解决RuntimeError: one of the variables needed for gradient computation has been modified by an inplace

跑pytorch模型报错：

RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda.FloatTensor [720, 64, 36, 36]], which is output 0 of TanhBackward0, is at version 1; expected version 0 instead. Hint: enable anomaly detection to find the operation that failed to compute its gradient, with torch.autograd.set_detect_anomaly(True).

根据[720, 64, 36, 36]，定位模型报错部分：

上网搜了一下，由于使用了分布式训练，在模型中使用Y +=X的操作时容易在进行此操作前，数据被修改。因此只要将Y=Y+X改为Y=Y.clone()+X即可