pytorch 报错:RuntimeError: CUDA error: device-side assert triggered
原因:
标签越界:
/opt/conda/conda-bld/pytorch_1544176307774/work/aten/src/THCUNN/ClassNLLCriterion.cu:105: void cunn_ClassNLLCriterion_updateOutput_kernel(Dtype *, Dtype *, Dtype *, long *, Dtype *, int, int, int, int, long) [with Dtype = float, Acctype = float]: block: [0,0,0], thread: [5,0,0] Assertion `t >= 0 && t < n_classes` failed.在这里插入代码片
原始标签是0—7, 提出未标注数据时将255标签未剔除,保留下来了。