 }TensorFlow.org에서 보기 |
 Google Colab에서 실행하기 |
 View source on GitHub |
개요
이 튜토리얼에서는 Neural Structured Learning(NSL) 프레임워크를 사용하여 이미지 분류에 적대적 학습(Goodfellow et al., 2014)을 사용하는 방법을 살펴봅니다.
적대적 학습의 핵심 아이디어는 유기적 훈련 데이터와 함께 적대적으로 교란된 데이터(적대적 예라고 함)로 모델을 훈련하는 것입니다. 적대적인 예는 의도적으로 모델을 잘못된 예측 또는 분류로 오도하기 위해 구성됩니다. 이러한 예를 사용한 훈련을 통해 모델은 예측할 때 적대적 교란에 대해 견고함을 유지하도록 학습합니다.
이 튜토리얼에서는 Neural Structured Learning 프레임워크를 사용하여 견고한 모델을 얻기 위해 적대적 학습을 적용하는 다음 절차를 설명합니다.
- 신경망을 기본 모델로 생성합니다. 이 튜토리얼에서 기본 모델은
tf.keras함수형 API로 생성됩니다. 이 절차는tf.keras순차 및 하위 클래스화 API로 생성된 모델과도 호환됩니다. - NSL 프레임워크에서 제공하는
AdversarialRegularization래퍼 클래스로 기본 모델을 래핑하여 새tf.keras.Model인스턴스를 만듭니다. 이 새로운 모델은 훈련 목표에서 정규화 항으로 적대적 손실을 포함합니다. - 훈련 데이터의 예를 특성 사전으로 변환합니다.
- 새 모델을 훈련하고 평가합니다.
설정
Neural Structured Learning 패키지를 설치합니다.
pip install --quiet neural-structured-learning라이브러리를 가져옵니다. neural_structured_learning을 nsl로 축약합니다.
import matplotlib.pyplot as plt
import neural_structured_learning as nsl
import numpy as np
import tensorflow as tf
import tensorflow_datasets as tfds
하이퍼 매개변수
모델 훈련 및 평가를 위해 하이퍼 매개변수(HParams 객체에서)를 수집하고 설명합니다.
입출력:
input_shape: 입력 텐서의 형상, 각 이미지는 1채널의 28x28 픽셀입니다.num_classes: 총 10개의 클래스가 있으며 10자리 [0-9]에 해당합니다.
모델 아키텍처:
conv_filters: 각각 컨볼루셔널 레이어에서 필터의 수를 지정하는 숫자의 목록입니다.kernel_size: 모든 컨볼루셔널 레이어가 공유하는 2D 컨볼루션 창의 크기입니다.pool_size: 각 최대 풀링 레이어에서 이미지를 축소하는 요소입니다.num_fc_units: 완전히 연결된 각 레이어의 단위(예: 너비) 수입니다.
훈련 및 평가:
batch_size: 훈련 및 평가에 사용되는 배치 크기입니다.epochs: 훈련 epoch의 수입니다.
적대적 학습:
adv_multiplier: 훈련 목표에서 레이블이 지정된 손실에 대한 적대적 손실의 가중치입니다.adv_step_size: 적대적 교란의 크기입니다.adv_grad_norm: 적대적 교란의 크기를 측정하기 위한 표준입니다.
class HParams(object):
def __init__(self):
self.input_shape = [28, 28, 1]
self.num_classes = 10
self.conv_filters = [32, 64, 64]
self.kernel_size = (3, 3)
self.pool_size = (2, 2)
self.num_fc_units = [64]
self.batch_size = 32
self.epochs = 5
self.adv_multiplier = 0.2
self.adv_step_size = 0.2
self.adv_grad_norm = 'infinity'
HPARAMS = HParams()
MNIST 데이터세트
MNIST 데이터세트에는 손으로 쓴 숫자('0'에서 '9'까지)의 회색조 이미지가 포함됩니다. 각 이미지는 저해상도(28x28 픽셀)에서 한 자리를 표시합니다. 관련된 작업은 이미지를 숫자당 하나씩 10개의 범주로 분류하는 것입니다.
여기서는 TensorFlow Datasets에서 MNIST 데이터세트를 로드합니다. 데이터 다운로드 및 tf.data.Dataset 생성을 처리합니다. 로드된 데이터세트에는 두 개의 하위 집합이 있습니다.
- 60,000개의 예제로
train하고 - 10,000개의 예제로
test합니다.
두 하위 집합의 예는 다음 두 키를 사용하여 특성 사전에 저장됩니다.
image: 0에서 255 사이의 픽셀 값 배열입니다.label: Groundtruth 레이블, 범위는 0~9입니다.
datasets = tfds.load('mnist')
train_dataset = datasets['train']
test_dataset = datasets['test']
IMAGE_INPUT_NAME = 'image'
LABEL_INPUT_NAME = 'label'
모델을 수치적으로 안정적으로 만들기 위해 normalize 함수에 데이터세트를 매핑하여 픽셀 값을 [0, 1]로 정규화합니다. 훈련 세트를 셔플하고 일괄 처리한 후 기본 모델의 훈련을 위해 예를 특성 튜플 (image, label)로 변환합니다. 또한, 나중에 사용할 수 있도록 튜플에서 사전으로 변환하는 함수를 제공합니다.
def normalize(features):
features[IMAGE_INPUT_NAME] = tf.cast(
features[IMAGE_INPUT_NAME], dtype=tf.float32) / 255.0
return features
def convert_to_tuples(features):
return features[IMAGE_INPUT_NAME], features[LABEL_INPUT_NAME]
def convert_to_dictionaries(image, label):
return {IMAGE_INPUT_NAME: image, LABEL_INPUT_NAME: label}
train_dataset = train_dataset.map(normalize).shuffle(10000).batch(HPARAMS.batch_size).map(convert_to_tuples)
test_dataset = test_dataset.map(normalize).batch(HPARAMS.batch_size).map(convert_to_tuples)
기본 모델
기본 모델은 3개의 컨볼루셔널 레이어와 2개의 완전 연결 레이어(HPARAMS에 정의된 대로)로 구성된 신경망입니다. 여기에서는 Keras 함수형 API를 사용하여 정의합니다. 다른 API 또는 모델 아키텍처를 자유롭게 사용해 보세요.
def build_base_model(hparams):
"""Builds a model according to the architecture defined in `hparams`."""
inputs = tf.keras.Input(
shape=hparams.input_shape, dtype=tf.float32, name=IMAGE_INPUT_NAME)
x = inputs
for i, num_filters in enumerate(hparams.conv_filters):
x = tf.keras.layers.Conv2D(
num_filters, hparams.kernel_size, activation='relu')(
x)
if i < len(hparams.conv_filters) - 1:
# max pooling between convolutional layers
x = tf.keras.layers.MaxPooling2D(hparams.pool_size)(x)
x = tf.keras.layers.Flatten()(x)
for num_units in hparams.num_fc_units:
x = tf.keras.layers.Dense(num_units, activation='relu')(x)
pred = tf.keras.layers.Dense(hparams.num_classes, activation='softmax')(x)
model = tf.keras.Model(inputs=inputs, outputs=pred)
return model
base_model = build_base_model(HPARAMS)
base_model.summary()
Model: "model" _________________________________________________________________ Layer (type) Output Shape Param # ================================================================= image (InputLayer) [(None, 28, 28, 1)] 0 _________________________________________________________________ conv2d (Conv2D) (None, 26, 26, 32) 320 _________________________________________________________________ max_pooling2d (MaxPooling2D) (None, 13, 13, 32) 0 _________________________________________________________________ conv2d_1 (Conv2D) (None, 11, 11, 64) 18496 _________________________________________________________________ max_pooling2d_1 (MaxPooling2 (None, 5, 5, 64) 0 _________________________________________________________________ conv2d_2 (Conv2D) (None, 3, 3, 64) 36928 _________________________________________________________________ flatten (Flatten) (None, 576) 0 _________________________________________________________________ dense (Dense) (None, 64) 36928 _________________________________________________________________ dense_1 (Dense) (None, 10) 650 ================================================================= Total params: 93,322 Trainable params: 93,322 Non-trainable params: 0 _________________________________________________________________
다음으로 기본 모델을 훈련하고 평가합니다.
base_model.compile(optimizer='adam', loss='sparse_categorical_crossentropy',
metrics=['acc'])
base_model.fit(train_dataset, epochs=HPARAMS.epochs)
Epoch 1/5 1/1875 [..............................] - ETA: 1:28:18 - loss: 2.3145 - acc: 0.0625 9/1875 [..............................] - ETA: 12s - loss: 2.2732 - acc: 0.2100 18/1875 [..............................] - ETA: 11s - loss: 2.2135 - acc: 0.2672 27/1875 [..............................] - ETA: 11s - loss: 2.1153 - acc: 0.3154 35/1875 [..............................] - ETA: 11s - loss: 2.0145 - acc: 0.3546 44/1875 [..............................] - ETA: 11s - loss: 1.9118 - acc: 0.3910 53/1875 [..............................] - ETA: 11s - loss: 1.8177 - acc: 0.4228 62/1875 [..............................] - ETA: 11s - loss: 1.7342 - acc: 0.4506 71/1875 [>.............................] - ETA: 10s - loss: 1.6605 - acc: 0.4747 80/1875 [>.............................] - ETA: 10s - loss: 1.5936 - acc: 0.4963 89/1875 [>.............................] - ETA: 10s - loss: 1.5331 - acc: 0.5158 98/1875 [>.............................] - ETA: 10s - loss: 1.4780 - acc: 0.5335 106/1875 [>.............................] - ETA: 10s - loss: 1.4333 - acc: 0.5477 115/1875 [>.............................] - ETA: 10s - loss: 1.3870 - acc: 0.5624 124/1875 [>.............................] - ETA: 10s - loss: 1.3449 - acc: 0.5757 133/1875 [=>............................] - ETA: 10s - loss: 1.3061 - acc: 0.5879 142/1875 [=>............................] - ETA: 10s - loss: 1.2703 - acc: 0.5992 151/1875 [=>............................] - ETA: 10s - loss: 1.2371 - acc: 0.6095 160/1875 [=>............................] - ETA: 10s - loss: 1.2063 - acc: 0.6192 169/1875 [=>............................] - ETA: 10s - loss: 1.1774 - acc: 0.6283 178/1875 [=>............................] - ETA: 10s - loss: 1.1503 - acc: 0.6368 187/1875 [=>............................] - ETA: 10s - loss: 1.1246 - acc: 0.6449 196/1875 [==>...........................] - ETA: 9s - loss: 1.1006 - acc: 0.6525 205/1875 [==>...........................] - ETA: 9s - loss: 1.0780 - acc: 0.6597 214/1875 [==>...........................] - ETA: 9s - loss: 1.0567 - acc: 0.6664 222/1875 [==>...........................] - ETA: 9s - loss: 1.0388 - acc: 0.6721 231/1875 [==>...........................] - ETA: 9s - loss: 1.0195 - acc: 0.6783 240/1875 [==>...........................] - ETA: 9s - loss: 1.0011 - acc: 0.6841 249/1875 [==>...........................] - ETA: 9s - loss: 0.9836 - acc: 0.6896 258/1875 [===>..........................] - ETA: 9s - loss: 0.9668 - acc: 0.6950 267/1875 [===>..........................] - ETA: 9s - loss: 0.9508 - acc: 0.7000 276/1875 [===>..........................] - ETA: 9s - loss: 0.9356 - acc: 0.7049 285/1875 [===>..........................] - ETA: 9s - loss: 0.9209 - acc: 0.7095 294/1875 [===>..........................] - ETA: 9s - loss: 0.9068 - acc: 0.7140 303/1875 [===>..........................] - ETA: 9s - loss: 0.8934 - acc: 0.7182 312/1875 [===>..........................] - ETA: 9s - loss: 0.8806 - acc: 0.7223 321/1875 [====>.........................] - ETA: 9s - loss: 0.8682 - acc: 0.7262 330/1875 [====>.........................] - ETA: 9s - loss: 0.8563 - acc: 0.7300 339/1875 [====>.........................] - ETA: 9s - loss: 0.8449 - acc: 0.7336 348/1875 [====>.........................] - ETA: 9s - loss: 0.8338 - acc: 0.7370 357/1875 [====>.........................] - ETA: 8s - loss: 0.8232 - acc: 0.7404 366/1875 [====>.........................] - ETA: 8s - loss: 0.8129 - acc: 0.7436 375/1875 [=====>........................] - ETA: 8s - loss: 0.8029 - acc: 0.7468 384/1875 [=====>........................] - ETA: 8s - loss: 0.7933 - acc: 0.7498 393/1875 [=====>........................] - ETA: 8s - loss: 0.7840 - acc: 0.7527 402/1875 [=====>........................] - ETA: 8s - loss: 0.7749 - acc: 0.7556 411/1875 [=====>........................] - ETA: 8s - loss: 0.7661 - acc: 0.7584 420/1875 [=====>........................] - ETA: 8s - loss: 0.7576 - acc: 0.7611 429/1875 [=====>........................] - ETA: 8s - loss: 0.7493 - acc: 0.7637 438/1875 [======>.......................] - ETA: 8s - loss: 0.7412 - acc: 0.7662 447/1875 [======>.......................] - ETA: 8s - loss: 0.7334 - acc: 0.7687 457/1875 [======>.......................] - ETA: 8s - loss: 0.7250 - acc: 0.7713 466/1875 [======>.......................] - ETA: 8s - loss: 0.7177 - acc: 0.7736 475/1875 [======>.......................] - ETA: 8s - loss: 0.7106 - acc: 0.7759 484/1875 [======>.......................] - ETA: 8s - loss: 0.7037 - acc: 0.7780 494/1875 [======>.......................] - ETA: 8s - loss: 0.6962 - acc: 0.7804 503/1875 [=======>......................] - ETA: 8s - loss: 0.6897 - acc: 0.7825 512/1875 [=======>......................] - ETA: 8s - loss: 0.6833 - acc: 0.7845 520/1875 [=======>......................] - ETA: 7s - loss: 0.6778 - acc: 0.7862 529/1875 [=======>......................] - ETA: 7s - loss: 0.6717 - acc: 0.7881 537/1875 [=======>......................] - ETA: 7s - loss: 0.6665 - acc: 0.7898 546/1875 [=======>......................] - ETA: 7s - loss: 0.6607 - acc: 0.7916 555/1875 [=======>......................] - ETA: 7s - loss: 0.6550 - acc: 0.7934 564/1875 [========>.....................] - ETA: 7s - loss: 0.6495 - acc: 0.7951 573/1875 [========>.....................] - ETA: 7s - loss: 0.6441 - acc: 0.7968 582/1875 [========>.....................] - ETA: 7s - loss: 0.6389 - acc: 0.7985 591/1875 [========>.....................] - ETA: 7s - loss: 0.6337 - acc: 0.8001 600/1875 [========>.....................] - ETA: 7s - loss: 0.6287 - acc: 0.8017 609/1875 [========>.....................] - ETA: 7s - loss: 0.6238 - acc: 0.8032 619/1875 [========>.....................] - ETA: 7s - loss: 0.6184 - acc: 0.8049 629/1875 [=========>....................] - ETA: 7s - loss: 0.6131 - acc: 0.8066 638/1875 [=========>....................] - ETA: 7s - loss: 0.6085 - acc: 0.8081 647/1875 [=========>....................] - ETA: 7s - loss: 0.6040 - acc: 0.8095 656/1875 [=========>....................] - ETA: 7s - loss: 0.5995 - acc: 0.8109 665/1875 [=========>....................] - ETA: 7s - loss: 0.5952 - acc: 0.8123 674/1875 [=========>....................] - ETA: 7s - loss: 0.5909 - acc: 0.8137 684/1875 [=========>....................] - ETA: 6s - loss: 0.5862 - acc: 0.8151 693/1875 [==========>...................] - ETA: 6s - loss: 0.5821 - acc: 0.8164 702/1875 [==========>...................] - ETA: 6s - loss: 0.5781 - acc: 0.8177 711/1875 [==========>...................] - ETA: 6s - loss: 0.5741 - acc: 0.8189 720/1875 [==========>...................] - ETA: 6s - loss: 0.5702 - acc: 0.8202 728/1875 [==========>...................] - ETA: 6s - loss: 0.5669 - acc: 0.8212 737/1875 [==========>...................] - ETA: 6s - loss: 0.5631 - acc: 0.8224 747/1875 [==========>...................] - ETA: 6s - loss: 0.5590 - acc: 0.8237 757/1875 [===========>..................] - ETA: 6s - loss: 0.5550 - acc: 0.8250 767/1875 [===========>..................] - ETA: 6s - loss: 0.5511 - acc: 0.8262 777/1875 [===========>..................] - ETA: 6s - loss: 0.5472 - acc: 0.8274 786/1875 [===========>..................] - ETA: 6s - loss: 0.5438 - acc: 0.8285 795/1875 [===========>..................] - ETA: 6s - loss: 0.5405 - acc: 0.8296 804/1875 [===========>..................] - ETA: 6s - loss: 0.5372 - acc: 0.8306 814/1875 [============>.................] - ETA: 6s - loss: 0.5336 - acc: 0.8318 824/1875 [============>.................] - ETA: 6s - loss: 0.5300 - acc: 0.8329 834/1875 [============>.................] - ETA: 6s - loss: 0.5266 - acc: 0.8340 844/1875 [============>.................] - ETA: 6s - loss: 0.5232 - acc: 0.8350 853/1875 [============>.................] - ETA: 5s - loss: 0.5202 - acc: 0.8360 862/1875 [============>.................] - ETA: 5s - loss: 0.5172 - acc: 0.8369 872/1875 [============>.................] - ETA: 5s - loss: 0.5139 - acc: 0.8380 882/1875 [=============>................] - ETA: 5s - loss: 0.5107 - acc: 0.8390 892/1875 [=============>................] - ETA: 5s - loss: 0.5076 - acc: 0.8400 901/1875 [=============>................] - ETA: 5s - loss: 0.5048 - acc: 0.8408 910/1875 [=============>................] - ETA: 5s - loss: 0.5021 - acc: 0.8417 919/1875 [=============>................] - ETA: 5s - loss: 0.4994 - acc: 0.8426 928/1875 [=============>................] - ETA: 5s - loss: 0.4967 - acc: 0.8434 938/1875 [==============>...............] - ETA: 5s - loss: 0.4938 - acc: 0.8443 948/1875 [==============>...............] - ETA: 5s - loss: 0.4910 - acc: 0.8452 957/1875 [==============>...............] - ETA: 5s - loss: 0.4884 - acc: 0.8460 967/1875 [==============>...............] - ETA: 5s - loss: 0.4856 - acc: 0.8469 977/1875 [==============>...............] - ETA: 5s - loss: 0.4829 - acc: 0.8478 986/1875 [==============>...............] - ETA: 5s - loss: 0.4805 - acc: 0.8486 995/1875 [==============>...............] - ETA: 5s - loss: 0.4781 - acc: 0.8493 1004/1875 [===============>..............] - ETA: 5s - loss: 0.4757 - acc: 0.8501 1013/1875 [===============>..............] - ETA: 4s - loss: 0.4734 - acc: 0.8508 1023/1875 [===============>..............] - ETA: 4s - loss: 0.4709 - acc: 0.8516 1033/1875 [===============>..............] - ETA: 4s - loss: 0.4683 - acc: 0.8524 1042/1875 [===============>..............] - ETA: 4s - loss: 0.4661 - acc: 0.8531 1050/1875 [===============>..............] - ETA: 4s - loss: 0.4641 - acc: 0.8538 1059/1875 [===============>..............] - ETA: 4s - loss: 0.4620 - acc: 0.8544 1068/1875 [================>.............] - ETA: 4s - loss: 0.4598 - acc: 0.8551 1077/1875 [================>.............] - ETA: 4s - loss: 0.4577 - acc: 0.8558 1086/1875 [================>.............] - ETA: 4s - loss: 0.4556 - acc: 0.8565 1095/1875 [================>.............] - ETA: 4s - loss: 0.4535 - acc: 0.8571 1105/1875 [================>.............] - ETA: 4s - loss: 0.4513 - acc: 0.8578 1114/1875 [================>.............] - ETA: 4s - loss: 0.4493 - acc: 0.8585 1123/1875 [================>.............] - ETA: 4s - loss: 0.4473 - acc: 0.8591 1132/1875 [=================>............] - ETA: 4s - loss: 0.4453 - acc: 0.8597 1141/1875 [=================>............] - ETA: 4s - loss: 0.4433 - acc: 0.8604 1150/1875 [=================>............] - ETA: 4s - loss: 0.4414 - acc: 0.8610 1159/1875 [=================>............] - ETA: 4s - loss: 0.4395 - acc: 0.8616 1168/1875 [=================>............] - ETA: 4s - loss: 0.4376 - acc: 0.8622 1177/1875 [=================>............] - ETA: 4s - loss: 0.4357 - acc: 0.8628 1186/1875 [=================>............] - ETA: 3s - loss: 0.4339 - acc: 0.8633 