)
      6 
----> 7 input_en = torch.tensor(inp_en_idx)
      8 hidden = lstm(input_en)[0]
      9 attn_weights = attn(hidden[-1])

NameError: name 'inp_en_idx' is not defined

我该如何解决这个错误？

Answer 1

什么是

inp_en_idx

你之前定义过 is 吗？使用前必须声明