NameError: name 'dataset' is not defined

Question

NameError                                 Traceback (most recent call last)
<ipython-input-35-1771b3d0e744> in <cell line: 3>()
      1 color_list = ["red", "orange", "green", "blue", "purple", "brown"]
      2 
----> 3 loader = DataLoader(dataset, batch_size=64, shuffle=True)
      4 embs = []
      5 colors = []

NameError: name 'dataset' is not defined

之后出现错误：

color_list = ["red", "orange", "green", "blue", "purple", "brown"]

loader = DataLoader(dataset, batch_size=64, shuffle=True)
embs = []
colors = []
for batch in loader:
    emb, pred = model(batch)
    embs.append(emb)
    colors += [color_list[y] for y in batch.y]
embs = torch.cat(embs, dim=0)

xs, ys = zip(*TSNE().fit_transform(embs.detach().numpy()))
plt.scatter(xs, ys, color=colors)

数据集定义为：

def train（数据集、任务、作者）： 如果任务 == '图形'： data_size = len(数据集) loader = DataLoader(dataset[:int(data_size * 0.8)], batch_size=64, shuffle=True) test_loader = DataLoader(dataset[int(data_size * 0.8):], batch_size=64, shuffle=True) 别的： test_loader = loader = DataLoader(dataset, batch_size=64, shuffle=True)