問題

運行tensorflow 的resnet網(wǎng)絡(luò)，預(yù)測結(jié)果，出現(xiàn)報錯

ValueError: Shape mismatch: The shape of labels (received (768,)) should equal the shape of logits e

原因

解決：use of sparse_softmax_cross_entropy. feed it one hot encoded labels which it doesn’t want.

Switch to just normal softmax_cross_entropy and see if that works.
這個問題是由于輸出層的類別數(shù)和訓(xùn)練數(shù)據(jù)shape不同導(dǎo)致。
在訓(xùn)練模型時，一般會將label使用one hot編碼，然后模型的loss使用交叉熵：

解決方法

Use this crossentropy loss function when there are two or more label classes. We expect labels to be provided in a one_hot representation. If you want to provide labels as integers, please use SparseCategoricalCrossentropy loss. There should be # classes floating point values per feature.

修改sparse_categorical_crossentropy為ategorical_crossentropy

?

?

《新程序員》：云原生和全面數(shù)字化實踐50位技術(shù)專家共同創(chuàng)作，文字、視頻、音頻交互閱讀

總結(jié)

以上是生活随笔為你收集整理的ValueError: Shape mismatch: The shape of labels (received (768,)) should equal the shape of logits e的全部內(nèi)容，希望文章能夠幫你解決所遇到的問題。

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