1196/1875 [==================>...........] - ETA: 3s - loss: 0.4319 - acc: 0.8640 1205/1875 [==================>...........] - ETA: 3s - loss: 0.4301 - acc: 0.8645 1214/1875 [==================>...........] - ETA: 3s - loss: 0.4283 - acc: 0.8651 1223/1875 [==================>...........] - ETA: 3s - loss: 0.4266 - acc: 0.8657 1232/1875 [==================>...........] - ETA: 3s - loss: 0.4248 - acc: 0.8662 1241/1875 [==================>...........] - ETA: 3s - loss: 0.4231 - acc: 0.8668 1250/1875 [===================>..........] - ETA: 3s - loss: 0.4214 - acc: 0.8673 1259/1875 [===================>..........] - ETA: 3s - loss: 0.4197 - acc: 0.8678 1268/1875 [===================>..........] - ETA: 3s - loss: 0.4180 - acc: 0.8684 1277/1875 [===================>..........] - ETA: 3s - loss: 0.4164 - acc: 0.8689 1286/1875 [===================>..........] - ETA: 3s - loss: 0.4148 - acc: 0.8694 1295/1875 [===================>..........] - ETA: 3s - loss: 0.4132 - acc: 0.8699 1304/1875 [===================>..........] - ETA: 3s - loss: 0.4116 - acc: 0.8704 1313/1875 [====================>.........] - ETA: 3s - loss: 0.4100 - acc: 0.8709 1322/1875 [====================>.........] - ETA: 3s - loss: 0.4084 - acc: 0.8714 1331/1875 [====================>.........] - ETA: 3s - loss: 0.4069 - acc: 0.8719 1340/1875 [====================>.........] - ETA: 3s - loss: 0.4053 - acc: 0.8724 1349/1875 [====================>.........] - ETA: 3s - loss: 0.4038 - acc: 0.8728 1358/1875 [====================>.........] - ETA: 2s - loss: 0.4023 - acc: 0.8733 1367/1875 [====================>.........] - ETA: 2s - loss: 0.4008 - acc: 0.8738 1376/1875 [=====================>........] - ETA: 2s - loss: 0.3993 - acc: 0.8743 1386/1875 [=====================>........] - ETA: 2s - loss: 0.3977 - acc: 0.8748 1395/1875 [=====================>........] - ETA: 2s - loss: 0.3963 - acc: 0.8752 1404/1875 [=====================>........] - ETA: 2s - loss: 0.3948 - acc: 0.8757 1413/1875 [=====================>........] - ETA: 2s - loss: 0.3934 - acc: 0.8761 1422/1875 [=====================>........] - ETA: 2s - loss: 0.3920 - acc: 0.8766 1430/1875 [=====================>........] - ETA: 2s - loss: 0.3908 - acc: 0.8769 1439/1875 [======================>.......] - ETA: 2s - loss: 0.3894 - acc: 0.8774 1449/1875 [======================>.......] - ETA: 2s - loss: 0.3879 - acc: 0.8779 1459/1875 [======================>.......] - ETA: 2s - loss: 0.3864 - acc: 0.8783 1469/1875 [======================>.......] - ETA: 2s - loss: 0.3849 - acc: 0.8788 1478/1875 [======================>.......] - ETA: 2s - loss: 0.3836 - acc: 0.8792 1487/1875 [======================>.......] - ETA: 2s - loss: 0.3823 - acc: 0.8796 1496/1875 [======================>.......] - ETA: 2s - loss: 0.3810 - acc: 0.8800 1505/1875 [=======================>......] - ETA: 2s - loss: 0.3797 - acc: 0.8804 1514/1875 [=======================>......] - ETA: 2s - loss: 0.3784 - acc: 0.8808 1523/1875 [=======================>......] - ETA: 2s - loss: 0.3772 - acc: 0.8812 1532/1875 [=======================>......] - ETA: 1s - loss: 0.3759 - acc: 0.8816 1541/1875 [=======================>......] - ETA: 1s - loss: 0.3747 - acc: 0.8820 1550/1875 [=======================>......] - ETA: 1s - loss: 0.3734 - acc: 0.8824 1559/1875 [=======================>......] - ETA: 1s - loss: 0.3722 - acc: 0.8828 1567/1875 [========================>.....] - ETA: 1s - loss: 0.3712 - acc: 0.8831 1583/1875 [========================>.....] - ETA: 1s - loss: 0.3690 - acc: 0.8838 1599/1875 [========================>.....] - ETA: 1s - loss: 0.3669 - acc: 0.8845 1614/1875 [========================>.....] - ETA: 1s - loss: 0.3650 - acc: 0.8851 1629/1875 [=========================>....] - ETA: 1s - loss: 0.3631 - acc: 0.8857 1644/1875 [=========================>....] - ETA: 1s - loss: 0.3612 - acc: 0.8863 1660/1875 [=========================>....] - ETA: 1s - loss: 0.3592 - acc: 0.8869 1676/1875 [=========================>....] - ETA: 1s - loss: 0.3573 - acc: 0.8875 1692/1875 [==========================>...] - ETA: 1s - loss: 0.3554 - acc: 0.8881 1708/1875 [==========================>...] - ETA: 0s - loss: 0.3535 - acc: 0.8887 1724/1875 [==========================>...] - ETA: 0s - loss: 0.3516 - acc: 0.8893 1739/1875 [==========================>...] - ETA: 0s - loss: 0.3499 - acc: 0.8898 1754/1875 [===========================>..] - ETA: 0s - loss: 0.3482 - acc: 0.8904 1770/1875 [===========================>..] - ETA: 0s - loss: 0.3464 - acc: 0.8909 1786/1875 [===========================>..] - ETA: 0s - loss: 0.3446 - acc: 0.8915 1801/1875 [===========================>..] - ETA: 0s - loss: 0.3430 - acc: 0.8920 1816/1875 [============================>.] - ETA: 0s - loss: 0.3414 - acc: 0.8925 1831/1875 [============================>.] - ETA: 0s - loss: 0.3398 - acc: 0.8930 1847/1875 [============================>.] - ETA: 0s - loss: 0.3381 - acc: 0.8936 1861/1875 [============================>.] - ETA: 0s - loss: 0.3366 - acc: 0.8940 1875/1875 [==============================] - ETA: 0s - loss: 0.3352 - acc: 0.8945 1875/1875 [==============================] - 13s 5ms/step - loss: 0.3351 - acc: 0.8945 Epoch 2/5 1/1875 [..............................] - ETA: 14:27 - loss: 0.0887 - acc: 0.9688 11/1875 [..............................] - ETA: 9s - loss: 0.0596 - acc: 0.9750 22/1875 [..............................] - ETA: 9s - loss: 0.0726 - acc: 0.9753 32/1875 [..............................] - ETA: 9s - loss: 0.0734 - acc: 0.9767 42/1875 [..............................] - ETA: 9s - loss: 0.0720 - acc: 0.9778 52/1875 [..............................] - ETA: 9s - loss: 0.0703 - acc: 0.9784 63/1875 [>.............................] - ETA: 9s - loss: 0.0688 - acc: 0.9790 73/1875 [>.............................] - ETA: 9s - loss: 0.0682 - acc: 0.9792 84/1875 [>.............................] - ETA: 9s - loss: 0.0675 - acc: 0.9794 94/1875 [>.............................] - ETA: 9s - loss: 0.0667 - acc: 0.9796 104/1875 [>.............................] - ETA: 9s - loss: 0.0662 - acc: 0.9797 114/1875 [>.............................] - ETA: 9s - loss: 0.0657 - acc: 0.9799 125/1875 [=>............................] - ETA: 9s - loss: 0.0651 - acc: 0.9801 136/1875 [=>............................] - ETA: 8s - loss: 0.0646 - acc: 0.9803 147/1875 [=>............................] - ETA: 8s - loss: 0.0641 - acc: 0.9805 157/1875 [=>............................] - ETA: 8s - loss: 0.0635 - acc: 0.9807 167/1875 [=>............................] - ETA: 8s - loss: 0.0629 - acc: 0.9809 177/1875 [=>............................] - ETA: 8s - loss: 0.0625 - acc: 0.9811 187/1875 [=>............................] - ETA: 8s - loss: 0.0621 - acc: 0.9812 197/1875 [==>...........................] - ETA: 8s - loss: 0.0617 - acc: 0.9814 208/1875 [==>...........................] - ETA: 8s - loss: 0.0613 - acc: 0.9815 218/1875 [==>...........................] - ETA: 8s - loss: 0.0610 - acc: 0.9816 229/1875 [==>...........................] - ETA: 8s - loss: 0.0606 - acc: 0.9818 240/1875 [==>...........................] - ETA: 8s - loss: 0.0603 - acc: 0.9819 250/1875 [===>..........................] - ETA: 8s - loss: 0.0599 - acc: 0.9820 260/1875 [===>..........................] - ETA: 8s - loss: 0.0596 - acc: 0.9821 270/1875 [===>..........................] - ETA: 8s - loss: 0.0593 - acc: 0.9822 281/1875 [===>..........................] - ETA: 8s - loss: 0.0591 - acc: 0.9823 291/1875 [===>..........................] - ETA: 8s - loss: 0.0588 - acc: 0.9824 301/1875 [===>..........................] - ETA: 8s - loss: 0.0586 - acc: 0.9824 311/1875 [===>..........................] - ETA: 8s - loss: 0.0584 - acc: 0.9825 321/1875 [====>.........................] - ETA: 8s - loss: 0.0582 - acc: 0.9825 331/1875 [====>.........................] - ETA: 7s - loss: 0.0580 - acc: 0.9826 341/1875 [====>.........................] - ETA: 7s - loss: 0.0579 - acc: 0.9826 351/1875 [====>.........................] - ETA: 7s - loss: 0.0577 - acc: 0.9827 361/1875 [====>.........................] - ETA: 7s - loss: 0.0575 - acc: 0.9827 371/1875 [====>.........................] - ETA: 7s - loss: 0.0574 - acc: 0.9828 381/1875 [=====>........................] - ETA: 7s - loss: 0.0572 - acc: 0.9828 390/1875 [=====>........................] - ETA: 7s - loss: 0.0571 - acc: 0.9828 400/1875 [=====>........................] - ETA: 7s - loss: 0.0569 - acc: 0.9829 410/1875 [=====>........................] - ETA: 7s - loss: 0.0568 - acc: 0.9829 420/1875 [=====>........................] - ETA: 7s - loss: 0.0566 - acc: 0.9830 430/1875 [=====>........................] - ETA: 7s - loss: 0.0565 - acc: 0.9830 440/1875 [======>.......................] - ETA: 7s - loss: 0.0563 - acc: 0.9830 450/1875 [======>.......................] - ETA: 7s - loss: 0.0562 - acc: 0.9831 460/1875 [======>.......................] - ETA: 7s - loss: 0.0560 - acc: 0.9831 470/1875 [======>.......................] - ETA: 7s - loss: 0.0559 - acc: 0.9832 480/1875 [======>.......................] - ETA: 7s - loss: 0.0558 - acc: 0.9832 490/1875 [======>.......................] - ETA: 7s - loss: 0.0556 - acc: 0.9832 500/1875 [=======>......................] - ETA: 7s - loss: 0.0555 - acc: 0.9833 510/1875 [=======>......................] - ETA: 7s - loss: 0.0554 - acc: 0.9833 520/1875 [=======>......................] - ETA: 7s - loss: 0.0553 - acc: 0.9833 529/1875 [=======>......................] - ETA: 7s - loss: 0.0552 - acc: 0.9834 539/1875 [=======>......................] - ETA: 6s - loss: 0.0551 - acc: 0.9834 549/1875 [=======>......................] - ETA: 6s - loss: 0.0550 - acc: 0.9834 560/1875 [=======>......................] - ETA: 6s - loss: 0.0548 - acc: 0.9835 570/1875 [========>.....................] - ETA: 6s - loss: 0.0547 - acc: 0.9835 580/1875 [========>.....................] - ETA: 6s - loss: 0.0546 - acc: 0.9835 591/1875 [========>.....................] - ETA: 6s - loss: 0.0545 - acc: 0.9836 600/1875 [========>.....................] - ETA: 6s - loss: 0.0544 - acc: 0.9836 610/1875 [========>.....................] - ETA: 6s - loss: 0.0543 - acc: 0.9836 620/1875 [========>.....................] - ETA: 6s - loss: 0.0542 - acc: 0.9837 630/1875 [=========>....................] - ETA: 6s - loss: 0.0541 - acc: 0.9837 641/1875 [=========>....................] - ETA: 6s - loss: 0.0540 - acc: 0.9837 652/1875 [=========>....................] - ETA: 6s - loss: 0.0539 - acc: 0.9838 662/1875 [=========>....................] - ETA: 6s - loss: 0.0538 - acc: 0.9838 672/1875 [=========>....................] - ETA: 6s - loss: 0.0537 - acc: 0.9838 681/1875 [=========>....................] - ETA: 6s - loss: 0.0536 - acc: 0.9839 691/1875 [==========>...................] - ETA: 6s - loss: 0.0535 - acc: 0.9839 701/1875 [==========>...................] - ETA: 6s - loss: 0.0534 - acc: 0.9839 711/1875 [==========>...................] - ETA: 6s - loss: 0.0533 - acc: 0.9840 721/1875 [==========>...................] - ETA: 6s - loss: 0.0532 - acc: 0.9840 731/1875 [==========>...................] - ETA: 5s - loss: 0.0532 - acc: 0.9840 740/1875 [==========>...................] - ETA: 5s - loss: 0.0531 - acc: 0.9840 750/1875 [===========>..................] - ETA: 5s - loss: 0.0530 - acc: 0.9841 760/1875 [===========>..................] - ETA: 5s - loss: 0.0530 - acc: 0.9841 770/1875 [===========>..................] - ETA: 5s - loss: 0.0529 - acc: 0.9841 780/1875 [===========>..................] - ETA: 5s - loss: 0.0528 - acc: 0.9841 790/1875 [===========>..................] - ETA: 5s - loss: 0.0528 - acc: 0.9841 800/1875 [===========>..................] - ETA: 5s - loss: 0.0527 - acc: 0.9842 810/1875 [===========>..................] - ETA: 5s - loss: 0.0527 - acc: 0.9842 820/1875 [============>.................] - ETA: 5s - loss: 0.0527 - acc: 0.9842 830/1875 [============>.................] - ETA: 5s - loss: 0.0526 - acc: 0.9842 840/1875 [============>.................] - ETA: 5s - loss: 0.0526 - acc: 0.9842 850/1875 [============>.................] - ETA: 5s - loss: 0.0525 - acc: 0.9842 860/1875 [============>.................] - ETA: 5s - loss: 0.0525 - acc: 0.9842 869/1875 [============>.................] - ETA: 5s - loss: 0.0525 - acc: 0.9843 878/1875 [=============>................] - ETA: 5s - loss: 0.0524 - acc: 0.9843 888/1875 [=============>................] - ETA: 5s - loss: 0.0524 - acc: 0.9843 898/1875 [=============>................] - ETA: 5s - loss: 0.0524 - acc: 0.9843 908/1875 [=============>................] - ETA: 5s - loss: 0.0523 - acc: 0.9843 918/1875 [=============>................] - ETA: 5s - loss: 0.0523 - acc: 0.9843 928/1875 [=============>................] - ETA: 4s - loss: 0.0523 - acc: 0.9843 938/1875 [==============>...............] - ETA: 4s - loss: 0.0523 - acc: 0.9843 948/1875 [==============>...............] - ETA: 4s - loss: 0.0522 - acc: 0.9844 958/1875 [==============>...............] - ETA: 4s - loss: 0.0522 - acc: 0.9844 968/1875 [==============>...............] - ETA: 4s - loss: 0.0522 - acc: 0.9844 978/1875 [==============>...............] - ETA: 4s - loss: 0.0521 - acc: 0.9844 988/1875 [==============>...............] - ETA: 4s - loss: 0.0521 - acc: 0.9844 998/1875 [==============>...............] - ETA: 4s - loss: 0.0521 - acc: 0.9844 1008/1875 [===============>..............] - ETA: 4s - loss: 0.0521 - acc: 0.9844 1018/1875 [===============>..............] - ETA: 4s - loss: 0.0520 - acc: 0.9844 1028/1875 [===============>..............] - ETA: 4s - loss: 0.0520 - acc: 0.9844 1038/1875 [===============>..............] - ETA: 4s - loss: 0.0520 - acc: 0.9844 1048/1875 [===============>..............] - ETA: 4s - loss: 0.0520 - acc: 0.9845 1058/1875 [===============>..............] - ETA: 4s - loss: 0.0519 - acc: 0.9845 1067/1875 [================>.............] - ETA: 4s - loss: 0.0519 - acc: 0.9845 1076/1875 [================>.............] - ETA: 4s - loss: 0.0519 - acc: 0.9845 1085/1875 [================>.............] - ETA: 4s - loss: 0.0519 - acc: 0.9845 1095/1875 [================>.............] - ETA: 4s - loss: 0.0518 - acc: 0.9845 1105/1875 [================>.............] - ETA: 4s - loss: 0.0518 - acc: 0.9845 1115/1875 [================>.............] - ETA: 3s - loss: 0.0518 - acc: 0.9845 1125/1875 [=================>............] - ETA: 3s - loss: 0.0518 - acc: 0.9845 1135/1875 [=================>............] - ETA: 3s - loss: 0.0518 - acc: 0.9845 1144/1875 [=================>............] - ETA: 3s - loss: 0.0517 - acc: 0.9845 1153/1875 [=================>............] - ETA: 3s - loss: 0.0517 - acc: 0.9845 1163/1875 [=================>............] - ETA: 3s - loss: 0.0517 - acc: 0.9845 1173/1875 [=================>............] - ETA: 3s - loss: 0.0517 - acc: 0.9845 1183/1875 [=================>............] - ETA: 3s - loss: 0.0516 - acc: 0.9846 1193/1875 [==================>...........] - ETA: 3s - loss: 0.0516 - acc: 0.9846 1203/1875 [==================>...........] - ETA: 3s - loss: 0.0516 - acc: 0.9846 1212/1875 [==================>...........] - ETA: 3s - loss: 0.0516 - acc: 0.9846 1222/1875 [==================>...........] - ETA: 3s - loss: 0.0515 - acc: 0.9846 1232/1875 [==================>...........] - ETA: 3s - loss: 0.0515 - acc: 0.9846 1241/1875 [==================>...........] - ETA: 3s - loss: 0.0515 - acc: 0.9846 1251/1875 [===================>..........] - ETA: 3s - loss: 0.0515 - acc: 0.9846 1261/1875 [===================>..........] - ETA: 3s - loss: 0.0514 - acc: 0.9846 1271/1875 [===================>..........] - ETA: 3s - loss: 0.0514 - acc: 0.9846 1281/1875 [===================>..........] - ETA: 3s - loss: 0.0514 - acc: 0.9846 1291/1875 [===================>..........] - ETA: 3s - loss: 0.0514 - acc: 0.9846 1302/1875 [===================>..........] - ETA: 3s - loss: 0.0513 - acc: 0.9846 1313/1875 [====================>.........] - ETA: 2s - loss: 0.0513 - acc: 0.9846 1323/1875 [====================>.........] - ETA: 2s - loss: 0.0513 - acc: 0.9846 1333/1875 [====================>.........] - ETA: 2s - loss: 0.0513 - acc: 0.9847 1343/1875 [====================>.........] - ETA: 2s - loss: 0.0512 - acc: 0.9847 1352/1875 [====================>.........] - ETA: 2s - loss: 0.0512 - acc: 0.9847 1362/1875 [====================>.........] - ETA: 