PyTorch报错：RuntimeError: CUDA error: device-side assert triggered at /pytorch/aten/src/THC/generic_runtimeerror: cuda runtime error (710) : device-si

训练模型报错

RuntimeError: cuda runtime error (710) : device-side assert triggered at /pytorch/aten/src/THC/generic/THCTensorMath.cu:29
1

问题解决

原因：标签（label）越界，一般从0开始

方法：更改你的标签值，从0开始

target_names = ['__label__1', '__label__2', '__label__3', '__label__4', '__label__5']
    for i in range(len(questlist)):
        if i < len(questlist):
            for j in range(len(attrlist)):
                if j < len(attrlist):
                    if attr_names[j] in attrlist[i]:
                        label = target_names[j]
                        line = " ".join(jieba.cut(questlist[i])) + "\t" + str(int(label[-1])-1) ## 标签值减1
                        linelist.append(line)
                        break
1
2
3
4
5
6
7
8
9
10

  line = " ".join(jieba.cut(questlist[i])) + "\t" + str(int(label[-1])-1) ## 标签值减1
1