摘要:
结论 torch.stack ,类似并联,结果是在相应的维度会增加一维; 所以要求每个tensor大小相等 torch.cat 类似串联,结果是在相应的维度值会增加,所以要求扩展的维度大小相等 1.torch.stack l = [] for i in range(0,3): x = torch.r 阅读全文
摘要:
memory_profiler是干嘛的 This is a python module for monitoring memory consumption of a process as well as line-by-line analysis of memory consumption for 阅读全文