2s - loss: 0.0512 - acc: 0.9847 1372/1875 [====================>.........] - ETA: 2s - loss: 0.0512 - acc: 0.9847 1383/1875 [=====================>........] - ETA: 2s - loss: 0.0511 - acc: 0.9847 1393/1875 [=====================>........] - ETA: 2s - loss: 0.0511 - acc: 0.9847 1403/1875 [=====================>........] - ETA: 2s - loss: 0.0511 - acc: 0.9847 1413/1875 [=====================>........] - ETA: 2s - loss: 0.0511 - acc: 0.9847 1422/1875 [=====================>........] - ETA: 2s - loss: 0.0510 - acc: 0.9847 1432/1875 [=====================>........] - ETA: 2s - loss: 0.0510 - acc: 0.9847 1443/1875 [======================>.......] - ETA: 2s - loss: 0.0510 - acc: 0.9847 1453/1875 [======================>.......] - ETA: 2s - loss: 0.0510 - acc: 0.9847 1463/1875 [======================>.......] - ETA: 2s - loss: 0.0509 - acc: 0.9847 1474/1875 [======================>.......] - ETA: 2s - loss: 0.0509 - acc: 0.9848 1484/1875 [======================>.......] - ETA: 2s - loss: 0.0509 - acc: 0.9848 1494/1875 [======================>.......] - ETA: 2s - loss: 0.0509 - acc: 0.9848 1503/1875 [=======================>......] - ETA: 1s - loss: 0.0508 - acc: 0.9848 1513/1875 [=======================>......] - ETA: 1s - loss: 0.0508 - acc: 0.9848 1523/1875 [=======================>......] - ETA: 1s - loss: 0.0508 - acc: 0.9848 1533/1875 [=======================>......] - ETA: 1s - loss: 0.0508 - acc: 0.9848 1543/1875 [=======================>......] - ETA: 1s - loss: 0.0508 - acc: 0.9848 1553/1875 [=======================>......] - ETA: 1s - loss: 0.0507 - acc: 0.9848 1562/1875 [=======================>......] - ETA: 1s - loss: 0.0507 - acc: 0.9848 1576/1875 [========================>.....] - ETA: 1s - loss: 0.0507 - acc: 0.9848 1592/1875 [========================>.....] - ETA: 1s - loss: 0.0506 - acc: 0.9848 1607/1875 [========================>.....] - ETA: 1s - loss: 0.0506 - acc: 0.9848 1622/1875 [========================>.....] - ETA: 1s - loss: 0.0506 - acc: 0.9849 1637/1875 [=========================>....] - ETA: 1s - loss: 0.0505 - acc: 0.9849 1652/1875 [=========================>....] - ETA: 1s - loss: 0.0505 - acc: 0.9849 1666/1875 [=========================>....] - ETA: 1s - loss: 0.0505 - acc: 0.9849 1681/1875 [=========================>....] - ETA: 0s - loss: 0.0504 - acc: 0.9849 1697/1875 [==========================>...] - ETA: 0s - loss: 0.0504 - acc: 0.9849 1711/1875 [==========================>...] - ETA: 0s - loss: 0.0504 - acc: 0.9849 1725/1875 [==========================>...] - ETA: 0s - loss: 0.0503 - acc: 0.9849 1740/1875 [==========================>...] - ETA: 0s - loss: 0.0503 - acc: 0.9849 1756/1875 [===========================>..] - ETA: 0s - loss: 0.0503 - acc: 0.9849 1770/1875 [===========================>..] - ETA: 0s - loss: 0.0502 - acc: 0.9849 1784/1875 [===========================>..] - ETA: 0s - loss: 0.0502 - acc: 0.9850 1799/1875 [===========================>..] - ETA: 0s - loss: 0.0502 - acc: 0.9850 1814/1875 [============================>.] - ETA: 0s - loss: 0.0501 - acc: 0.9850 1828/1875 [============================>.] - ETA: 0s - loss: 0.0501 - acc: 0.9850 1843/1875 [============================>.] - ETA: 0s - loss: 0.0501 - acc: 0.9850 1858/1875 [============================>.] - ETA: 0s - loss: 0.0500 - acc: 0.9850 1872/1875 [============================>.] - ETA: 0s - loss: 0.0500 - acc: 0.9850 1875/1875 [==============================] - 10s 5ms/step - loss: 0.0500 - acc: 0.9850 Epoch 3/5 1/1875 [..............................] - ETA: 13:16 - loss: 0.0126 - acc: 1.0000 12/1875 [..............................] - ETA: 9s - loss: 0.0369 - acc: 0.9882 23/1875 [..............................] - ETA: 8s - loss: 0.0377 - acc: 0.9884 34/1875 [..............................] - ETA: 8s - loss: 0.0388 - acc: 0.9878 45/1875 [..............................] - ETA: 8s - loss: 0.0388 - acc: 0.9876 56/1875 [..............................] - ETA: 8s - loss: 0.0382 - acc: 0.9877 66/1875 [>.............................] - ETA: 8s - loss: 0.0376 - acc: 0.9877 76/1875 [>.............................] - ETA: 8s - loss: 0.0373 - acc: 0.9877 87/1875 [>.............................] - ETA: 8s - loss: 0.0373 - acc: 0.9878 97/1875 [>.............................] - ETA: 8s - loss: 0.0377 - acc: 0.9877 108/1875 [>.............................] - ETA: 8s - loss: 0.0381 - acc: 0.9876 118/1875 [>.............................] - ETA: 8s - loss: 0.0384 - acc: 0.9876 128/1875 [=>............................] - ETA: 8s - loss: 0.0386 - acc: 0.9876 138/1875 [=>............................] - ETA: 8s - loss: 0.0387 - acc: 0.9875 148/1875 [=>............................] - ETA: 8s - loss: 0.0388 - acc: 0.9875 158/1875 [=>............................] - ETA: 8s - loss: 0.0390 - acc: 0.9875 169/1875 [=>............................] - ETA: 8s - loss: 0.0391 - acc: 0.9874 179/1875 [=>............................] - ETA: 8s - loss: 0.0393 - acc: 0.9874 189/1875 [==>...........................] - ETA: 8s - loss: 0.0394 - acc: 0.9873 199/1875 [==>...........................] - ETA: 8s - loss: 0.0395 - acc: 0.9872 209/1875 [==>...........................] - ETA: 8s - loss: 0.0396 - acc: 0.9872 219/1875 [==>...........................] - ETA: 8s - loss: 0.0396 - acc: 0.9871 230/1875 [==>...........................] - ETA: 8s - loss: 0.0397 - acc: 0.9871 241/1875 [==>...........................] - ETA: 8s - loss: 0.0398 - acc: 0.9871 252/1875 [===>..........................] - ETA: 8s - loss: 0.0399 - acc: 0.9870 263/1875 [===>..........................] - ETA: 8s - loss: 0.0399 - acc: 0.9870 274/1875 [===>..........................] - ETA: 7s - loss: 0.0400 - acc: 0.9870 284/1875 [===>..........................] - ETA: 7s - loss: 0.0400 - acc: 0.9870 294/1875 [===>..........................] - ETA: 7s - loss: 0.0399 - acc: 0.9870 304/1875 [===>..........................] - ETA: 7s - loss: 0.0399 - acc: 0.9869 315/1875 [====>.........................] - ETA: 7s - loss: 0.0399 - acc: 0.9869 326/1875 [====>.........................] - ETA: 7s - loss: 0.0399 - acc: 0.9869 337/1875 [====>.........................] - ETA: 7s - loss: 0.0399 - acc: 0.9869 347/1875 [====>.........................] - ETA: 7s - loss: 0.0398 - acc: 0.9869 357/1875 [====>.........................] - ETA: 7s - loss: 0.0397 - acc: 0.9869 367/1875 [====>.........................] - ETA: 7s - loss: 0.0397 - acc: 0.9869 377/1875 [=====>........................] - ETA: 7s - loss: 0.0396 - acc: 0.9870 388/1875 [=====>........................] - ETA: 7s - loss: 0.0395 - acc: 0.9870 399/1875 [=====>........................] - ETA: 7s - loss: 0.0395 - acc: 0.9870 409/1875 [=====>........................] - ETA: 7s - loss: 0.0394 - acc: 0.9870 418/1875 [=====>........................] - ETA: 7s - loss: 0.0394 - acc: 0.9870 427/1875 [=====>........................] - ETA: 7s - loss: 0.0394 - acc: 0.9870 437/1875 [=====>........................] - ETA: 7s - loss: 0.0393 - acc: 0.9871 448/1875 [======>.......................] - ETA: 7s - loss: 0.0393 - acc: 0.9871 459/1875 [======>.......................] - ETA: 7s - loss: 0.0392 - acc: 0.9871 470/1875 [======>.......................] - ETA: 7s - loss: 0.0392 - acc: 0.9871 481/1875 [======>.......................] - ETA: 6s - loss: 0.0392 - acc: 0.9871 491/1875 [======>.......................] - ETA: 6s - loss: 0.0392 - acc: 0.9871 502/1875 [=======>......................] - ETA: 6s - loss: 0.0391 - acc: 0.9871 512/1875 [=======>......................] - ETA: 6s - loss: 0.0391 - acc: 0.9872 523/1875 [=======>......................] - ETA: 6s - loss: 0.0390 - acc: 0.9872 534/1875 [=======>......................] - ETA: 6s - loss: 0.0390 - acc: 0.9872 545/1875 [=======>......................] - ETA: 6s - loss: 0.0389 - acc: 0.9872 556/1875 [=======>......................] - ETA: 6s - loss: 0.0389 - acc: 0.9873 567/1875 [========>.....................] - ETA: 6s - loss: 0.0388 - acc: 0.9873 578/1875 [========>.....................] - ETA: 6s - loss: 0.0387 - acc: 0.9873 589/1875 [========>.....................] - ETA: 6s - loss: 0.0387 - acc: 0.9873 599/1875 [========>.....................] - ETA: 6s - loss: 0.0386 - acc: 0.9874 610/1875 [========>.....................] - ETA: 6s - loss: 0.0385 - acc: 0.9874 621/1875 [========>.....................] - ETA: 6s - loss: 0.0385 - acc: 0.9874 632/1875 [=========>....................] - ETA: 6s - loss: 0.0384 - acc: 0.9874 642/1875 [=========>....................] - ETA: 6s - loss: 0.0384 - acc: 0.9874 652/1875 [=========>....................] - ETA: 6s - loss: 0.0383 - acc: 0.9875 662/1875 [=========>....................] - ETA: 6s - loss: 0.0383 - acc: 0.9875 673/1875 [=========>....................] - ETA: 6s - loss: 0.0382 - acc: 0.9875 684/1875 [=========>....................] - ETA: 5s - loss: 0.0382 - acc: 0.9875 695/1875 [==========>...................] - ETA: 5s - loss: 0.0381 - acc: 0.9875 706/1875 [==========>...................] - ETA: 5s - loss: 0.0381 - acc: 0.9876 717/1875 [==========>...................] - ETA: 5s - loss: 0.0380 - acc: 0.9876 728/1875 [==========>...................] - ETA: 5s - loss: 0.0380 - acc: 0.9876 739/1875 [==========>...................] - ETA: 5s - loss: 0.0380 - acc: 0.9876 750/1875 [===========>..................] - ETA: 5s - loss: 0.0379 - acc: 0.9876 761/1875 [===========>..................] - ETA: 5s - loss: 0.0379 - acc: 0.9876 772/1875 [===========>..................] - ETA: 5s - loss: 0.0378 - acc: 0.9877 782/1875 [===========>..................] - ETA: 5s - loss: 0.0378 - acc: 0.9877 792/1875 [===========>..................] - ETA: 5s - loss: 0.0378 - acc: 0.9877 803/1875 [===========>..................] - ETA: 5s - loss: 0.0377 - acc: 0.9877 813/1875 [============>.................] - ETA: 5s - loss: 0.0377 - acc: 0.9877 824/1875 [============>.................] - ETA: 5s - loss: 0.0376 - acc: 0.9877 835/1875 [============>.................] - ETA: 5s - loss: 0.0376 - acc: 0.9877 846/1875 [============>.................] - ETA: 5s - loss: 0.0376 - acc: 0.9878 857/1875 [============>.................] - ETA: 5s - loss: 0.0375 - acc: 0.9878 867/1875 [============>.................] - ETA: 4s - loss: 0.0375 - acc: 0.9878 877/1875 [=============>................] - ETA: 4s - loss: 0.0375 - acc: 0.9878 888/1875 [=============>................] - ETA: 4s - loss: 0.0375 - acc: 0.9878 899/1875 [=============>................] - ETA: 4s - loss: 0.0374 - acc: 0.9878 909/1875 [=============>................] - ETA: 4s - loss: 0.0374 - acc: 0.9878 920/1875 [=============>................] - ETA: 4s - loss: 0.0374 - acc: 0.9879 931/1875 [=============>................] - ETA: 4s - loss: 0.0373 - acc: 0.9879 942/1875 [==============>...............] - ETA: 4s - loss: 0.0373 - acc: 0.9879 952/1875 [==============>...............] - ETA: 4s - loss: 0.0373 - acc: 0.9879 963/1875 [==============>...............] - ETA: 4s - loss: 0.0372 - acc: 0.9879 974/1875 [==============>...............] - ETA: 4s - loss: 0.0372 - acc: 0.9879 985/1875 [==============>...............] - ETA: 4s - loss: 0.0372 - acc: 0.9879 995/1875 [==============>...............] - ETA: 4s - loss: 0.0372 - acc: 0.9879 1006/1875 [===============>..............] - ETA: 4s - loss: 0.0371 - acc: 0.9879 1017/1875 [===============>..............] - ETA: 4s - loss: 0.0371 - acc: 0.9880 1028/1875 [===============>..............] - ETA: 4s - loss: 0.0371 - acc: 0.9880 1039/1875 [===============>..............] - ETA: 4s - loss: 0.0371 - acc: 0.9880 1050/1875 [===============>..............] - ETA: 4s - loss: 0.0370 - acc: 0.9880 1061/1875 [===============>..............] - ETA: 4s - loss: 0.0370 - acc: 0.9880 1072/1875 [================>.............] - ETA: 3s - loss: 0.0370 - acc: 0.9880 1082/1875 [================>.............] - ETA: 3s - loss: 0.0370 - acc: 0.9880 1092/1875 [================>.............] - ETA: 3s - loss: 0.0369 - acc: 0.9880 1103/1875 [================>.............] - ETA: 3s - loss: 0.0369 - acc: 0.9880 1114/1875 [================>.............] - ETA: 3s - loss: 0.0369 - acc: 0.9881 1124/1875 [================>.............] - ETA: 3s - loss: 0.0368 - acc: 0.9881 1135/1875 [=================>............] - ETA: 3s - loss: 0.0368 - acc: 0.9881 1146/1875 [=================>............] - ETA: 3s - loss: 0.0368 - acc: 0.9881 1157/1875 [=================>............] - ETA: 3s - loss: 0.0368 - acc: 0.9881 1168/1875 [=================>............] - ETA: 3s - loss: 0.0367 - acc: 0.9881 1179/1875 [=================>............] - ETA: 3s - loss: 0.0367 - acc: 0.9881 1190/1875 [==================>...........] - ETA: 3s - loss: 0.0367 - acc: 0.9881 1201/1875 [==================>...........] - ETA: 3s - loss: 0.0366 - acc: 0.9882 1212/1875 [==================>...........] - ETA: 3s - loss: 0.0366 - acc: 0.9882 1223/1875 [==================>...........] - ETA: 3s - loss: 0.0366 - acc: 0.9882 1234/1875 [==================>...........] - ETA: 3s - loss: 0.0366 - acc: 0.9882 1244/1875 [==================>...........] - ETA: 3s - loss: 0.0365 - acc: 0.9882 1255/1875 [===================>..........] - ETA: 3s - loss: 0.0365 - acc: 0.9882 1266/1875 [===================>..........] - ETA: 3s - loss: 0.0365 - acc: 0.9882 1277/1875 [===================>..........] - ETA: 2s - loss: 0.0364 - acc: 0.9882 1288/1875 [===================>..........] - ETA: 2s - loss: 0.0364 - acc: 0.9882 1299/1875 [===================>..........] - ETA: 2s - loss: 0.0364 - acc: 0.9883 1310/1875 [===================>..........] - ETA: 2s - loss: 0.0364 - acc: 0.9883 1321/1875 [====================>.........] - ETA: 2s - loss: 0.0363 - acc: 0.9883 1332/1875 [====================>.........] - ETA: 2s - loss: 0.0363 - acc: 0.9883 1343/1875 [====================>.........] - ETA: 2s - loss: 0.0363 - acc: 0.9883 1354/1875 [====================>.........] - ETA: 2s - loss: 0.0363 - acc: 0.9883 1365/1875 [====================>.........] - ETA: 2s - loss: 0.0362 - acc: 0.9883 1376/1875 [=====================>........] - ETA: 2s - loss: 0.0362 - acc: 0.9883 1386/1875 [=====================>........] - ETA: 2s - loss: 0.0362 - acc: 0.9883 1397/1875 [=====================>........] - ETA: 2s - loss: 0.0362 - acc: 0.9883 1408/1875 [=====================>........] - ETA: 2s - loss: 0.0361 - acc: 0.9883 1419/1875 [=====================>........] - ETA: 2s - loss: 0.0361 - acc: 0.9883 1430/1875 [=====================>........] - ETA: 2s - loss: 0.0361 - acc: 0.9883 1441/1875 [======================>.......] - ETA: 2s - loss: 0.0361 - acc: 0.9884 1452/1875 [======================>.......] - ETA: 2s - loss: 0.0361 - acc: 0.9884 1462/1875 [======================>.......] - ETA: 2s - loss: 0.0360 - acc: 0.9884 1472/1875 [======================>.......] - ETA: 1s - loss: 0.0360 - acc: 0.9884 1483/1875 [======================>.......] - ETA: 1s - loss: 0.0360 - acc: 0.9884 1494/1875 [======================>.......] - ETA: 1s - loss: 0.0360 - acc: 0.9884 1505/1875 [=======================>......] - ETA: 1s - loss: 0.0359 - acc: 0.9884 1516/1875 [=======================>......] - ETA: 1s - loss: 0.0359 - acc: 0.9884 1527/1875 [=======================>......] - ETA: 1s - loss: 0.0359 - acc: 0.9884 1537/1875 [=======================>......] - ETA: 1s - loss: 0.0359 - acc: 0.9884 1547/1875 [=======================>......] - ETA: 1s - loss: 0.0359 - acc: 0.9884 1558/1875 [=======================>......] - ETA: 1s - loss: 0.0358 - acc: 0.9884 1569/1875 [========================>.....] - ETA: 1s - loss: 0.0358 - acc: 0.9884 1585/1875 [========================>.....] - ETA: 1s - loss: 0.0358 - acc: 0.9885 1601/1875 [========================>.....] - ETA: 1s - loss: 0.0357 - acc: 0.9885 1616/1875 [========================>.....] - ETA: 1s - loss: 0.0357 - acc: 0.9885 1630/1875 [=========================>....] - ETA: 1s - loss: 0.0357 - acc: 0.9885 1646/1875 [=========================>....] - ETA: 1s - loss: 0.0357 - acc: 0.9885 1661/1875 [=========================>....] - ETA: 1s - loss: 0.0356 - acc: 0.9885 1677/1875 [=========================>....] - ETA: 0s - loss: 0.0356 - acc: 0.9885 1692/1875 [==========================>...] - ETA: 0s - loss: 0.0356 - acc: 0.9885 1708/1875 [==========================>...] - ETA: 0s - loss: 0.0356 - acc: 0.9885 1723/1875 [==========================>...] - ETA: 0s - loss: 0.0355 - acc: 0.9885 1737/1875 [==========================>...] - ETA: 0s - loss: 0.0355 - acc: 0.9886 1752/1875 [===========================>..] - ETA: 0s - loss: 0.0355 - acc: 0.9886 1767/1875 [===========================>..] - ETA: 0s - loss: 0.0355 - acc: 0.9886 1783/1875 [===========================>..] - ETA: 0s - loss: 0.0354 - acc: 0.9886 1799/1875 [===========================>..] - ETA: 0s - loss: 0.0354 - acc: 0.9886 1814/1875 [============================>.] - ETA: 0s - loss: 0.0354 - acc: 0.9886 1830/1875 [============================>.] - ETA: 0s - loss: 0.0354 - acc: 0.9886 1845/1875 [============================>.] - ETA: 0s - loss: 0.0354 - acc: 0.9886 1857/1875 [============================>.] - ETA: 0s - loss: 0.0353 - acc: 0.9886 1867/1875 [============================>.] - ETA: 0s - loss: 0.0353 - acc: 0.9886 1875/1875 [==============================] - 9s 5ms/step - loss: 0.0353 - acc: 0.9886 Epoch 4/5 1/1875 [..............................] - ETA: 13:20 - loss: 0.0059 - acc: 1.0000 12/1875 [..............................] - ETA: 8s - loss: 0.0165 - acc: 0.9951 23/1875 [..............................] - ETA: 9s - loss: 0.0197 - acc: 0.9940 33/1875 [..............................] - ETA: 9s - loss: 0.0232 - acc: 0.9924 42/1875 [..............................] - ETA: 9s - loss: 0.0245 - acc: 0.9920 52/1875 [..............................] - ETA: 9s - loss: 0.0265 - acc: 0.9915 63/1875 [>.............................] - ETA: 9s - loss: 0.0279 - acc: 0.9910 74/1875 [>.............................] - ETA: 8s - loss: 0.0288 - acc: 0.9908 85/1875 [>.............................] - ETA: 8s - loss: 0.0292 - acc: 0.9906 96/1875 [>.............................] - ETA: 8s - loss: 0.0297 - acc: 0.9905 106/1875 [>.............................] - ETA: 8s - loss: 0.0301 - acc: 0.9904 116/1875 [>.............................] - ETA: 8s - loss: 0.0304 - acc: 0.9903 127/1875 [=>............................] - ETA: 8s - loss: 0.0307 - acc: 0.9902 138/1875 [=>............................] - ETA: 8s - loss: 0.0309 - acc: 0.9902 149/1875 [=>............................] - ETA: 8s - loss: 0.0311 - acc: 0.9901 160/1875 [=>............................] - ETA: 8s - loss: 0.0312 - acc: 0.9901 170/1875 [=>............................] - ETA: 8s - loss: 0.0312 - acc: 0.9901 180/1875 [=>............................] - ETA: 8s - loss: 0.0312 - acc: 0.9901 190/1875 [==>...........................] - ETA: 8s - loss: 0.0312 - acc: 0.9901 200/1875 [==>...........................] - ETA: 8s - loss: 0.0312 - acc: 0.9902 210/1875 [==>...........................] - ETA: 8s - loss: 0.0312 - acc: 0.9902 220/1875 [==>...........................] - ETA: 8s - loss: 0.0311 - acc: 0.9902 230/1875 [==>...........................] - ETA: 8s - loss: 0.0310 - acc: 0.9902 240/1875 [==>...........................] - ETA: 8s - loss: 0.0310 - acc: 0.9903 250/1875 [===>..........................] - ETA: 8s - loss: 0.0309 - acc: 0.9903 260/1875 [===>..........................] - ETA: 8s - loss: 0.0308 - acc: 0.9904 270/1875 [===>..........................] - ETA: 8s - loss: 0.0307 - acc: 0.9904 281/1875 [===>..........................] - ETA: 8s - loss: 0.0306 - acc: 0.9905 291/1875 [===>..........................] - ETA: 8s - loss: 0.0305 - acc: 0.9905 301/1875 [===>..........................] - ETA: 7s - loss: 0.0304 - acc: 0.9905 311/1875 [===>..........................] - ETA: 7s - loss: 0.0303 - acc: 0.9905 320/1875 [====>.........................] - ETA: 7s - loss: 0.0302 - acc: 0.9906 328/1875 [====>.........................] - ETA: 7s - loss: 0.0302 - acc: 0.9906 339/1875 [====>.........................] - ETA: 7s - loss: 0.0301 - acc: 0.9906 350/1875 [====>.........................] - ETA: 7s - loss: 0.0300 - acc: 0.9907 361/1875 [====>.........................] - ETA: 7s - loss: 0.0299 - acc: 0.9907 372/1875 [====>.........................] - ETA: 7s - loss: 0.0299 - acc: 0.9907 382/1875 [=====>........................] - ETA: 7s - loss: 0.0298 - acc: 0.9907 392/1875 [=====>........................] - ETA: 7s - loss: 0.0297 - acc: 0.9908 403/1875 [=====>........................] - ETA: 7s - loss: 0.0297 - acc: 0.9908 415/1875 [=====>........................] - ETA: 7s - loss: 0.0296 - acc: 0.9908 427/1875 [=====>........................] - ETA: 7s - loss: 0.0295 - acc: 0.9908 438/1875 [======>.......................] - ETA: 7s - loss: 0.0295 - acc: 0.9908 449/1875 [======>.......................] - ETA: 7s - loss: 0.0294 - acc: 0.9909 460/1875 [======>.......................] - ETA: 7s - loss: 0.0294 - acc: 0.9909 470/1875 [======>.......................] - ETA: 7s - loss: 0.0293 - acc: 0.9909 481/1875 [======>.......................] - ETA: 7s - loss: 0.0293 - acc: 0.9909 492/1875 [======>.......................] - ETA: 6s - loss: 0.0292 - acc: 0.9909 503/1875 [=======>......................] - ETA: 6s - loss: 0.0292 - acc: 0.9909 514/1875 [=======>......................] - ETA: 6s - loss: 0.0291 - acc: 0.9910 525/1875 [=======>......................] - ETA: 6s - loss: 0.0290 - acc: 0.9910 535/1875 [=======>......................] - ETA: 6s - loss: 0.0290 - acc: 0.9910 545/1875 [=======>......................] - ETA: 6s - loss: 0.0289 - acc: 0.9910 556/1875 [=======>......................] - ETA: 6s - loss: 0.0289 - acc: 0.9910 566/1875 [========>.....................] - ETA: 6s - loss: 0.0288 - acc: 0.9910 573/1875 [========>.....................] - ETA: 6s - loss: 0.0288 - acc: 0.9911 577/1875 [========>.....................] - ETA: 6s - loss: 0.0287 - acc: 0.9911 582/1875 [========>.....................] - ETA: 6s - loss: 0.0287 - acc: 0.9911 587/1875 [========>.....................] - ETA: 6s - loss: 0.0287 - acc: 0.9911 592/1875 [========>.....................] - ETA: 6s - loss: 0.0287 - acc: 0.9911 597/1875 [========>.....................] - ETA: 6s - loss: 0.0286 - acc: 0.9911 602/1875 [========>.....................] - ETA: 6s - loss: 0.0286 - acc: 0.9911 607/1875 [========>.....................] - ETA: 6s - loss: 0.0286 - acc: 0.9911 612/1875 [========>.....................] - ETA: 6s - loss: 0.0286 - acc: 0.9911 617/1875 [========>.....................] - ETA: 6s - loss: 0.0285 - acc: 0.9911 624/1875 [========>.....................] - ETA: 6s - loss: 0.0285 - acc: 0.9911 635/1875 [=========>....................] - ETA: 6s - loss: 0.0285 - acc: 0.9912 645/1875 [=========>....................] - ETA: 6s - loss: 0.0284 - acc: 0.9912 656/1875 [=========>....................] - ETA: 6s - loss: 0.0284 - acc: 0.9912 667/1875 [=========>....................] - ETA: 6s - loss: 0.0284 - acc: 0.9912 677/1875 [=========>....................] - ETA: 6s - loss: 0.0284 - acc: 0.9912 687/1875 [=========>....................] - ETA: 6s - loss: 0.0283 - acc: 0.9912 698/1875 [==========>...................] - ETA: 6s - loss: 0.0283 - acc: 0.9912 708/1875 [==========>...................] - ETA: 6s - loss: 0.0283 - acc: 0.9912 719/1875 [==========>...................] - ETA: 6s - loss: 0.0283 - acc: 0.9912 730/1875 [==========>...................] - ETA: 6s - loss: 0.0282 - acc: 0.9912 741/1875 [==========>...................] - ETA: 6s - loss: 0.0282 - acc: 0.9912 752/1875 [===========>..................] - ETA: 6s - loss: 0.0282 - acc: 0.9913 762/1875 [===========>..................] - ETA: 6s - loss: 0.0282 - acc: 0.9913 772/1875 [===========>..................] - ETA: 5s - loss: 0.0281 - acc: 0.9913 782/1875 [===========>..................] - ETA: 5s - loss: 0.0281 - acc: 0.9913 792/1875 [===========>..................] - ETA: 5s - loss: 0.0281 - acc: 0.9913 802/1875 [===========>..................] - ETA: 5s - loss: 0.0281 - acc: 0.9913 812/1875 [===========>..................] - ETA: 5s - loss: 0.0281 - acc: 0.9913 823/1875 [============>.................] - ETA: 5s - loss: 0.0280 - acc: 0.9913 833/1875 [============>.................] - ETA: 5s - loss: 0.0280 - acc: 0.9913 843/1875 [============>.................] - ETA: 5s - loss: 0.0280 - acc: 0.9913 854/1875 [============>.................] - ETA: 5s - loss: 0.0280 - acc: 0.9913 865/1875 [============>.................] - ETA: 5s - loss: 0.0280 - acc: 0.9913 875/1875 [=============>................] - ETA: 5s - loss: 0.0280 - acc: 0.9914 886/1875 [=============>................] - ETA: 5s - loss: 0.0279 - acc: 0.9914 897/1875 [=============>................] - ETA: 5s - loss: 0.0279 - acc: 0.9914 907/1875 [=============>................] - ETA: 5s - loss: 0.0279 - acc: 0.9914 918/1875 [=============>................] - ETA: 5s - loss: 0.0279 - acc: 0.9914 929/1875 [=============>................] - ETA: 5s - loss: 0.0279 - acc: 0.9914 940/1875 [==============>...............] - ETA: 4s - loss: 0.0279 - acc: 0.9914 951/1875 [==============>...............] - ETA: 4s - loss: 0.0279 - acc: 0.9914 961/1875 [==============>...............] - ETA: 4s - loss: 0.0279 - acc: 0.9914 971/1875 [==============>...............] - ETA: 4s - loss: 0.0279 - acc: 0.9914 981/1875 [==============>...............] - ETA: 4s - loss: 0.0278 - acc: 0.9914 992/1875 [==============>...............] - ETA: 4s - loss: 0.0278 - acc: 0.9914 1002/1875 [===============>..............] - ETA: 4s - loss: 0.0278 - acc: 0.9914 1012/1875 [===============>..............] - ETA: 4s - loss: 0.0278 - acc: 0.9914 1023/1875 [===============>..............] - ETA: 4s - loss: 0.0278 - acc: 0.9914 1034/1875 [===============>..............] - ETA: 4s - loss: 0.0278 - acc: 0.9914 1044/1875 [===============>..............] - ETA: 4s - loss: 0.0278 - acc: 0.9914 1054/1875 [===============>..............] - ETA: 4s - loss: 0.0278 - acc: 0.9914 1064/1875 [================>.............] - ETA: 4s - loss: 0.0278 - acc: 0.9914 1074/1875 [================>.............] - ETA: 4s - loss: 0.0278 - acc: 0.9914 1085/1875 [================>.............] - ETA: 4s - loss: 0.0278 - acc: 0.9914 1096/1875 [================>.............] - ETA: 4s - loss: 0.0278 - acc: 0.9914 1107/1875 [================>.............] - ETA: 4s - loss: 0.0278 - acc: 0.9914 1117/1875 [================>.............] - ETA: 4s - loss: 0.0278 - acc: 0.9914 1127/1875 [=================>............] - ETA: 3s - loss: 0.0278 - acc: 0.9914 1137/1875 [=================>............] - ETA: 3s - loss: 0.0278 - acc: 0.9914 1147/1875 [=================>............] - ETA: 3s - loss: 0.0278 - acc: 0.9914 1158/1875 [=================>............] - ETA: 3s - loss: 0.0278 - acc: 0.9914 1169/1875 [=================>............] - ETA: 3s - loss: 0.0278 - acc: 0.9914 1180/1875 [=================>............] - ETA: 3s - loss: 0.0278 - acc: 0.9914 1189/1875 [==================>...........] - ETA: 3s - loss: 0.0278 - acc: 0.9914 1199/1875 [==================>...........] - ETA: 3s - loss: 0.0278 - acc: 0.9914 1209/1875 [==================>...........] - ETA: 3s - loss: 0.0277 - acc: 0.9914 1220/1875 [==================>...........] - ETA: 3s - loss: 0.0277 - acc: 0.9914 1231/1875 [==================>...........] - ETA: 3s - loss: 0.0277 - acc: 0.9914 1241/1875 [==================>...........] - ETA: 3s - loss: 0.0277 - acc: 0.9914 1251/1875 [===================>..........] - ETA: 3s - loss: 0.0277 - acc: 0.9914 1261/1875 [===================>..........] - ETA: 3s - loss: 0.0277 - acc: 0.9914 1270/1875 [===================>..........] - ETA: 3s - loss: 0.0277 - acc: 0.9914 1280/1875 [===================>..........] - ETA: 3s - loss: 0.0277 - acc: 0.9914 1291/1875 [===================>..........] - ETA: 3s - loss: 0.0277 - acc: 0.9914 1301/1875 [===================>..........] - ETA: 3s - loss: 0.0277 - acc: 0.9914 1311/1875 [===================>..........] - ETA: 2s - loss: 0.0277 - acc: 0.9914 1321/1875 [====================>.........] - ETA: 2s - loss: 0.0277 - acc: 0.9914 1332/1875 [====================>.........] - ETA: 2s - loss: 0.0277 - acc: 0.9914 1342/1875 [====================>.........] - ETA: 2s - loss: 0.0277 - acc: 0.9914 1352/1875 [====================>.........] - ETA: 2s - loss: 0.0277 - acc: 0.9914 1362/1875 [====================>.........] - ETA: 2s - loss: 0.0277 - acc: 0.9914 1373/1875 [====================>.........] - ETA: 2s - loss: 0.0277 - acc: 0.9914 1384/1875 [=====================>........] - ETA: 2s - loss: 0.0277 - acc: 0.9914 1394/1875 [=====================>........] - ETA: 2s - loss: 0.0277 - acc: 0.9914 1404/1875 [=====================>........] - ETA: 2s - loss: 0.0277 - acc: 0.9914 1413/1875 [=====================>........] - ETA: 2s - loss: 0.0276 - acc: 0.9914 1423/1875 [=====================>........] - ETA: 2s - loss: 0.0276 - acc: 0.9914 1433/1875 [=====================>........] - ETA: 2s - loss: 0.0276 - acc: 0.9914 1444/1875 [======================>.......] - ETA: 2s - loss: 0.0276 - acc: 0.9915 1455/1875 [======================>.......] - ETA: 2s - loss: 0.0276 - acc: 0.9915 1466/1875 [======================>.......] - ETA: 2s - loss: 0.0276 - acc: 0.9915 1476/1875 [======================>.......] - ETA: 2s - loss: 0.0276 - acc: 0.9915 1486/1875 [======================>.......] - ETA: 2s - loss: 0.0276 - acc: 0.9915 1497/1875 [======================>.......] - ETA: 1s - loss: 0.0276 - acc: 0.9915 1507/1875 [=======================>......] - ETA: 1s - loss: 0.0275 - acc: 0.9915 1517/1875 [=======================>......] - ETA: 1s - loss: 0.0275 - acc: 0.9915 1528/1875 [=======================>......] - ETA: 1s - loss: 0.0275 - acc: 0.9915 1539/1875 [=======================>......] - ETA: 1s - loss: 0.0275 - acc: 0.9915 1549/1875 [=======================>......] - ETA: 1s - loss: 0.0275 - acc: 0.9915 1559/1875 [=======================>......] - ETA: 1s - loss: 0.0275 - acc: 0.9915 1572/1875 [========================>.....] - ETA: 1s - loss: 0.0275 - acc: 0.9915 1587/1875 [========================>.....] - ETA: 1s - loss: 0.0274 - acc: 0.9915 1602/1875 [========================>.....] - ETA: 1s - loss: 0.0274 - acc: 0.9915 1617/1875 [========================>.....] - ETA: 1s - loss: 0.0274 - acc: 0.9915 1632/1875 [=========================>....] - ETA: 1s - loss: 0.0274 - acc: 0.9915 1646/1875 [=========================>....] - ETA: 1s - loss: 0.0273 - acc: 0.9915 1660/1875 [=========================>....] - ETA: 1s - loss: 0.0273 - acc: 0.9915 1675/1875 [=========================>....] - ETA: 1s - loss: 0.0273 - acc: 0.9916 1690/1875 [==========================>...] - ETA: 0s - loss: 0.0273 - acc: 0.9916 1704/1875 [==========================>...] - ETA: 0s - loss: 0.0273 - acc: 0.9916 1719/1875 [==========================>...] - ETA: 0s - loss: 0.0272 - acc: 0.9916 1734/1875 [==========================>...] - ETA: 0s - loss: 0.0272 - acc: 0.9916 1749/1875 [==========================>...] - ETA: 0s - loss: 0.0272 - acc: 0.9916 1764/1875 [===========================>..] - ETA: 0s - loss: 0.0272 - acc: 0.9916 1779/1875 [===========================>..] - ETA: 0s - loss: 0.0272 - acc: 0.9916 1795/1875 [===========================>..] - ETA: 0s - loss: 0.0271 - acc: 0.9916 1810/1875 [===========================>..] - ETA: 0s - loss: 0.0271 - acc: 0.9916 1825/1875 [============================>.] - ETA: 0s - loss: 0.0271 - acc: 0.9916 1840/1875 [============================>.] - ETA: 0s - loss: 0.0271 - acc: 0.9916 1854/1875 [============================>.] - ETA: 0s - loss: 0.0271 - acc: 0.9916 1868/1875 [============================>.] - ETA: 0s - loss: 0.0270 - acc: 0.9916 1875/1875 [==============================] - 10s 5ms/step - loss: 0.0270 - acc: 0.9916 Epoch 5/5 1/1875 [..............................] - ETA: 14:15 - loss: 0.1184 - acc: 0.9688 11/1875 [..............................] - ETA: 9s - loss: 0.0471 - acc: 0.9856 22/1875 [..............................] - ETA: 9s - loss: 0.0398 - acc: 0.9873 32/1875 [..............................] - ETA: 9s - loss: 0.0365 - acc: 0.9883 42/1875 [..............................] - ETA: 9s - loss: 0.0334 - acc: 0.9893 52/1875 [..............................] - ETA: 9s - loss: 0.0308 - acc: 0.9902 63/1875 [>.............................] - ETA: 9s - loss: 0.0292 - acc: 0.9908 73/1875 [>.............................] - ETA: 9s - loss: 0.0282 - acc: 0.9910 83/1875 [>.............................] - ETA: 9s - loss: 0.0278 - acc: 0.9910 93/1875 [>.............................] - ETA: 9s - loss: 0.0275 - acc: 0.9910 103/1875 [>.............................] - ETA: 9s - loss: 0.0272 - acc: 0.9910 113/1875 [>.............................] - ETA: 9s - loss: 0.0269 - acc: 0.9910 123/1875 [>.............................] - ETA: 8s - loss: 0.0266 - acc: 0.9910 134/1875 [=>............................] - ETA: 8s - loss: 0.0263 - acc: 0.9910 145/1875 [=>............................] - ETA: 8s - loss: 0.0262 - acc: 0.9910 156/1875 [=>............................] - ETA: 8s - loss: 0.0261 - acc: 0.9910 166/1875 [=>............................] - ETA: 8s - loss: 0.0261 - acc: 0.9909 176/1875 [=>............................] - ETA: 8s - loss: 0.0260 - acc: 0.9909 185/1875 [=>............................] - ETA: 8s - loss: 0.0259 - acc: 0.9910 195/1875 [==>...........................] - ETA: 8s - loss: 0.0257 - acc: 0.9910 205/1875 [==>...........................] - ETA: 8s - loss: 0.0256 - acc: 0.9910 215/1875 [==>...........................] - ETA: 8s - loss: 0.0255 - acc: 0.9910 225/1875 [==>...........................] - ETA: 8s - loss: 0.0253 - acc: 0.9911 234/1875 [==>...........................] - ETA: 8s - loss: 0.0253 - acc: 0.9911 245/1875 [==>...........................] - ETA: 8s - loss: 0.0252 - acc: 0.9911 255/1875 [===>..........................] - ETA: 8s - loss: 0.0251 - acc: 0.9911 265/1875 [===>..........................] - ETA: 8s - loss: 0.0251 - acc: 0.9911 276/1875 [===>..........................] - ETA: 8s - loss: 0.0250 - acc: 0.9911 287/1875 [===>..........................] - ETA: 8s - loss: 0.0249 - acc: 0.9911 298/1875 [===>..........................] - ETA: 8s - loss: 0.0248 - acc: 0.9912 308/1875 [===>..........................] - ETA: 8s - loss: 0.0248 - acc: 0.9912 318/1875 [====>.........................] - ETA: 7s - loss: 0.0247 - acc: 0.9912 328/1875 [====>.........................] - ETA: 7s - loss: 0.0246 - acc: 0.9913 338/1875 [====>.........................] - ETA: 7s - loss: 0.0245 - acc: 0.9913 349/1875 [====>.........................] - ETA: 7s - loss: 0.0244 - acc: 0.9913 360/1875 [====>.........................] - ETA: 7s - loss: 0.0243 - acc: 0.9913 371/1875 [====>.........................] - ETA: 7s - loss: 0.0243 - acc: 0.9913 382/1875 [=====>........................] - ETA: 7s - loss: 0.0242 - acc: 0.9914 391/1875 [=====>........................] - ETA: 7s - loss: 0.0241 - acc: 0.9914 401/1875 [=====>........................] - ETA: 7s - loss: 0.0241 - acc: 0.9914 411/1875 [=====>........................] - ETA: 7s - loss: 0.0240 - acc: 0.9914 422/1875 [=====>........................] - ETA: 7s - loss: 0.0240 - acc: 0.9914 433/1875 [=====>........................] - ETA: 7s - loss: 0.0239 - acc: 0.9914 444/1875 [======>.......................] - ETA: 7s - loss: 0.0238 - acc: 0.9914 455/1875 [======>.......................] - ETA: 7s - loss: 0.0238 - acc: 0.9915 464/1875 [======>.......................] - ETA: 7s - loss: 0.0238 - acc: 0.9915 474/1875 [======>.......................] - ETA: 7s - loss: 0.0237 - acc: 0.9915 484/1875 [======>.......................] - ETA: 7s - loss: 0.0237 - acc: 0.9915 494/1875 [======>.......................] - ETA: 7s - loss: 0.0237 - acc: 0.9915 505/1875 [=======>......................] - ETA: 7s - loss: 0.0237 - acc: 0.9915 515/1875 [=======>......................] - ETA: 6s - loss: 0.0237 - acc: 0.9915 526/1875 [=======>......................] - ETA: 6s - loss: 0.0237 - acc: 0.9915 536/1875 [=======>......................] - ETA: 6s - loss: 0.0238 - acc: 0.9915 547/1875 [=======>......................] - ETA: 6s - loss: 0.0238 - acc: 0.9915 558/1875 [=======>......................] - ETA: 6s - loss: 0.0238 - acc: 0.9915 569/1875 [========>.....................] - ETA: 6s - loss: 0.0238 - acc: 0.9915 580/1875 [========>.....................] - ETA: 6s - loss: 0.0237 - acc: 0.9915 591/1875 [========>.....................] - ETA: 6s - loss: 0.0237 - acc: 0.9915 601/1875 [========>.....................] - ETA: 6s - loss: 0.0237 - acc: 0.9915 611/1875 [========>.....................] - ETA: 6s - loss: 0.0238 - acc: 0.9915 621/1875 [========>.....................] - ETA: 6s - loss: 0.0238 - acc: 0.9915 631/1875 [=========>....................] - ETA: 6s - loss: 0.0237 - acc: 0.9915 642/1875 [=========>....................] - ETA: 6s - loss: 0.0237 - acc: 0.9915 653/1875 [=========>....................] - ETA: 6s - loss: 0.0237 - acc: 0.9915 664/1875 [=========>....................] - ETA: 6s - loss: 0.0237 - acc: 0.9915 674/1875 [=========>....................] - ETA: 6s - loss: 0.0237 - acc: 0.9916 685/1875 [=========>....................] - ETA: 6s - loss: 0.0237 - acc: 0.9916 695/1875 [==========>...................] - ETA: 6s - loss: 0.0237 - acc: 0.9916 705/1875 [==========>...................] - ETA: 5s - loss: 0.0237 - acc: 0.9916 715/1875 [==========>...................] - ETA: 5s - loss: 0.0236 - acc: 0.9916 726/1875 [==========>...................] - ETA: 5s - loss: 0.0236 - acc: 0.9916 737/1875 [==========>...................] - ETA: 5s - loss: 0.0236 - acc: 0.9916 748/1875 [==========>...................] - ETA: 5s - loss: 0.0236 - acc: 0.9916 758/1875 [===========>..................] - ETA: 5s - loss: 0.0236 - acc: 0.9916 768/1875 [===========>..................] - ETA: 5s - loss: 0.0236 - acc: 0.9916 778/1875 [===========>..................] - ETA: 5s - loss: 0.0236 - acc: 0.9916 788/1875 [===========>..................] - ETA: 5s - loss: 0.0236 - acc: 0.9916 798/1875 [===========>..................] - ETA: 5s - loss: 0.0235 - acc: 0.9916 809/1875 [===========>..................] - ETA: 5s - loss: 0.0235 - acc: 0.9917 820/1875 [============>.................] - ETA: 5s - loss: 0.0235 - acc: 0.9917 831/1875 [============>.................] - ETA: 5s - loss: 0.0235 - acc: 0.9917 842/1875 [============>.................] - ETA: 5s - loss: 0.0235 - acc: 0.9917 853/1875 [============>.................] - ETA: 5s - loss: 0.0235 - acc: 0.9917 864/1875 [============>.................] - ETA: 5s - loss: 0.0235 - acc: 0.9917 875/1875 [=============>................] - ETA: 5s - loss: 0.0235 - acc: 0.9917 886/1875 [=============>................] - ETA: 5s - loss: 0.0235 - acc: 0.9917 897/1875 [=============>................] - ETA: 4s - loss: 0.0235 - acc: 0.9917 907/1875 [=============>................] - ETA: 4s - loss: 0.0234 - acc: 0.9917 917/1875 [=============>................] - ETA: 4s - loss: 0.0234 - acc: 0.9917 928/1875 [=============>................] - ETA: 4s - loss: 0.0234 - acc: 0.9918 939/1875 [==============>...............] - ETA: 4s - loss: 0.0234 - acc: 0.9918 950/1875 [==============>...............] - ETA: 4s - loss: 0.0234 - acc: 0.9918 961/1875 [==============>...............] - ETA: 4s - loss: 0.0234 - acc: 0.9918 971/1875 [==============>...............] - ETA: 4s - loss: 0.0234 - acc: 0.9918 981/1875 [==============>...............] - ETA: 4s - loss: 0.0234 - acc: 0.9918 992/1875 [==============>...............] - ETA: 4s - loss: 0.0234 - acc: 0.9918 1003/1875 [===============>..............] - ETA: 4s - loss: 0.0234 - acc: 0.9918 1013/1875 [===============>..............] - ETA: 4s - loss: 0.0233 - acc: 0.9918 1022/1875 [===============>..............] - ETA: 4s - loss: 0.0233 - acc: 0.9918 1033/1875 [===============>..............] - ETA: 4s - loss: 0.0233 - acc: 0.9918 1044/1875 [===============>..............] - ETA: 4s - loss: 0.0233 - acc: 0.9918 1054/1875 [===============>..............] - ETA: 4s - loss: 0.0233 - acc: 0.9918 1064/1875 [================>.............] - ETA: 4s - loss: 0.0233 - acc: 0.9919 1074/1875 [================>.............] - ETA: 4s - loss: 0.0233 - acc: 0.9919 1085/1875 [================>.............] - ETA: 3s - loss: 0.0233 - acc: 0.9919 1096/1875 [================>.............] - ETA: 3s - loss: 0.0232 - acc: 0.9919 1106/1875 [================>.............] - ETA: 3s - loss: 0.0232 - acc: 0.9919 1117/1875 [================>.............] - ETA: 3s - loss: 0.0232 - acc: 0.9919 1127/1875 [=================>............] - ETA: 3s - loss: 0.0232 - acc: 0.9919 1138/1875 [=================>............] - ETA: 3s - loss: 0.0232 - acc: 0.9919 1148/1875 [=================>............] - ETA: 3s - loss: 0.0231 - acc: 0.9919 1159/1875 [=================>............] - ETA: 3s - loss: 0.0231 - acc: 0.9919 1169/1875 [=================>............] - ETA: 3s - loss: 0.0231 - acc: 0.9920 1180/1875 [=================>............] - ETA: 3s - loss: 0.0231 - acc: 0.9920 1190/1875 [==================>...........] - ETA: 3s - loss: 0.0231 - acc: 0.9920 1201/1875 [==================>...........] - ETA: 3s - loss: 0.0230 - acc: 0.9920 1211/1875 [==================>...........] - ETA: 3s - loss: 0.0230 - acc: 0.9920 1221/1875 [==================>...........] - ETA: 3s - loss: 0.0230 - acc: 0.9920 1230/1875 [==================>...........] - ETA: 3s - loss: 0.0230 - acc: 0.9920 1240/1875 [==================>...........] - ETA: 3s - loss: 0.0230 - acc: 0.9920 1251/1875 [===================>..........] - ETA: 3s - loss: 0.0229 - acc: 0.9920 1261/1875 [===================>..........] - ETA: 3s - loss: 0.0229 - acc: 0.9920 1272/1875 [===================>..........] - ETA: 3s - loss: 0.0229 - acc: 0.9921 1282/1875 [===================>..........] - ETA: 3s - loss: 0.0229 - acc: 0.9921 1293/1875 [===================>..........] - ETA: 2s - loss: 0.0229 - acc: 0.9921 1304/1875 [===================>..........] - ETA: 2s - loss: 0.0228 - acc: 0.9921 1314/1875 [====================>.........] - ETA: 2s - loss: 0.0228 - acc: 0.9921 1324/1875 [====================>.........] - ETA: 2s - loss: 0.0228 - acc: 0.9921 1334/1875 [====================>.........] - ETA: 2s - loss: 0.0228 - acc: 0.9921 1345/1875 [====================>.........] - ETA: 2s - loss: 0.0228 - acc: 0.9921 1356/1875 [====================>.........] - ETA: 2s - loss: 0.0227 - acc: 0.9921 1366/1875 [====================>.........] - ETA: 2s - loss: 0.0227 - acc: 0.9921 1377/1875 [=====================>........] - ETA: 2s - loss: 0.0227 - acc: 0.9922 1387/1875 [=====================>........] - ETA: 2s - loss: 0.0227 - acc: 0.9922 1398/1875 [=====================>........] - ETA: 2s - loss: 0.0227 - acc: 0.9922 1408/1875 [=====================>........] - ETA: 2s - loss: 0.0226 - acc: 0.9922 1419/1875 [=====================>........] - ETA: 2s - loss: 0.0226 - acc: 0.9922 1430/1875 [=====================>........] - ETA: 2s - loss: 0.0226 - acc: 0.9922 1441/1875 [======================>.......] - ETA: 2s - loss: 0.0226 - acc: 0.9922 1451/1875 [======================>.......] - ETA: 2s - loss: 0.0226 - acc: 0.9922 1461/1875 [======================>.......] - ETA: 2s - loss: 0.0225 - acc: 0.9922 1471/1875 [======================>.......] - ETA: 2s - loss: 0.0225 - acc: 0.9922 1482/1875 [======================>.......] - ETA: 1s - loss: 0.0225 - acc: 0.9922 1492/1875 [======================>.......] - ETA: 1s - loss: 0.0225 - acc: 0.9923 1503/1875 [=======================>......] - ETA: 1s - loss: 0.0225 - acc: 0.9923 1514/1875 [=======================>......] - ETA: 1s - loss: 0.0224 - acc: 0.9923 1524/1875 [=======================>......] - ETA: 1s - loss: 0.0224 - acc: 0.9923 1534/1875 [=======================>......] - ETA: 1s - loss: 0.0224 - acc: 0.9923 1544/1875 [=======================>......] - ETA: 1s - loss: 0.0224 - acc: 0.9923 1555/1875 [=======================>......] - ETA: 1s - loss: 0.0224 - acc: 0.9923 1566/1875 [========================>.....] - ETA: 1s - loss: 0.0224 - acc: 0.9923 1581/1875 [========================>.....] - ETA: 1s - loss: 0.0223 - acc: 0.9923 1596/1875 [========================>.....] - ETA: 1s - loss: 0.0223 - acc: 0.9923 1611/1875 [========================>.....] - ETA: 1s - loss: 0.0223 - acc: 0.9923 1627/1875 [=========================>....] - ETA: 1s - loss: 0.0223 - acc: 0.9924 1643/1875 [=========================>....] - ETA: 1s - loss: 0.0222 - acc: 0.9924 1659/1875 [=========================>....] - ETA: 1s - loss: 0.0222 - acc: 0.9924 1675/1875 [=========================>....] - ETA: 0s - loss: 0.0222 - acc: 0.9924 1690/1875 [==========================>...] - ETA: 0s - loss: 0.0222 - acc: 0.9924 1705/1875 [==========================>...] - ETA: 0s - loss: 0.0221 - acc: 0.9924 1720/1875 [==========================>...] - ETA: 0s - loss: 0.0221 - acc: 0.9924 1735/1875 [==========================>...] - ETA: 0s - loss: 0.0221 - acc: 0.9924 1751/1875 [===========================>..] - ETA: 0s - loss: 0.0221 - acc: 0.9924 1766/1875 [===========================>..] - ETA: 0s - loss: 0.0220 - acc: 0.9924 1781/1875 [===========================>..] - ETA: 0s - loss: 0.0220 - acc: 0.9925 1797/1875 [===========================>..] - ETA: 0s - loss: 0.0220 - acc: 0.9925 1812/1875 [===========================>..] - ETA: 0s - loss: 0.0220 - acc: 0.9925 1827/1875 [============================>.] - ETA: 0s - loss: 0.0220 - acc: 0.9925 1841/1875 [============================>.] - ETA: 0s - loss: 0.0219 - acc: 0.9925 1856/1875 [============================>.] - ETA: 0s - loss: 0.0219 - acc: 0.9925 1871/1875 [============================>.] - ETA: 0s - loss: 0.0219 - acc: 0.9925 1875/1875 [==============================] - 9s 5ms/step - loss: 0.0219 - acc: 0.9925 <tensorflow.python.keras.callbacks.History at 0x7fbf483ae940>
results = base_model.evaluate(test_dataset)
named_results = dict(zip(base_model.metrics_names, results))
print('\naccuracy:', named_results['acc'])
1/313 [..............................] - ETA: 1:05 - loss: 0.0516 - acc: 0.9688 19/313 [>.............................] - ETA: 0s - loss: 0.0415 - acc: 0.9885 38/313 [==>...........................] - ETA: 0s - loss: 0.0304 - acc: 0.9918 57/313 [====>.........................] - ETA: 0s - loss: 0.0300 - acc: 0.9923 75/313 [======>.......................] - ETA: 0s - loss: 0.0338 - acc: 0.9908 94/313 [========>.....................] - ETA: 0s - loss: 0.0317 - acc: 0.9910 112/313 [=========>....................] - ETA: 0s - loss: 0.0300 - acc: 0.9914 130/313 [===========>..................] - ETA: 0s - loss: 0.0320 - acc: 0.9904 148/313 [=============>................] - ETA: 0s - loss: 0.0327 - acc: 0.9903 167/313 [===============>..............] - ETA: 0s - loss: 0.0332 - acc: 0.9901 185/313 [================>.............] - ETA: 0s - loss: 0.0315 - acc: 0.9905 203/313 [==================>...........] - ETA: 0s - loss: 0.0305 - acc: 0.9908 220/313 [====================>.........] - ETA: 0s - loss: 0.0304 - acc: 0.9908 238/313 [=====================>........] - ETA: 0s - loss: 0.0323 - acc: 0.9904 256/313 [=======================>......] - ETA: 0s - loss: 0.0337 - acc: 0.9904 274/313 [=========================>....] - ETA: 0s - loss: 0.0343 - acc: 0.9903 293/313 [===========================>..] - ETA: 0s - loss: 0.0338 - acc: 0.9905 312/313 [============================>.] - ETA: 0s - loss: 0.0333 - acc: 0.9905 313/313 [==============================] - 1s 3ms/step - loss: 0.0333 - acc: 0.9905 accuracy: 0.9904999732971191
기본 모델이 테스트세트에서 99% 정확성을 달성하는 것을 볼 수 있습니다. 아래의 적대적 교란에서의 견고성에서 얼마나 견고한지 확인할 수 있습니다.
적대적 정규화 모델
여기에서는 NSL 프레임워크를 사용하여 몇 줄의 코드로 적대적 훈련을 Keras 모델에 통합하는 방법을 보여줍니다. 기본 모델은 훈련 목표에 적대적 정규화를 포함하는 새로운 tf.Keras.Model을 만들기 위해 래핑됩니다.
먼저, 도우미 함수 nsl.configs.make_adv_reg_config를 사용하여 모든 관련 하이퍼 매개변수가 포함된 config 객체를 만듭니다.
adv_config = nsl.configs.make_adv_reg_config(
multiplier=HPARAMS.adv_multiplier,
adv_step_size=HPARAMS.adv_step_size,
adv_grad_norm=HPARAMS.adv_grad_norm
)
이제 AdversarialRegularization으로 기본 모델을 래핑할 수 있습니다. 여기에서 새로운 기본 모델 (base_adv_model)을 생성하여 기존 모델(base_model)을 나중에 비교하는 데 사용할 수 있습니다.
반환된 adv_model은 tf.keras.Model 객체이며, 훈련 목표에는 적대적 손실에 대한 정규화 항이 포함됩니다. 손실을 계산하려면 모델이 일반 입력(특성 image) 외에도 레이블 정보(특성 label)에 액세스할 수 있어야 합니다. 이러한 이유로 데이터세트의 예를 튜플에서 다시 사전으로 변환합니다. 그리고 label_keys 매개변수를 통해 레이블 정보가 포함된 특성을 모델에 알립니다.
base_adv_model = build_base_model(HPARAMS)
adv_model = nsl.keras.AdversarialRegularization(
base_adv_model,
label_keys=[LABEL_INPUT_NAME],
adv_config=adv_config
)
train_set_for_adv_model = train_dataset.map(convert_to_dictionaries)
test_set_for_adv_model = test_dataset.map(convert_to_dictionaries)
다음으로 적대적 정규화 모델을 컴파일, 훈련 및 평가합니다. "손실 사전에서 누락된 출력"과 같은 경고가 있을 수 있으며, adv_model이 총 손실을 계산하기 위해 기본 구현에 의존하지 않기 때문에 괜찮습니다.
