Here is training log for mnist-cnn-beautiful-model.gguf:
```console
(tf) examples/mnist(norm_agree) $ python mnist-tf.py train mnist-cnn-beautiful-model 25
x_train shape: (60000, 28, 28, 1)
60000 train samples
10000 test samples
Model: "sequential"
_________________________________________________________________
Layer (type) Output Shape Param #
conv2d (Conv2D) (None, 24, 24, 32) 832
conv2d_1 (Conv2D) (None, 20, 20, 32) 25632
batch_normalization (Batch (None, 20, 20, 32) 128
Normalization)
max_pooling2d (MaxPooling2 (None, 10, 10, 32) 0
D)
conv2d_2 (Conv2D) (None, 8, 8, 64) 18496
conv2d_3 (Conv2D) (None, 6, 6, 64) 36928
batch_normalization_1 (Bat (None, 6, 6, 64) 256
chNormalization)
max_pooling2d_1 (MaxPoolin (None, 3, 3, 64) 0
g2D)
flatten (Flatten) (None, 576) 0
dropout (Dropout) (None, 576) 0
dense (Dense) (None, 10) 5770
Total params: 88042 (343.91 KB) Trainable params: 87850 (343.16 KB) Non-trainable params: 192 (768.00 Byte)
Epoch 1/25 422/422 [==============================] - 19s 45ms/step - loss: 0.2546 - accuracy: 0.9237 - val_loss: 0.0823 - val_accuracy: 0.9798 Epoch 2/25 422/422 [==============================] - 19s 46ms/step - loss: 0.0754 - accuracy: 0.9769 - val_loss: 0.0400 - val_accuracy: 0.9883 Epoch 3/25 422/422 [==============================] - 19s 46ms/step - loss: 0.0530 - accuracy: 0.9837 - val_loss: 0.0297 - val_accuracy: 0.9912 Epoch 4/25 422/422 [==============================] - 20s 48ms/step - loss: 0.0422 - accuracy: 0.9866 - val_loss: 0.0292 - val_accuracy: 0.9915 Epoch 5/25 422/422 [==============================] - 21s 49ms/step - loss: 0.0372 - accuracy: 0.9882 - val_loss: 0.0317 - val_accuracy: 0.9913 Epoch 6/25 422/422 [==============================] - 21s 50ms/step - loss: 0.0304 - accuracy: 0.9900 - val_loss: 0.0365 - val_accuracy: 0.9893 Epoch 7/25 422/422 [==============================] - 21s 50ms/step - loss: 0.0284 - accuracy: 0.9911 - val_loss: 0.0483 - val_accuracy: 0.9872 Epoch 8/25 422/422 [==============================] - 21s 51ms/step - loss: 0.0273 - accuracy: 0.9912 - val_loss: 0.0327 - val_accuracy: 0.9917 Epoch 9/25 422/422 [==============================] - 22s 51ms/step - loss: 0.0237 - accuracy: 0.9926 - val_loss: 0.0276 - val_accuracy: 0.9925 Epoch 10/25 422/422 [==============================] - 22s 51ms/step - loss: 0.0220 - accuracy: 0.9928 - val_loss: 0.0334 - val_accuracy: 0.9933 Epoch 11/25 422/422 [==============================] - 22s 51ms/step - loss: 0.0228 - accuracy: 0.9927 - val_loss: 0.0368 - val_accuracy: 0.9907 Epoch 12/25 422/422 [==============================] - 22s 52ms/step - loss: 0.0208 - accuracy: 0.9931 - val_loss: 0.0263 - val_accuracy: 0.9927 Epoch 13/25 422/422 [==============================] - 22s 52ms/step - loss: 0.0181 - accuracy: 0.9937 - val_loss: 0.0348 - val_accuracy: 0.9912 Epoch 14/25 422/422 [==============================] - 22s 52ms/step - loss: 0.0190 - accuracy: 0.9937 - val_loss: 0.0269 - val_accuracy: 0.9943 Epoch 15/25 422/422 [==============================] - 22s 53ms/step - loss: 0.0150 - accuracy: 0.9952 - val_loss: 0.0363 - val_accuracy: 0.9918 Epoch 16/25 422/422 [==============================] - 22s 53ms/step - loss: 0.0141 - accuracy: 0.9950 - val_loss: 0.0302 - val_accuracy: 0.9935 Epoch 17/25 422/422 [==============================] - 22s 53ms/step - loss: 0.0150 - accuracy: 0.9951 - val_loss: 0.0358 - val_accuracy: 0.9923 Epoch 18/25 422/422 [==============================] - 22s 52ms/step - loss: 0.0145 - accuracy: 0.9951 - val_loss: 0.0477 - val_accuracy: 0.9905 Epoch 19/25 422/422 [==============================] - 22s 52ms/step - loss: 0.0138 - accuracy: 0.9954 - val_loss: 0.0327 - val_accuracy: 0.9928 Epoch 20/25 422/422 [==============================] - 22s 52ms/step - loss: 0.0126 - accuracy: 0.9958 - val_loss: 0.0442 - val_accuracy: 0.9915 Epoch 21/25 422/422 [==============================] - 22s 52ms/step - loss: 0.0122 - accuracy: 0.9958 - val_loss: 0.0317 - val_accuracy: 0.9922 Epoch 22/25 422/422 [==============================] - 22s 52ms/step - loss: 0.0127 - accuracy: 0.9959 - val_loss: 0.0317 - val_accuracy: 0.9938 Epoch 23/25 422/422 [==============================] - 22s 51ms/step - loss: 0.0101 - accuracy: 0.9966 - val_loss: 0.0356 - val_accuracy: 0.9918 Epoch 24/25 422/422 [==============================] - 22s 52ms/step - loss: 0.0115 - accuracy: 0.9961 - val_loss: 0.0276 - val_accuracy: 0.9942 Epoch 25/25 422/422 [==============================] - 22s 52ms/step - loss: 0.0092 - accuracy: 0.9968 - val_loss: 0.0347 - val_accuracy: 0.9935 Test loss: 0.030567098408937454 Test accuracy: 0.9915000200271606 Keras model saved to 'mnist-cnn-beautiful-model' ```
- Downloads last month
- 13