在使用pytorch的指數(shù)衰減學(xué)習(xí)率時(shí)，出現(xiàn)報(bào)錯(cuò)UserWarning: Detected call of `lr_scheduler.step()` before `optimizer.step()`. In PyTorch 1.1.0 and later, you should call them in the opposite order: `optimizer.step()` before `lr_scheduler.step()`. ?Failure to do this will result in PyTorch skipping the first value of the learning rate schedule.

原因是如報(bào)錯(cuò)所說(shuō)，在“optimizer.step()”之前對(duì)“l(fā)r_scheduler.step()”的進(jìn)行了調(diào)用，如下錯(cuò)誤的代碼所示：

for i in range(epoch):net.scheduler.step()net.set_mode_train(True)for j, (x, y) in enumerate(train_loader):cost_pred, err = net.fit(x, y)net.epoch = i

所以應(yīng)該把lr_scheduler.step()放在每次epoch訓(xùn)練完成之后：

for i in range(epoch):net.set_mode_train(True)for j, (x, y) in enumerate(train_loader):cost_pred, err = net.fit(x, y)net.scheduler.step()net.epoch = i

?

總結(jié)

以上是生活随笔為你收集整理的Detected call of `lr_scheduler.step()` before `optimizer.step()`.的全部?jī)?nèi)容，希望文章能夠幫你解決所遇到的問(wèn)題。

如果覺得生活随笔網(wǎng)站內(nèi)容還不錯(cuò)，歡迎將生活随笔推薦給好友。