adv_model.compile(optimizer='adam', loss='sparse_categorical_crossentropy',
metrics=['acc'])
adv_model.fit(train_set_for_adv_model, epochs=HPARAMS.epochs)
Epoch 1/5 WARNING:tensorflow:The dtype of the watched tensor must be floating (e.g. tf.float32), got tf.int64 WARNING:tensorflow:The dtype of the watched tensor must be floating (e.g. tf.float32), got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:AutoGraph could not transform <bound method Socket.send of <zmq.sugar.socket.Socket object at 0x7fc0055961e8>> and will run it as-is. Please report this to the TensorFlow team. When filing the bug, set the verbosity to 10 (on Linux, `export AUTOGRAPH_VERBOSITY=10`) and attach the full output. Cause: <cyfunction Socket.send at 0x7fc00349c048> is not a module, class, method, function, traceback, frame, or code object To silence this warning, decorate the function with @tf.autograph.experimental.do_not_convert WARNING:absl:Cannot perturb features dict_keys(['label'])WARNING:tensorflow:AutoGraph could not transform <bound method Socket.send of <zmq.sugar.socket.Socket object at 0x7fc0055961e8>> and will run it as-is. Please report this to the TensorFlow team. When filing the bug, set the verbosity to 10 (on Linux, `export AUTOGRAPH_VERBOSITY=10`) and attach the full output. Cause: <cyfunction Socket.send at 0x7fc00349c048> is not a module, class, method, function, traceback, frame, or code object To silence this warning, decorate the function with @tf.autograph.experimental.do_not_convert WARNING: AutoGraph could not transform <bound method Socket.send of <zmq.sugar.socket.Socket object at 0x7fc0055961e8>> and will run it as-is. Please report this to the TensorFlow team. When filing the bug, set the verbosity to 10 (on Linux, `export AUTOGRAPH_VERBOSITY=10`) and attach the full output. Cause: <cyfunction Socket.send at 0x7fc00349c048> is not a module, class, method, function, traceback, frame, or code object To silence this warning, decorate the function with @tf.autograph.experimental.do_not_convert WARNING:tensorflow:The dtype of the watched tensor must be floating (e.g. tf.float32), got tf.int64 WARNING:tensorflow:The dtype of the watched tensor must be floating (e.g. tf.float32), got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 1/1875 [..............................] - ETA: 3:28:07 - loss: 2.7895 - sparse_categorical_crossentropy: 2.3104 - sparse_categorical_accuracy: 0.0312 - scaled_adversarial_loss: 0.4791 8/1875 [..............................] - ETA: 13s - loss: 2.7568 - sparse_categorical_crossentropy: 2.2808 - sparse_categorical_accuracy: 0.1597 - scaled_adversarial_loss: 0.4760 15/1875 [..............................] - ETA: 13s - loss: 2.7188 - sparse_categorical_crossentropy: 2.2455 - sparse_categorical_accuracy: 0.2348 - scaled_adversarial_loss: 0.4733 22/1875 [..............................] - ETA: 13s - loss: 2.6603 - sparse_categorical_crossentropy: 2.1894 - sparse_categorical_accuracy: 0.2883 - scaled_adversarial_loss: 0.4709 29/1875 [..............................] - ETA: 13s - loss: 2.5875 - sparse_categorical_crossentropy: 2.1182 - sparse_categorical_accuracy: 0.3284 - scaled_adversarial_loss: 0.4693 37/1875 [..............................] - ETA: 13s - loss: 2.5002 - sparse_categorical_crossentropy: 2.0314 - sparse_categorical_accuracy: 0.3657 - scaled_adversarial_loss: 0.4688 44/1875 [..............................] - ETA: 13s - loss: 2.4283 - sparse_categorical_crossentropy: 1.9596 - sparse_categorical_accuracy: 0.3921 - scaled_adversarial_loss: 0.4687 51/1875 [..............................] - ETA: 13s - loss: 2.3592 - sparse_categorical_crossentropy: 1.8913 - sparse_categorical_accuracy: 0.4159 - scaled_adversarial_loss: 0.4679 58/1875 [..............................] - ETA: 13s - loss: 2.2928 - sparse_categorical_crossentropy: 1.8268 - sparse_categorical_accuracy: 0.4381 - scaled_adversarial_loss: 0.4660 65/1875 [>.............................] - ETA: 13s - loss: 2.2301 - sparse_categorical_crossentropy: 1.7661 - sparse_categorical_accuracy: 0.4585 - scaled_adversarial_loss: 0.4640 72/1875 [>.............................] - ETA: 13s - loss: 2.1707 - sparse_categorical_crossentropy: 1.7092 - sparse_categorical_accuracy: 0.4775 - scaled_adversarial_loss: 0.4616 79/1875 [>.............................] - ETA: 13s - loss: 2.1157 - sparse_categorical_crossentropy: 1.6566 - sparse_categorical_accuracy: 0.4946 - scaled_adversarial_loss: 0.4591 86/1875 [>.............................] - ETA: 13s - loss: 2.0648 - sparse_categorical_crossentropy: 1.6081 - sparse_categorical_accuracy: 0.5103 - scaled_adversarial_loss: 0.4568 93/1875 [>.............................] - ETA: 13s - loss: 2.0171 - sparse_categorical_crossentropy: 1.5630 - sparse_categorical_accuracy: 0.5248 - scaled_adversarial_loss: 0.4541 100/1875 [>.............................] - ETA: 13s - loss: 1.9718 - sparse_categorical_crossentropy: 1.5206 - sparse_categorical_accuracy: 0.5384 - scaled_adversarial_loss: 0.4512 106/1875 [>.............................] - ETA: 13s - loss: 1.9353 - sparse_categorical_crossentropy: 1.4866 - sparse_categorical_accuracy: 0.5493 - scaled_adversarial_loss: 0.4487 112/1875 [>.............................] - ETA: 13s - loss: 1.9008 - sparse_categorical_crossentropy: 1.4545 - sparse_categorical_accuracy: 0.5595 - scaled_adversarial_loss: 0.4463 119/1875 [>.............................] - ETA: 13s - loss: 1.8628 - sparse_categorical_crossentropy: 1.4194 - sparse_categorical_accuracy: 0.5706 - scaled_adversarial_loss: 0.4434 126/1875 [=>............................] - ETA: 13s - loss: 1.8267 - sparse_categorical_crossentropy: 1.3862 - sparse_categorical_accuracy: 0.5810 - scaled_adversarial_loss: 0.4405 133/1875 [=>............................] - ETA: 13s - loss: 1.7925 - sparse_categorical_crossentropy: 1.3550 - sparse_categorical_accuracy: 0.5908 - scaled_adversarial_loss: 0.4375 140/1875 [=>............................] - ETA: 13s - loss: 1.7602 - sparse_categorical_crossentropy: 1.3255 - sparse_categorical_accuracy: 0.6000 - scaled_adversarial_loss: 0.4346 148/1875 [=>............................] - ETA: 13s - loss: 1.7251 - sparse_categorical_crossentropy: 1.2937 - sparse_categorical_accuracy: 0.6099 - scaled_adversarial_loss: 0.4314 155/1875 [=>............................] - ETA: 13s - loss: 1.6958 - sparse_categorical_crossentropy: 1.2673 - sparse_categorical_accuracy: 0.6182 - scaled_adversarial_loss: 0.4285 162/1875 [=>............................] - ETA: 13s - loss: 1.6680 - sparse_categorical_crossentropy: 1.2423 - sparse_categorical_accuracy: 0.6259 - scaled_adversarial_loss: 0.4257 169/1875 [=>............................] - ETA: 12s - loss: 1.6415 - sparse_categorical_crossentropy: 1.2186 - sparse_categorical_accuracy: 0.6333 - scaled_adversarial_loss: 0.4229 176/1875 [=>............................] - ETA: 12s - loss: 1.6164 - sparse_categorical_crossentropy: 1.1961 - sparse_categorical_accuracy: 0.6403 - scaled_adversarial_loss: 0.4203 183/1875 [=>............................] - ETA: 12s - loss: 1.5925 - sparse_categorical_crossentropy: 1.1748 - sparse_categorical_accuracy: 0.6469 - scaled_adversarial_loss: 0.4178 190/1875 [==>...........................] - ETA: 12s - loss: 1.5698 - sparse_categorical_crossentropy: 1.1545 - sparse_categorical_accuracy: 0.6532 - scaled_adversarial_loss: 0.4153 197/1875 [==>...........................] - ETA: 12s - loss: 1.5479 - sparse_categorical_crossentropy: 1.1351 - sparse_categorical_accuracy: 0.6592 - scaled_adversarial_loss: 0.4128 205/1875 [==>...........................] - ETA: 12s - loss: 1.5240 - sparse_categorical_crossentropy: 1.1140 - sparse_categorical_accuracy: 0.6658 - scaled_adversarial_loss: 0.4100 212/1875 [==>...........................] - ETA: 12s - loss: 1.5040 - sparse_categorical_crossentropy: 1.0964 - sparse_categorical_accuracy: 0.6712 - scaled_adversarial_loss: 0.4076 219/1875 [==>...........................] - ETA: 12s - loss: 1.4847 - sparse_categorical_crossentropy: 1.0794 - sparse_categorical_accuracy: 0.6764 - scaled_adversarial_loss: 0.4052 226/1875 [==>...........................] - ETA: 12s - loss: 1.4661 - sparse_categorical_crossentropy: 1.0632 - sparse_categorical_accuracy: 0.6814 - scaled_adversarial_loss: 0.4029 233/1875 [==>...........................] - ETA: 12s - loss: 1.4481 - sparse_categorical_crossentropy: 1.0475 - sparse_categorical_accuracy: 0.6862 - scaled_adversarial_loss: 0.4005 240/1875 [==>...........................] - ETA: 12s - loss: 1.4307 - sparse_categorical_crossentropy: 1.0324 - sparse_categorical_accuracy: 0.6909 - scaled_adversarial_loss: 0.3983 247/1875 [==>...........................] - ETA: 12s - loss: 1.4139 - sparse_categorical_crossentropy: 1.0179 - sparse_categorical_accuracy: 0.6953 - scaled_adversarial_loss: 0.3960 254/1875 [===>..........................] - ETA: 12s - loss: 1.3977 - sparse_categorical_crossentropy: 1.0039 - sparse_categorical_accuracy: 0.6996 - scaled_adversarial_loss: 0.3939 261/1875 [===>..........................] - ETA: 12s - loss: 1.3821 - sparse_categorical_crossentropy: 0.9903 - sparse_categorical_accuracy: 0.7038 - scaled_adversarial_loss: 0.3917 268/1875 [===>..........................] - ETA: 12s - loss: 1.3669 - sparse_categorical_crossentropy: 0.9773 - sparse_categorical_accuracy: 0.7078 - scaled_adversarial_loss: 0.3896 275/1875 [===>..........................] - ETA: 12s - loss: 1.3522 - sparse_categorical_crossentropy: 0.9647 - sparse_categorical_accuracy: 0.7116 - scaled_adversarial_loss: 0.3875 282/1875 [===>..........................] - ETA: 12s - loss: 1.3381 - sparse_categorical_crossentropy: 0.9526 - sparse_categorical_accuracy: 0.7154 - scaled_adversarial_loss: 0.3855 289/1875 [===>..........................] - ETA: 11s - loss: 1.3243 - sparse_categorical_crossentropy: 0.9408 - sparse_categorical_accuracy: 0.7190 - scaled_adversarial_loss: 0.3835 296/1875 [===>..........................] - ETA: 11s - loss: 1.3110 - sparse_categorical_crossentropy: 0.9294 - sparse_categorical_accuracy: 0.7225 - scaled_adversarial_loss: 0.3815 303/1875 [===>..........................] - ETA: 11s - loss: 1.2980 - sparse_categorical_crossentropy: 0.9184 - sparse_categorical_accuracy: 0.7259 - scaled_adversarial_loss: 0.3796 310/1875 [===>..........................] - ETA: 11s - loss: 1.2854 - sparse_categorical_crossentropy: 0.9077 - sparse_categorical_accuracy: 0.7291 - scaled_adversarial_loss: 0.3777 317/1875 [====>.........................] - ETA: 11s - loss: 1.2732 - sparse_categorical_crossentropy: 0.8974 - sparse_categorical_accuracy: 0.7323 - scaled_adversarial_loss: 0.3759 324/1875 [====>.........................] - ETA: 11s - loss: 1.2613 - sparse_categorical_crossentropy: 0.8873 - sparse_categorical_accuracy: 0.7354 - scaled_adversarial_loss: 0.3740 332/1875 [====>.........................] - ETA: 11s - loss: 1.2482 - sparse_categorical_crossentropy: 0.8762 - sparse_categorical_accuracy: 0.7388 - scaled_adversarial_loss: 0.3720 339/1875 [====>.........................] - ETA: 11s - loss: 1.2370 - sparse_categorical_crossentropy: 0.8668 - sparse_categorical_accuracy: 0.7417 - scaled_adversarial_loss: 0.3702 346/1875 [====>.........................] - ETA: 11s - loss: 1.2260 - sparse_categorical_crossentropy: 0.8576 - sparse_categorical_accuracy: 0.7445 - scaled_adversarial_loss: 0.3685 353/1875 [====>.........................] - ETA: 11s - loss: 1.2154 - sparse_categorical_crossentropy: 0.8486 - sparse_categorical_accuracy: 0.7472 - scaled_adversarial_loss: 0.3668 360/1875 [====>.........................] - ETA: 11s - loss: 1.2050 - sparse_categorical_crossentropy: 0.8399 - sparse_categorical_accuracy: 0.7499 - scaled_adversarial_loss: 0.3651 367/1875 [====>.........................] - ETA: 11s - loss: 1.1949 - sparse_categorical_crossentropy: 0.8315 - sparse_categorical_accuracy: 0.7525 - scaled_adversarial_loss: 0.3634 374/1875 [====>.........................] - ETA: 11s - loss: 1.1850 - sparse_categorical_crossentropy: 0.8232 - sparse_categorical_accuracy: 0.7550 - scaled_adversarial_loss: 0.3618 381/1875 [=====>........................] - ETA: 11s - loss: 1.1753 - sparse_categorical_crossentropy: 0.8152 - sparse_categorical_accuracy: 0.7574 - scaled_adversarial_loss: 0.3602 388/1875 [=====>........................] - ETA: 11s - loss: 1.1659 - sparse_categorical_crossentropy: 0.8073 - sparse_categorical_accuracy: 0.7598 - scaled_adversarial_loss: 0.3586 395/1875 [=====>........................] - ETA: 11s - loss: 1.1567 - sparse_categorical_crossentropy: 0.7997 - sparse_categorical_accuracy: 0.7621 - scaled_adversarial_loss: 0.3570 402/1875 [=====>........................] - ETA: 11s - loss: 1.1478 - sparse_categorical_crossentropy: 0.7922 - sparse_categorical_accuracy: 0.7644 - scaled_adversarial_loss: 0.3555 409/1875 [=====>........................] - ETA: 11s - loss: 1.1390 - sparse_categorical_crossentropy: 0.7850 - sparse_categorical_accuracy: 0.7666 - scaled_adversarial_loss: 0.3540 416/1875 [=====>........................] - ETA: 10s - loss: 1.1304 - sparse_categorical_crossentropy: 0.7779 - sparse_categorical_accuracy: 0.7687 - scaled_adversarial_loss: 0.3525 423/1875 [=====>........................] - ETA: 10s - loss: 1.1221 - sparse_categorical_crossentropy: 0.7710 - sparse_categorical_accuracy: 0.7708 - scaled_adversarial_loss: 0.3511 430/1875 [=====>........................] - ETA: 10s - loss: 1.1139 - sparse_categorical_crossentropy: 0.7642 - sparse_categorical_accuracy: 0.7729 - scaled_adversarial_loss: 0.3497 437/1875 [=====>........................] - ETA: 10s - loss: 1.1059 - sparse_categorical_crossentropy: 0.7576 - sparse_categorical_accuracy: 0.7749 - scaled_adversarial_loss: 0.3483 444/1875 [======>.......................] - ETA: 10s - loss: 1.0981 - sparse_categorical_crossentropy: 0.7512 - sparse_categorical_accuracy: 0.7768 - scaled_adversarial_loss: 0.3469 451/1875 [======>.......................] - ETA: 10s - loss: 1.0904 - sparse_categorical_crossentropy: 0.7449 - sparse_categorical_accuracy: 0.7787 - scaled_adversarial_loss: 0.3455 458/1875 [======>.......................] - ETA: 10s - loss: 1.0829 - sparse_categorical_crossentropy: 0.7387 - sparse_categorical_accuracy: 0.7806 - scaled_adversarial_loss: 0.3442 465/1875 [======>.......................] - ETA: 10s - loss: 1.0755 - sparse_categorical_crossentropy: 0.7327 - sparse_categorical_accuracy: 0.7825 - scaled_adversarial_loss: 0.3428 472/1875 [======>.......................] - ETA: 10s - loss: 1.0683 - sparse_categorical_crossentropy: 0.7268 - sparse_categorical_accuracy: 0.7842 - scaled_adversarial_loss: 0.3415 479/1875 [======>.......................] - ETA: 10s - loss: 1.0612 - sparse_categorical_crossentropy: 0.7210 - sparse_categorical_accuracy: 0.7860 - scaled_adversarial_loss: 0.3402 486/1875 [======>.......................] - ETA: 10s - loss: 1.0542 - sparse_categorical_crossentropy: 0.7153 - sparse_categorical_accuracy: 0.7877 - scaled_adversarial_loss: 0.3390 493/1875 [======>.......................] - ETA: 10s - loss: 1.0474 - sparse_categorical_crossentropy: 0.7097 - sparse_categorical_accuracy: 0.7894 - scaled_adversarial_loss: 0.3377 500/1875 [=======>......................] - ETA: 10s - loss: 1.0407 - sparse_categorical_crossentropy: 0.7043 - sparse_categorical_accuracy: 0.7911 - scaled_adversarial_loss: 0.3365 506/1875 [=======>......................] - ETA: 10s - loss: 1.0351 - sparse_categorical_crossentropy: 0.6997 - sparse_categorical_accuracy: 0.7924 - scaled_adversarial_loss: 0.3354 512/1875 [=======>......................] - ETA: 10s - loss: 1.0295 - sparse_categorical_crossentropy: 0.6952 - sparse_categorical_accuracy: 0.7938 - scaled_adversarial_loss: 0.3344 518/1875 [=======>......................] - ETA: 10s - loss: 1.0240 - sparse_categorical_crossentropy: 0.6907 - sparse_categorical_accuracy: 0.7951 - scaled_adversarial_loss: 0.3333 525/1875 [=======>......................] - ETA: 10s - loss: 1.0178 - sparse_categorical_crossentropy: 0.6856 - sparse_categorical_accuracy: 0.7967 - scaled_adversarial_loss: 0.3322 532/1875 [=======>......................] - ETA: 10s - loss: 1.0116 - sparse_categorical_crossentropy: 0.6806 - sparse_categorical_accuracy: 0.7982 - scaled_adversarial_loss: 0.3310 539/1875 [=======>......................] - ETA: 10s - loss: 1.0056 - sparse_categorical_crossentropy: 0.6757 - sparse_categorical_accuracy: 0.7997 - scaled_adversarial_loss: 0.3298 546/1875 [=======>......................] - ETA: 10s - loss: 0.9996 - sparse_categorical_crossentropy: 0.6709 - sparse_categorical_accuracy: 0.8011 - scaled_adversarial_loss: 0.3287 553/1875 [=======>......................] - ETA: 9s - loss: 0.9938 - sparse_categorical_crossentropy: 0.6662 - sparse_categorical_accuracy: 0.8025 - scaled_adversarial_loss: 0.3276 560/1875 [=======>......................] - ETA: 9s - loss: 0.9881 - sparse_categorical_crossentropy: 0.6616 - sparse_categorical_accuracy: 0.8039 - scaled_adversarial_loss: 0.3265 567/1875 [========>.....................] - ETA: 9s - loss: 0.9824 - sparse_categorical_crossentropy: 0.6570 - sparse_categorical_accuracy: 0.8053 - scaled_adversarial_loss: 0.3254 574/1875 [========>.....................] - ETA: 9s - loss: 0.9769 - sparse_categorical_crossentropy: 0.6526 - sparse_categorical_accuracy: 0.8066 - scaled_adversarial_loss: 0.3243 581/1875 [========>.....................] - ETA: 9s - loss: 0.9715 - sparse_categorical_crossentropy: 0.6482 - sparse_categorical_accuracy: 0.8079 - scaled_adversarial_loss: 0.3232 588/1875 [========>.....................] - ETA: 9s - loss: 0.9661 - sparse_categorical_crossentropy: 0.6439 - sparse_categorical_accuracy: 0.8092 - scaled_adversarial_loss: 0.3222 595/1875 [========>.....................] - ETA: 9s - loss: 0.9609 - sparse_categorical_crossentropy: 0.6397 - sparse_categorical_accuracy: 0.8105 - scaled_adversarial_loss: 0.3212 602/1875 [========>.....................] - ETA: 9s - loss: 0.9557 - sparse_categorical_crossentropy: 0.6355 - sparse_categorical_accuracy: 0.8118 - scaled_adversarial_loss: 0.3201 609/1875 [========>.....................] - ETA: 9s - loss: 0.9506 - sparse_categorical_crossentropy: 0.6314 - sparse_categorical_accuracy: 0.8130 - scaled_adversarial_loss: 0.3191 616/1875 [========>.....................] - ETA: 9s - loss: 0.9456 - sparse_categorical_crossentropy: 0.6274 - sparse_categorical_accuracy: 0.8142 - scaled_adversarial_loss: 0.3181 623/1875 [========>.....................] - ETA: 9s - loss: 0.9406 - sparse_categorical_crossentropy: 0.6235 - sparse_categorical_accuracy: 0.8154 - scaled_adversarial_loss: 0.3172 630/1875 [=========>....................] - ETA: 9s - loss: 0.9358 - sparse_categorical_crossentropy: 0.6196 - sparse_categorical_accuracy: 0.8166 - scaled_adversarial_loss: 0.3162 638/1875 [=========>....................] - ETA: 9s - loss: 0.9303 - sparse_categorical_crossentropy: 0.6152 - sparse_categorical_accuracy: 0.8179 - scaled_adversarial_loss: 0.3151 645/1875 [=========>....................] - ETA: 9s - loss: 0.9256 - sparse_categorical_crossentropy: 0.6115 - sparse_categorical_accuracy: 0.8190 - scaled_adversarial_loss: 0.3141 652/1875 [=========>....................] - ETA: 9s - loss: 0.9210 - sparse_categorical_crossentropy: 0.6078 - sparse_categorical_accuracy: 0.8201 - scaled_adversarial_loss: 0.3132 659/1875 [=========>....................] - ETA: 9s - loss: 0.9165 - sparse_categorical_crossentropy: 0.6042 - sparse_categorical_accuracy: 