python - Disable train accuracy testing after each epoch

Question

When I use tf.keras.Model.fit(), I get train accuracy testing after every epoch:

Epoch 1/50
420/420 - 3s - loss: 0.1289 - accuracy: 0.9633 - lr: 0.0010

I use CUDA, so the learning process itself is fast, but accuracy testing after each epoch greatly slow down the training process. Even more when I do the hyperparameters tuning.

So, how can I disable it?

标签： pythontensorflowkerastf.keras