If you want to provide labels as integers, please use SparseCategoricalCrossentropy loss. We expect labels to be provided in a onehot representation. дёЉиЁгЃ® xxx г‚’дЅїгЃЈгЃ¦ optimizer='xxxx' гЃ®г‚€гЃ†гЃ«ж–‡е—е€—гЃ§жЊ‡е®љгЃ™г‚‹гЃ“гЃЁг‚‚гЂЃ optimizer=tf.optimizers. Use this crossentropy loss function when there are two or more label classes. For such a model with output shape of (None, 10), the conventional way is to have the target outputs converted to the one-hot encoded array to match with the output shape, however, with the help of the sparsecategoricalcrossentropy loss function, we can skip that step and keep the integers as targets. If a scalar is provided, then the loss is simply scaled by the.

sampleweight: Optional sampleweight acts as reduction weighting coefficient for the per-sample losses. dN-1) ypred: The predicted values, of shape (batchsize, d0. compile ( optimizer = 'adam', loss = 'sparse_categorical_crossentropy', metrics = )

Sparse categorical cross entropy