0.8212 - scaled_adversarial_loss: 0.3123 667/1875 [=========>....................] - ETA: 9s - loss: 0.9113 - sparse_categorical_crossentropy: 0.6001 - sparse_categorical_accuracy: 0.8224 - scaled_adversarial_loss: 0.3112 674/1875 [=========>....................] - ETA: 9s - loss: 0.9069 - sparse_categorical_crossentropy: 0.5966 - sparse_categorical_accuracy: 0.8235 - scaled_adversarial_loss: 0.3103 680/1875 [=========>....................] - ETA: 9s - loss: 0.9032 - sparse_categorical_crossentropy: 0.5936 - sparse_categorical_accuracy: 0.8243 - scaled_adversarial_loss: 0.3096 687/1875 [=========>....................] - ETA: 8s - loss: 0.8989 - sparse_categorical_crossentropy: 0.5902 - sparse_categorical_accuracy: 0.8254 - scaled_adversarial_loss: 0.3087 694/1875 [==========>...................] - ETA: 8s - loss: 0.8947 - sparse_categorical_crossentropy: 0.5868 - sparse_categorical_accuracy: 0.8264 - scaled_adversarial_loss: 0.3078 701/1875 [==========>...................] - ETA: 8s - loss: 0.8905 - sparse_categorical_crossentropy: 0.5835 - sparse_categorical_accuracy: 0.8274 - scaled_adversarial_loss: 0.3069 708/1875 [==========>...................] - ETA: 8s - loss: 0.8864 - sparse_categorical_crossentropy: 0.5803 - sparse_categorical_accuracy: 0.8283 - scaled_adversarial_loss: 0.3061 716/1875 [==========>...................] - ETA: 8s - loss: 0.8817 - sparse_categorical_crossentropy: 0.5766 - sparse_categorical_accuracy: 0.8294 - scaled_adversarial_loss: 0.3051 723/1875 [==========>...................] - ETA: 8s - loss: 0.8777 - sparse_categorical_crossentropy: 0.5734 - sparse_categorical_accuracy: 0.8304 - scaled_adversarial_loss: 0.3043 730/1875 [==========>...................] - ETA: 8s - loss: 0.8738 - sparse_categorical_crossentropy: 0.5703 - sparse_categorical_accuracy: 0.8313 - scaled_adversarial_loss: 0.3035 737/1875 [==========>...................] - ETA: 8s - loss: 0.8699 - sparse_categorical_crossentropy: 0.5673 - sparse_categorical_accuracy: 0.8322 - scaled_adversarial_loss: 0.3026 744/1875 [==========>...................] - ETA: 8s - loss: 0.8661 - sparse_categorical_crossentropy: 0.5642 - sparse_categorical_accuracy: 0.8331 - scaled_adversarial_loss: 0.3018 752/1875 [===========>..................] - ETA: 8s - loss: 0.8617 - sparse_categorical_crossentropy: 0.5608 - sparse_categorical_accuracy: 0.8341 - scaled_adversarial_loss: 0.3009 760/1875 [===========>..................] - ETA: 8s - loss: 0.8575 - sparse_categorical_crossentropy: 0.5574 - sparse_categorical_accuracy: 0.8351 - scaled_adversarial_loss: 0.3000 767/1875 [===========>..................] - ETA: 8s - loss: 0.8538 - sparse_categorical_crossentropy: 0.5546 - sparse_categorical_accuracy: 0.8360 - scaled_adversarial_loss: 0.2992 774/1875 [===========>..................] - ETA: 8s - loss: 0.8502 - sparse_categorical_crossentropy: 0.5517 - sparse_categorical_accuracy: 0.8369 - scaled_adversarial_loss: 0.2985 781/1875 [===========>..................] - ETA: 8s - loss: 0.8466 - sparse_categorical_crossentropy: 0.5489 - sparse_categorical_accuracy: 0.8377 - scaled_adversarial_loss: 0.2977 788/1875 [===========>..................] - ETA: 8s - loss: 0.8431 - sparse_categorical_crossentropy: 0.5461 - sparse_categorical_accuracy: 0.8385 - scaled_adversarial_loss: 0.2970 795/1875 [===========>..................] - ETA: 8s - loss: 0.8396 - sparse_categorical_crossentropy: 0.5434 - sparse_categorical_accuracy: 0.8393 - scaled_adversarial_loss: 0.2962 802/1875 [===========>..................] - ETA: 8s - loss: 0.8362 - sparse_categorical_crossentropy: 0.5407 - sparse_categorical_accuracy: 0.8401 - scaled_adversarial_loss: 0.2955 809/1875 [===========>..................] - ETA: 8s - loss: 0.8328 - sparse_categorical_crossentropy: 0.5380 - sparse_categorical_accuracy: 0.8409 - scaled_adversarial_loss: 0.2947 817/1875 [============>.................] - ETA: 7s - loss: 0.8289 - sparse_categorical_crossentropy: 0.5350 - sparse_categorical_accuracy: 0.8418 - scaled_adversarial_loss: 0.2939 824/1875 [============>.................] - ETA: 7s - loss: 0.8256 - sparse_categorical_crossentropy: 0.5324 - sparse_categorical_accuracy: 0.8426 - scaled_adversarial_loss: 0.2932 831/1875 [============>.................] - ETA: 7s - loss: 0.8224 - sparse_categorical_crossentropy: 0.5299 - sparse_categorical_accuracy: 0.8434 - scaled_adversarial_loss: 0.2925 838/1875 [============>.................] - ETA: 7s - loss: 0.8192 - sparse_categorical_crossentropy: 0.5274 - sparse_categorical_accuracy: 0.8441 - scaled_adversarial_loss: 0.2918 845/1875 [============>.................] - ETA: 7s - loss: 0.8160 - sparse_categorical_crossentropy: 0.5249 - sparse_categorical_accuracy: 0.8448 - scaled_adversarial_loss: 0.2911 852/1875 [============>.................] - ETA: 7s - loss: 0.8128 - sparse_categorical_crossentropy: 0.5224 - sparse_categorical_accuracy: 0.8456 - scaled_adversarial_loss: 0.2904 859/1875 [============>.................] - ETA: 7s - loss: 0.8097 - sparse_categorical_crossentropy: 0.5200 - sparse_categorical_accuracy: 0.8463 - scaled_adversarial_loss: 0.2897 866/1875 [============>.................] - ETA: 7s - loss: 0.8066 - sparse_categorical_crossentropy: 0.5176 - sparse_categorical_accuracy: 0.8470 - scaled_adversarial_loss: 0.2890 873/1875 [============>.................] - ETA: 7s - loss: 0.8036 - sparse_categorical_crossentropy: 0.5152 - sparse_categorical_accuracy: 0.8477 - scaled_adversarial_loss: 0.2884 880/1875 [=============>................] - ETA: 7s - loss: 0.8006 - sparse_categorical_crossentropy: 0.5129 - sparse_categorical_accuracy: 0.8484 - scaled_adversarial_loss: 0.2877 887/1875 [=============>................] - ETA: 7s - loss: 0.7976 - sparse_categorical_crossentropy: 0.5106 - sparse_categorical_accuracy: 0.8491 - scaled_adversarial_loss: 0.2870 894/1875 [=============>................] - ETA: 7s - loss: 0.7946 - sparse_categorical_crossentropy: 0.5083 - sparse_categorical_accuracy: 0.8498 - scaled_adversarial_loss: 0.2864 901/1875 [=============>................] - ETA: 7s - loss: 0.7917 - sparse_categorical_crossentropy: 0.5060 - sparse_categorical_accuracy: 0.8505 - scaled_adversarial_loss: 0.2857 908/1875 [=============>................] - ETA: 7s - loss: 0.7888 - sparse_categorical_crossentropy: 0.5038 - sparse_categorical_accuracy: 0.8511 - scaled_adversarial_loss: 0.2851 915/1875 [=============>................] - ETA: 7s - loss: 0.7860 - sparse_categorical_crossentropy: 0.5016 - sparse_categorical_accuracy: 0.8518 - scaled_adversarial_loss: 0.2844 922/1875 [=============>................] - ETA: 7s - loss: 0.7832 - sparse_categorical_crossentropy: 0.4994 - sparse_categorical_accuracy: 0.8524 - scaled_adversarial_loss: 0.2838 929/1875 [=============>................] - ETA: 7s - loss: 0.7804 - sparse_categorical_crossentropy: 0.4972 - sparse_categorical_accuracy: 0.8531 - scaled_adversarial_loss: 0.2832 936/1875 [=============>................] - ETA: 7s - loss: 0.7776 - sparse_categorical_crossentropy: 0.4951 - sparse_categorical_accuracy: 0.8537 - scaled_adversarial_loss: 0.2826 943/1875 [==============>...............] - ETA: 7s - loss: 0.7749 - sparse_categorical_crossentropy: 0.4930 - sparse_categorical_accuracy: 0.8543 - scaled_adversarial_loss: 0.2820 951/1875 [==============>...............] - ETA: 6s - loss: 0.7718 - sparse_categorical_crossentropy: 0.4906 - sparse_categorical_accuracy: 0.8550 - scaled_adversarial_loss: 0.2813 958/1875 [==============>...............] - ETA: 6s - loss: 0.7692 - sparse_categorical_crossentropy: 0.4885 - sparse_categorical_accuracy: 0.8557 - scaled_adversarial_loss: 0.2806 965/1875 [==============>...............] - ETA: 6s - loss: 0.7665 - sparse_categorical_crossentropy: 0.4865 - sparse_categorical_accuracy: 0.8563 - scaled_adversarial_loss: 0.2800 972/1875 [==============>...............] - ETA: 6s - loss: 0.7639 - sparse_categorical_crossentropy: 0.4845 - sparse_categorical_accuracy: 0.8569 - scaled_adversarial_loss: 0.2794 979/1875 [==============>...............] - ETA: 6s - loss: 0.7613 - sparse_categorical_crossentropy: 0.4825 - sparse_categorical_accuracy: 0.8575 - scaled_adversarial_loss: 0.2789 986/1875 [==============>...............] - ETA: 6s - loss: 0.7588 - sparse_categorical_crossentropy: 0.4805 - sparse_categorical_accuracy: 0.8580 - scaled_adversarial_loss: 0.2783 992/1875 [==============>...............] - ETA: 6s - loss: 0.7566 - sparse_categorical_crossentropy: 0.4788 - sparse_categorical_accuracy: 0.8585 - scaled_adversarial_loss: 0.2778 999/1875 [==============>...............] - ETA: 6s - loss: 0.7541 - sparse_categorical_crossentropy: 0.4769 - sparse_categorical_accuracy: 0.8591 - scaled_adversarial_loss: 0.2772 1006/1875 [===============>..............] - ETA: 6s - loss: 0.7516 - sparse_categorical_crossentropy: 0.4750 - sparse_categorical_accuracy: 0.8597 - scaled_adversarial_loss: 0.2766 1013/1875 [===============>..............] - ETA: 6s - loss: 0.7492 - sparse_categorical_crossentropy: 0.4731 - sparse_categorical_accuracy: 0.8602 - scaled_adversarial_loss: 0.2760 1020/1875 [===============>..............] - ETA: 6s - loss: 0.7467 - sparse_categorical_crossentropy: 0.4712 - sparse_categorical_accuracy: 0.8608 - scaled_adversarial_loss: 0.2755 1027/1875 [===============>..............] - ETA: 6s - loss: 0.7443 - sparse_categorical_crossentropy: 0.4694 - sparse_categorical_accuracy: 0.8613 - scaled_adversarial_loss: 0.2749 1034/1875 [===============>..............] - ETA: 6s - loss: 0.7419 - sparse_categorical_crossentropy: 0.4676 - sparse_categorical_accuracy: 0.8619 - scaled_adversarial_loss: 0.2743 1041/1875 [===============>..............] - ETA: 6s - loss: 0.7395 - sparse_categorical_crossentropy: 0.4657 - sparse_categorical_accuracy: 0.8624 - scaled_adversarial_loss: 0.2738 1048/1875 [===============>..............] - ETA: 6s - loss: 0.7372 - sparse_categorical_crossentropy: 0.4639 - sparse_categorical_accuracy: 0.8630 - scaled_adversarial_loss: 0.2732 1055/1875 [===============>..............] - ETA: 6s - loss: 0.7349 - sparse_categorical_crossentropy: 0.4622 - sparse_categorical_accuracy: 0.8635 - scaled_adversarial_loss: 0.2727 1062/1875 [===============>..............] - ETA: 6s - loss: 0.7326 - sparse_categorical_crossentropy: 0.4604 - sparse_categorical_accuracy: 0.8640 - scaled_adversarial_loss: 0.2721 1069/1875 [================>.............] - ETA: 6s - loss: 0.7303 - sparse_categorical_crossentropy: 0.4587 - sparse_categorical_accuracy: 0.8645 - scaled_adversarial_loss: 0.2716 1076/1875 [================>.............] - ETA: 6s - loss: 0.7280 - sparse_categorical_crossentropy: 0.4570 - sparse_categorical_accuracy: 0.8650 - scaled_adversarial_loss: 0.2711 1083/1875 [================>.............] - ETA: 5s - loss: 0.7258 - sparse_categorical_crossentropy: 0.4552 - sparse_categorical_accuracy: 0.8655 - scaled_adversarial_loss: 0.2705 1090/1875 [================>.............] - ETA: 5s - loss: 0.7236 - sparse_categorical_crossentropy: 0.4536 - sparse_categorical_accuracy: 0.8661 - scaled_adversarial_loss: 0.2700 1098/1875 [================>.............] - ETA: 5s - loss: 0.7210 - sparse_categorical_crossentropy: 0.4516 - sparse_categorical_accuracy: 0.8666 - scaled_adversarial_loss: 0.2694 1105/1875 [================>.............] - ETA: 5s - loss: 0.7189 - sparse_categorical_crossentropy: 0.4500 - sparse_categorical_accuracy: 0.8671 - scaled_adversarial_loss: 0.2689 1111/1875 [================>.............] - ETA: 5s - loss: 0.7170 - sparse_categorical_crossentropy: 0.4486 - sparse_categorical_accuracy: 0.8675 - scaled_adversarial_loss: 0.2684 1118/1875 [================>.............] - ETA: 5s - loss: 0.7149 - sparse_categorical_crossentropy: 0.4469 - sparse_categorical_accuracy: 0.8680 - scaled_adversarial_loss: 0.2679 1125/1875 [=================>............] - ETA: 5s - loss: 0.7127 - sparse_categorical_crossentropy: 0.4453 - sparse_categorical_accuracy: 0.8685 - scaled_adversarial_loss: 0.2674 1132/1875 [=================>............] - ETA: 5s - loss: 0.7106 - sparse_categorical_crossentropy: 0.4437 - sparse_categorical_accuracy: 0.8690 - scaled_adversarial_loss: 0.2669 1139/1875 [=================>............] - ETA: 5s - loss: 0.7086 - sparse_categorical_crossentropy: 0.4421 - sparse_categorical_accuracy: 0.8694 - scaled_adversarial_loss: 0.2664 1146/1875 [=================>............] - ETA: 5s - loss: 0.7065 - sparse_categorical_crossentropy: 0.4406 - sparse_categorical_accuracy: 0.8699 - scaled_adversarial_loss: 0.2659 1153/1875 [=================>............] - ETA: 5s - loss: 0.7044 - sparse_categorical_crossentropy: 0.4390 - sparse_categorical_accuracy: 0.8704 - scaled_adversarial_loss: 0.2654 1160/1875 [=================>............] - ETA: 5s - loss: 0.7024 - sparse_categorical_crossentropy: 0.4375 - sparse_categorical_accuracy: 0.8708 - scaled_adversarial_loss: 0.2649 1167/1875 [=================>............] - ETA: 5s - loss: 0.7004 - sparse_categorical_crossentropy: 0.4359 - sparse_categorical_accuracy: 0.8713 - scaled_adversarial_loss: 0.2644 1174/1875 [=================>............] - ETA: 5s - loss: 0.6984 - sparse_categorical_crossentropy: 0.4344 - sparse_categorical_accuracy: 0.8717 - scaled_adversarial_loss: 0.2640 1181/1875 [=================>............] - ETA: 5s - loss: 0.6964 - sparse_categorical_crossentropy: 0.4329 - sparse_categorical_accuracy: 0.8722 - scaled_adversarial_loss: 0.2635 1188/1875 [==================>...........] - ETA: 5s - loss: 0.6945 - sparse_categorical_crossentropy: 0.4315 - sparse_categorical_accuracy: 0.8726 - scaled_adversarial_loss: 0.2630 1195/1875 [==================>...........] - ETA: 5s - loss: 0.6925 - sparse_categorical_crossentropy: 0.4300 - sparse_categorical_accuracy: 0.8730 - scaled_adversarial_loss: 0.2625 1202/1875 [==================>...........] - ETA: 5s - loss: 0.6906 - sparse_categorical_crossentropy: 0.4285 - sparse_categorical_accuracy: 0.8735 - scaled_adversarial_loss: 0.2621 1209/1875 [==================>...........] - ETA: 5s - loss: 0.6887 - sparse_categorical_crossentropy: 0.4271 - sparse_categorical_accuracy: 0.8739 - scaled_adversarial_loss: 0.2616 1216/1875 [==================>...........] - ETA: 4s - loss: 0.6868 - sparse_categorical_crossentropy: 0.4257 - sparse_categorical_accuracy: 0.8743 - scaled_adversarial_loss: 0.2611 1222/1875 [==================>...........] - ETA: 4s - loss: 0.6852 - sparse_categorical_crossentropy: 0.4245 - sparse_categorical_accuracy: 0.8747 - scaled_adversarial_loss: 0.2607 1229/1875 [==================>...........] - ETA: 4s - loss: 0.6833 - sparse_categorical_crossentropy: 0.4231 - sparse_categorical_accuracy: 0.8751 - scaled_adversarial_loss: 0.2603 1237/1875 [==================>...........] - ETA: 4s - loss: 0.6812 - sparse_categorical_crossentropy: 0.4215 - sparse_categorical_accuracy: 0.8756 - scaled_adversarial_loss: 0.2598 1244/1875 [==================>...........] - ETA: 4s - loss: 0.6794 - sparse_categorical_crossentropy: 0.4201 - sparse_categorical_accuracy: 0.8760 - scaled_adversarial_loss: 0.2593 1251/1875 [===================>..........] - ETA: 4s - loss: 0.6776 - sparse_categorical_crossentropy: 0.4187 - sparse_categorical_accuracy: 0.8764 - scaled_adversarial_loss: 0.2589 1258/1875 [===================>..........] - ETA: 4s - loss: 0.6758 - sparse_categorical_crossentropy: 0.4174 - sparse_categorical_accuracy: 0.8768 - scaled_adversarial_loss: 0.2584 1265/1875 [===================>..........] - ETA: 4s - loss: 0.6740 - sparse_categorical_crossentropy: 0.4160 - sparse_categorical_accuracy: 0.8772 - scaled_adversarial_loss: 0.2580 1272/1875 [===================>..........] - ETA: 4s - loss: 0.6723 - sparse_categorical_crossentropy: 0.4147 - sparse_categorical_accuracy: 0.8776 - scaled_adversarial_loss: 0.2576 1278/1875 [===================>..........] - ETA: 4s - loss: 0.6708 - sparse_categorical_crossentropy: 0.4136 - sparse_categorical_accuracy: 0.8779 - scaled_adversarial_loss: 0.2572 1285/1875 [===================>..........] - ETA: 4s - loss: 0.6690 - sparse_categorical_crossentropy: 0.4123 - sparse_categorical_accuracy: 0.8783 - scaled_adversarial_loss: 0.2568 1292/1875 [===================>..........] - ETA: 4s - loss: 0.6673 - sparse_categorical_crossentropy: 0.4110 - sparse_categorical_accuracy: 0.8787 - scaled_adversarial_loss: 0.2563 1299/1875 [===================>..........] - ETA: 4s - loss: 0.6656 - sparse_categorical_crossentropy: 0.4097 - sparse_categorical_accuracy: 0.8791 - scaled_adversarial_loss: 0.2559 1306/1875 [===================>..........] - ETA: 4s - loss: 0.6639 - sparse_categorical_crossentropy: 0.4084 - sparse_categorical_accuracy: 0.8794 - scaled_adversarial_loss: 0.2555 1313/1875 [====================>.........] - ETA: 4s - loss: 0.6622 - sparse_categorical_crossentropy: 0.4071 - sparse_categorical_accuracy: 0.8798 - scaled_adversarial_loss: 0.2551 1320/1875 [====================>.........] - ETA: 4s - loss: 0.6605 - sparse_categorical_crossentropy: 0.4059 - sparse_categorical_accuracy: 0.8802 - scaled_adversarial_loss: 0.2546 1327/1875 [====================>.........] - ETA: 4s - loss: 0.6589 - sparse_categorical_crossentropy: 0.4046 - sparse_categorical_accuracy: 0.8806 - scaled_adversarial_loss: 0.2542 1334/1875 [====================>.........] - ETA: 4s - loss: 0.6572 - sparse_categorical_crossentropy: 0.4034 - sparse_categorical_accuracy: 0.8809 - scaled_adversarial_loss: 0.2538 1341/1875 [====================>.........] - ETA: 4s - loss: 0.6556 - sparse_categorical_crossentropy: 0.4022 - sparse_categorical_accuracy: 0.8813 - scaled_adversarial_loss: 0.2534 1348/1875 [====================>.........] - ETA: 3s - loss: 0.6540 - sparse_categorical_crossentropy: 0.4010 - sparse_categorical_accuracy: 0.8816 - scaled_adversarial_loss: 0.2530 1355/1875 [====================>.........] - ETA: 3s - loss: 0.6524 - sparse_categorical_crossentropy: 0.3998 - sparse_categorical_accuracy: 0.8820 - scaled_adversarial_loss: 0.2526 1362/1875 [====================>.........] - ETA: 3s - loss: 0.6508 - sparse_categorical_crossentropy: 0.3986 - sparse_categorical_accuracy: 0.8823 - scaled_adversarial_loss: 0.2522 1369/1875 [====================>.........] - ETA: 3s - loss: 0.6492 - sparse_categorical_crossentropy: 0.3974 - sparse_categorical_accuracy: 0.8827 - scaled_adversarial_loss: 0.2518 1376/1875 [=====================>........] - ETA: 3s - loss: 0.6476 - sparse_categorical_crossentropy: 0.3962 - sparse_categorical_accuracy: 0.8830 - scaled_adversarial_loss: 0.2514 1383/1875 [=====================>........] - ETA: 3s - loss: 0.6460 - sparse_categorical_crossentropy: 0.3950 - sparse_categorical_accuracy: 0.8834 - scaled_adversarial_loss: 0.2510 1390/1875 [=====================>........] - ETA: 3s - loss: 0.6445 - sparse_categorical_crossentropy: 0.3939 - sparse_categorical_accuracy: 0.8837 - scaled_adversarial_loss: 0.2506 1397/1875 [=====================>........] - ETA: 3s - loss: 0.6430 - sparse_categorical_crossentropy: 0.3927 - sparse_categorical_accuracy: 0.8841 - scaled_adversarial_loss: 0.2502 1404/1875 [=====================>........] - ETA: 3s - loss: 0.6414 - sparse_categorical_crossentropy: 0.3916 - sparse_categorical_accuracy: 0.8844 - scaled_adversarial_loss: 0.2498 1411/1875 [=====================>........] - ETA: 3s - loss: 0.6399 - sparse_categorical_crossentropy: 0.3905 - sparse_categorical_accuracy: 0.8847 - scaled_adversarial_loss: 0.2495 1418/1875 [=====================>........] - ETA: 3s - loss: 0.6384 - sparse_categorical_crossentropy: 0.3893 - sparse_categorical_accuracy: 0.8851 - scaled_adversarial_loss: 0.2491 1425/1875 [=====================>........] - ETA: 3s - loss: 0.6369 - sparse_categorical_crossentropy: 0.3882 - sparse_categorical_accuracy: 0.8854 - scaled_adversarial_loss: 0.2487 1432/1875 [=====================>........] - ETA: 3s - loss: 0.6354 - sparse_categorical_crossentropy: 0.3871 - sparse_categorical_accuracy: 0.8857 - scaled_adversarial_loss: 0.2483 1439/1875 [======================>.......] - ETA: 3s - loss: 0.6340 - sparse_categorical_crossentropy: 0.3860 - sparse_categorical_accuracy: 0.8860 - scaled_adversarial_loss: 0.2479 1446/1875 [======================>.......] - ETA: 3s - loss: 0.6325 - sparse_categorical_crossentropy: 0.3849 - sparse_categorical_accuracy: 0.8864 - scaled_adversarial_loss: 0.2476 1453/1875 [======================>.......] - ETA: 3s - loss: 0.6310 - sparse_categorical_crossentropy: 0.3838 - sparse_categorical_accuracy: 0.8867 - scaled_adversarial_loss: 0.2472 1460/1875 [======================>.......] - ETA: 3s - loss: 0.6296 - sparse_categorical_crossentropy: 0.3828 - sparse_categorical_accuracy: 0.8870 - scaled_adversarial_loss: 0.2468 1467/1875 [======================>.......] - ETA: 3s - loss: 0.6282 - sparse_categorical_crossentropy: 0.3817 - sparse_categorical_accuracy: 0.8873 - scaled_adversarial_loss: 0.2465 1474/1875 [======================>.......] - ETA: 3s - loss: 0.6267 - sparse_categorical_crossentropy: 0.3806 - sparse_categorical_accuracy: 0.8876 - scaled_adversarial_loss: 0.2461 1480/1875 [======================>.......] - ETA: 2s - loss: 0.6255 - sparse_categorical_crossentropy: 0.3797 - sparse_categorical_accuracy: 0.8879 - scaled_adversarial_loss: 0.2458 1487/1875 [======================>.......] - ETA: 2s - loss: 0.6241 - sparse_categorical_crossentropy: 0.3787 - sparse_categorical_accuracy: 0.8882 - scaled_adversarial_loss: 0.2454 1494/1875 [======================>.......] - ETA: 2s - loss: 0.6227 - sparse_categorical_crossentropy: 0.3777 - sparse_categorical_accuracy: 0.8885 - scaled_adversarial_loss: 0.2451 1501/1875 [=======================>......] - ETA: 2s - loss: 0.6213 - sparse_categorical_crossentropy: 0.3766 - sparse_categorical_accuracy: 0.8888 - scaled_adversarial_loss: 0.2447 1508/1875 [=======================>......] - ETA: 2s - loss: 0.6200 - sparse_categorical_crossentropy: 0.3756 - sparse_categorical_accuracy: 0.8891 - scaled_adversarial_loss: 0.2443 1515/1875 [=======================>......] - ETA: 2s - loss: 0.6186 - sparse_categorical_crossentropy: 0.3746 - sparse_categorical_accuracy: 0.8894 - scaled_adversarial_loss: 0.2440 1522/1875 [=======================>......] - ETA: 2s - loss: 0.6172 - sparse_categorical_crossentropy: 0.3736 - sparse_categorical_accuracy: 0.8897 - scaled_adversarial_loss: 0.2436 1529/1875 [=======================>......] - ETA: 2s - loss: 0.6159 - sparse_categorical_crossentropy: 0.3726 - sparse_categorical_accuracy: 0.8900 - scaled_adversarial_loss: 0.2433 1536/1875 [=======================>......] - ETA: 2s - loss: 0.6145 - sparse_categorical_crossentropy: 0.3716 - sparse_categorical_accuracy: 0.8903 - scaled_adversarial_loss: 0.2429 1543/1875 [=======================>......] - ETA: 2s - loss: 0.6132 - sparse_categorical_crossentropy: 0.3706 - sparse_categorical_accuracy: 0.8906 - scaled_adversarial_loss: 0.2426 1550/1875 [=======================>......] - ETA: 2s - loss: 0.6119 - sparse_categorical_crossentropy: 0.3696 - sparse_categorical_accuracy: 0.8909 - scaled_adversarial_loss: 0.2422 1557/1875 [=======================>......] - ETA: 2s - loss: 0.6106 - sparse_categorical_crossentropy: 0.3687 - sparse_categorical_accuracy: 0.8912 - scaled_adversarial_loss: 0.2419 1563/1875 [========================>.....] - ETA: 2s - loss: 0.6095 - sparse_categorical_crossentropy: 0.3678 - sparse_categorical_accuracy: 0.8914 - scaled_adversarial_loss: 0.2416 1572/1875 [========================>.....] - ETA: 2s - loss: 0.6078 - sparse_categorical_crossentropy: 0.3666 - sparse_categorical_accuracy: 0.8918 - scaled_adversarial_loss: 0.2412 1582/1875 [========================>.....] - ETA: 2s - loss: 0.6060 - sparse_categorical_crossentropy: 0.3653 - sparse_categorical_accuracy: 0.8922 - scaled_adversarial_loss: 0.2407 1591/1875 [========================>.....] - ETA: 2s - loss: 0.6043 - sparse_categorical_crossentropy: 0.3641 - sparse_categorical_accuracy: 0.8925 - scaled_adversarial_loss: 0.2403 1600/1875 [========================>.....] - ETA: 2s - loss: 0.6027 - sparse_categorical_crossentropy: 0.3629 - sparse_categorical_accuracy: 0.8929 - scaled_adversarial_loss: 0.2399 1609/1875 [========================>.....] - ETA: 2s - loss: 0.6011 - sparse_categorical_crossentropy: 0.3617 - sparse_categorical_accuracy: 0.8932 - scaled_adversarial_loss: 0.2394 1618/1875 [========================>.....] - ETA: 1s - loss: 0.5995 - sparse_categorical_crossentropy: 0.3605 - sparse_categorical_accuracy: 0.8936 - scaled_adversarial_loss: 0.2390 1627/1875 [=========================>....] - ETA: 1s - loss: 0.5980 - sparse_categorical_crossentropy: 0.3594 - sparse_categorical_accuracy: 0.8939 - scaled_adversarial_loss: 0.2386 1636/1875 [=========================>....] - ETA: 1s - loss: 0.5964 - sparse_categorical_crossentropy: 0.3582 - sparse_categorical_accuracy: 0.8942 - scaled_adversarial_loss: 0.2382 1645/1875 [=========================>....] - ETA: 1s - loss: 0.5948 - sparse_categorical_crossentropy: 0.3571 - sparse_categorical_accuracy: 0.8946 - scaled_adversarial_loss: 0.2378 1654/1875 [=========================>....] - ETA: 1s - loss: 0.5933 - sparse_categorical_crossentropy: 0.3559 - sparse_categorical_accuracy: 0.8949 - scaled_adversarial_loss: 0.2374 1663/1875 [=========================>....] - ETA: 1s - loss: 0.5918 - sparse_categorical_crossentropy: 0.3548 - sparse_categorical_accuracy: 0.8952 - scaled_adversarial_loss: 0.2370 1672/1875 [=========================>....] - ETA: 1s - loss: 0.5903 - sparse_categorical_crossentropy: 0.3537 - sparse_categorical_accuracy: 0.8956 - scaled_adversarial_loss: 0.2366 1681/1875 [=========================>....] - ETA: 1s - loss: 0.5888 - sparse_categorical_crossentropy: 0.3526 - sparse_categorical_accuracy: 0.8959 - scaled_adversarial_loss: 0.2362 1690/1875 [==========================>...] - ETA: 1s - loss: 0.5873 - sparse_categorical_crossentropy: 0.3515 - sparse_categorical_accuracy: 0.8962 - scaled_adversarial_loss: 0.2358 1699/1875 [==========================>...] - ETA: 1s - loss: 0.5858 - sparse_categorical_crossentropy: 0.3504 - sparse_categorical_accuracy: 0.8965 - scaled_adversarial_loss: 0.2354 1708/1875 [==========================>...] - ETA: 1s - loss: 0.5843 - sparse_categorical_crossentropy: 0.3493 - sparse_categorical_accuracy: 0.8968 - scaled_adversarial_loss: 0.2350 1717/1875 [==========================>...] - ETA: 1s - loss: 0.5829 - sparse_categorical_crossentropy: 0.3483 - sparse_categorical_accuracy: 0.8972 - scaled_adversarial_loss: 0.2346 1726/1875 [==========================>...] - ETA: 1s - loss: 0.5814 - sparse_categorical_crossentropy: 0.3472 - sparse_categorical_accuracy: 0.8975 - scaled_adversarial_loss: 0.2342 1736/1875 [==========================>...] - ETA: 1s - loss: 0.5798 - sparse_categorical_crossentropy: 0.3460 - sparse_categorical_accuracy: 0.8978 - scaled_adversarial_loss: 0.2338 1745/1875 [==========================>...] - ETA: 0s - loss: 0.5784 - sparse_categorical_crossentropy: 0.3450 - sparse_categorical_accuracy: 0.8981 - scaled_adversarial_loss: 0.2334 1754/1875 [===========================>..] - ETA: 0s - loss: 0.5770 - sparse_categorical_crossentropy: 0.3440 - sparse_categorical_accuracy: 0.8984 - scaled_adversarial_loss: 0.2330 1763/1875 [===========================>..] - ETA: 0s - loss: 0.5756 - sparse_categorical_crossentropy: 0.3429 - sparse_categorical_accuracy: 0.8987 - scaled_adversarial_loss: 0.2327 1772/1875 [===========================>..] - ETA: 0s - loss: 0.5742 - sparse_categorical_crossentropy: 0.3419 - sparse_categorical_accuracy: 0.8990 - scaled_adversarial_loss: 0.2323 1781/1875 [===========================>..] - ETA: 0s - loss: 0.5728 - sparse_categorical_crossentropy: 0.3409 - sparse_categorical_accuracy: 0.8993 - scaled_adversarial_loss: 0.2319 1790/1875 [===========================>..] - ETA: 0s - loss: 0.5715 - sparse_categorical_crossentropy: 0.3399 - sparse_categorical_accuracy: 0.8996 - scaled_adversarial_loss: 0.2316 1798/1875 [===========================>..] - ETA: 0s - loss: 0.5702 - sparse_categorical_crossentropy: 0.3390 - sparse_categorical_accuracy: 0.8999 - scaled_adversarial_loss: 0.2312 1807/1875 [===========================>..] - ETA: 0s - loss: 0.5689 - sparse_categorical_crossentropy: 0.3380 - sparse_categorical_accuracy: 0.9002 - scaled_adversarial_loss: 0.2309 1816/1875 [============================>.] - ETA: 0s - loss: 0.5675 - sparse_categorical_crossentropy: 0.3371 - sparse_categorical_accuracy: 0.9004 - scaled_adversarial_loss: 0.2305 1825/1875 [============================>.] - ETA: 0s - loss: 0.5662 - sparse_categorical_crossentropy: 0.3361 - sparse_categorical_accuracy: 0.9007 - scaled_adversarial_loss: 0.2301 1834/1875 [============================>.] - ETA: 0s - loss: 0.5649 - sparse_categorical_crossentropy: 0.3351 - sparse_categorical_accuracy: 0.9010 - scaled_adversarial_loss: 0.2298 1843/1875 [============================>.] - ETA: 0s - loss: 0.5636 - sparse_categorical_crossentropy: 0.3342 - sparse_categorical_accuracy: 0.9013 - scaled_adversarial_loss: 0.2294 1852/1875 [============================>.] - ETA: 0s - loss: 0.5623 - sparse_categorical_crossentropy: 0.3332 - sparse_categorical_accuracy: 0.9016 - scaled_adversarial_loss: 0.2291 1861/1875 [============================>.] - ETA: 0s - loss: 0.5610 - sparse_categorical_crossentropy: 0.3323 - sparse_categorical_accuracy: 0.9018 - scaled_adversarial_loss: 0.2287 1870/1875 [============================>.] - ETA: 0s - loss: 0.5597 - sparse_categorical_crossentropy: 0.3313 - sparse_categorical_accuracy: 0.9021 - scaled_adversarial_loss: 0.2284 1875/1875 [==============================] - 20s 7ms/step - loss: 0.5588 - sparse_categorical_crossentropy: 0.3307 - sparse_categorical_accuracy: 0.9023 - scaled_adversarial_loss: 0.2281 Epoch 2/5 1/1875 [..............................] - ETA: 13:11 - loss: 0.0708 - sparse_categorical_crossentropy: 0.0169 - sparse_categorical_accuracy: 1.0000 - scaled_adversarial_loss: 0.0539 9/1875 [..............................] - ETA: 13s - loss: 0.1883 - sparse_categorical_crossentropy: 0.0798 - sparse_categorical_accuracy: 0.9868 - scaled_adversarial_loss: 0.1085 16/1875 [..............................] - ETA: 13s - loss: 0.1774 - sparse_categorical_crossentropy: 0.0728 - sparse_categorical_accuracy: 0.9881 - scaled_adversarial_loss: 0.1045 23/1875 [..............................] - ETA: 13s - loss: 0.1722 - sparse_categorical_crossentropy: 0.0687 - sparse_categorical_accuracy: 0.9876 - scaled_adversarial_loss: 0.1034 30/1875 [..............................] - ETA: 13s - loss: 0.1669 - sparse_categorical_crossentropy: 0.0648 - sparse_categorical_accuracy: 0.9876 - scaled_adversarial_loss: 0.1020 37/1875 [..............................] - ETA: 13s - loss: 0.1636 - sparse_categorical_crossentropy: 0.0626 - sparse_categorical_accuracy: 0.9872 - scaled_adversarial_loss: 0.1010 44/1875 [..............................] - ETA: 13s - loss: 0.1618 - sparse_categorical_crossentropy: 0.0612 - sparse_categorical_accuracy: 0.9868 - scaled_adversarial_loss: 0.1006 51/1875 [..............................] - ETA: 13s - loss: 0.1603 - sparse_categorical_crossentropy: 0.0599 - sparse_categorical_accuracy: 0.9867 - scaled_adversarial_loss: 0.1003 58/1875 [..............................] - ETA: 13s - loss: 0.1590 - sparse_categorical_crossentropy: 0.0588 - sparse_categorical_accuracy: 0.9865 - scaled_adversarial_loss: 0.1002 65/1875 [>.............................] - ETA: 13s - loss: 0.1579 - sparse_categorical_crossentropy: 0.0579 - sparse_categorical_accuracy: 0.9864 - scaled_adversarial_loss: 0.1001 72/1875 [>.............................] - ETA: 13s - loss: 0.1575 - sparse_categorical_crossentropy: 0.0575 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.1000 79/1875 [>.............................] - ETA: 13s - loss: 0.1576 - sparse_categorical_crossentropy: 0.0575 - sparse_categorical_accuracy: 0.9858 - scaled_adversarial_loss: 0.1001 86/1875 [>.............................] - ETA: 13s - loss: 0.1578 - sparse_categorical_crossentropy: 0.0576 - sparse_categorical_accuracy: 0.9855 - scaled_adversarial_loss: 0.1002 93/1875 [>.............................] - ETA: 13s - loss: 0.1578 - sparse_categorical_crossentropy: 0.0577 - sparse_categorical_accuracy: 0.9853 - scaled_adversarial_loss: 0.1002 100/1875 [>.............................] - ETA: 13s - loss: 0.1579 - sparse_categorical_crossentropy: 0.0577 - sparse_categorical_accuracy: 0.9850 - scaled_adversarial_loss: 0.1002 107/1875 [>.............................] - ETA: 13s - loss: 0.1578 - sparse_categorical_crossentropy: 0.0576 - sparse_categorical_accuracy: 0.9849 - scaled_adversarial_loss: 0.1002 114/1875 [>.............................] - ETA: 13s - loss: 0.1576 - sparse_categorical_crossentropy: 0.0575 - sparse_categorical_accuracy: 0.9847 - scaled_adversarial_loss: 0.1001 121/1875 [>.............................] - ETA: 13s - loss: 0.1574 - sparse_categorical_crossentropy: 0.0573 - sparse_categorical_accuracy: 0.9847 - scaled_adversarial_loss: 0.1001 128/1875 [=>............................] - ETA: 13s - loss: 0.1571 - sparse_categorical_crossentropy: 0.0570 - sparse_categorical_accuracy: 0.9846 - scaled_adversarial_loss: 0.1001 135/1875 [=>............................] - ETA: 13s - loss: 0.1569 - sparse_categorical_crossentropy: 0.0569 - sparse_categorical_accuracy: 0.9845 - scaled_adversarial_loss: 0.1001 142/1875 [=>............................] - ETA: 13s - loss: 0.1567 - sparse_categorical_crossentropy: 0.0567 - sparse_categorical_accuracy: 0.9845 - scaled_adversarial_loss: 0.1000 150/1875 [=>............................] - ETA: 12s - loss: 0.1563 - sparse_categorical_crossentropy: 0.0564 - sparse_categorical_accuracy: 0.9845 - scaled_adversarial_loss: 0.0999 158/1875 [=>............................] - ETA: 12s - loss: 0.1561 - sparse_categorical_crossentropy: 0.0562 - sparse_categorical_accuracy: 0.9845 - scaled_adversarial_loss: 0.0999 165/1875 [=>............................] - ETA: 12s - loss: 0.1558 - sparse_categorical_crossentropy: 0.0559 - sparse_categorical_accuracy: 0.9845 - scaled_adversarial_loss: 0.0998 172/1875 [=>............................] - ETA: 12s - loss: 0.1554 - sparse_categorical_crossentropy: 0.0557 - sparse_categorical_accuracy: 0.9845 - scaled_adversarial_loss: 0.0997 179/1875 [=>............................] - ETA: 12s - loss: 0.1550 - sparse_categorical_crossentropy: 0.0554 - sparse_categorical_accuracy: 0.9845 - scaled_adversarial_loss: 0.0996 187/1875 [=>............................] - ETA: 12s - loss: 0.1547 - sparse_categorical_crossentropy: 0.0552 - sparse_categorical_accuracy: 0.9845 - scaled_adversarial_loss: 0.0995 195/1875 [==>...........................] - ETA: 12s - loss: 0.1544 - sparse_categorical_crossentropy: 0.0550 - sparse_categorical_accuracy: 0.9845 - scaled_adversarial_loss: 0.0994 202/1875 [==>...........................] - ETA: 12s - loss: 0.1541 - sparse_categorical_crossentropy: 0.0548 - sparse_categorical_accuracy: 0.9845 - scaled_adversarial_loss: 0.0993 209/1875 [==>...........................] - ETA: 12s - loss: 0.1539 - sparse_categorical_crossentropy: 0.0546 - sparse_categorical_accuracy: 0.9845 - scaled_adversarial_loss: 0.0992 216/1875 [==>...........................] - ETA: 12s - loss: 0.1536 - sparse_categorical_crossentropy: 0.0544 - sparse_categorical_accuracy: 0.9845 - scaled_adversarial_loss: 0.0992 224/1875 [==>...........................] - ETA: 12s - loss: 0.1533 - sparse_categorical_crossentropy: 0.0542 - sparse_categorical_accuracy: 0.9846 - scaled_adversarial_loss: 0.0991 231/1875 [==>...........................] - ETA: 12s - loss: 0.1530 - sparse_categorical_crossentropy: 0.0540 - sparse_categorical_accuracy: 0.9846 - scaled_adversarial_loss: 0.0990 238/1875 [==>...........................] - ETA: 12s - loss: 0.1527 - sparse_categorical_crossentropy: 0.0538 - sparse_categorical_accuracy: 0.9846 - scaled_adversarial_loss: 0.0989 246/1875 [==>...........................] - ETA: 12s - loss: 0.1524 - sparse_categorical_crossentropy: 0.0536 - sparse_categorical_accuracy: 0.9846 - scaled_adversarial_loss: 0.0988 253/1875 [===>..........................] - ETA: 12s - loss: 0.1522 - sparse_categorical_crossentropy: 0.0534 - sparse_categorical_accuracy: 0.9846 - scaled_adversarial_loss: 0.0988 260/1875 [===>..........................] - ETA: 12s - loss: 0.1520 - sparse_categorical_crossentropy: 0.0532 - sparse_categorical_accuracy: 0.9846 - scaled_adversarial_loss: 0.0987 267/1875 [===>..........................] - ETA: 11s - loss: 0.1517 - sparse_categorical_crossentropy: 0.0531 - sparse_categorical_accuracy: 0.9847 - scaled_adversarial_loss: 0.0987 274/1875 [===>..........................] - ETA: 11s - loss: 0.1515 - sparse_categorical_crossentropy: 0.0529 - sparse_categorical_accuracy: 0.9847 - scaled_adversarial_loss: 0.0986 281/1875 [===>..........................] - ETA: 11s - loss: 0.1513 - sparse_categorical_crossentropy: 0.0528 - sparse_categorical_accuracy: 0.9847 - scaled_adversarial_loss: 0.0986 288/1875 [===>..........................] - ETA: 11s - loss: 0.1511 - sparse_categorical_crossentropy: 0.0526 - sparse_categorical_accuracy: 0.9847 - scaled_adversarial_loss: 0.0985 295/1875 [===>..........................] - ETA: 11s - loss: 0.1509 - sparse_categorical_crossentropy: 0.0524 - sparse_categorical_accuracy: 0.9847 - scaled_adversarial_loss: 0.0985 302/1875 [===>..........................] - ETA: 11s - loss: 0.1507 - sparse_categorical_crossentropy: 0.0523 - sparse_categorical_accuracy: 0.9847 - scaled_adversarial_loss: 0.0984 309/1875 [===>..........................] - ETA: 11s - loss: 0.1505 - sparse_categorical_crossentropy: 0.0521 - sparse_categorical_accuracy: 0.9847 - scaled_adversarial_loss: 0.0983 316/1875 [====>.........................] - ETA: 11s - loss: 0.1502 - sparse_categorical_crossentropy: 0.0520 - sparse_categorical_accuracy: 0.9847 - scaled_adversarial_loss: 0.0983 323/1875 [====>.........................] - ETA: 11s - loss: 0.1500 - sparse_categorical_crossentropy: 0.0518 - sparse_categorical_accuracy: 0.9847 - scaled_adversarial_loss: 0.0982 330/1875 [====>.........................] - ETA: 11s - loss: 0.1498 - sparse_categorical_crossentropy: 0.0517 - sparse_categorical_accuracy: 0.9847 - scaled_adversarial_loss: 0.0981 337/1875 [====>.........................] - ETA: 11s - loss: 0.1495 - sparse_categorical_crossentropy: 0.0515 - sparse_categorical_accuracy: 0.9847 - scaled_adversarial_loss: 0.0980 344/1875 [====>.........................] - ETA: 11s - loss: 0.1493 - sparse_categorical_crossentropy: 0.0514 - sparse_categorical_accuracy: 0.9848 - scaled_adversarial_loss: 0.0979 351/1875 [====>.........................] - ETA: 11s - loss: 0.1490 - sparse_categorical_crossentropy: 0.0512 - sparse_categorical_accuracy: 0.9848 - scaled_adversarial_loss: 0.0978 358/1875 [====>.........................] - ETA: 11s - loss: 0.1488 - sparse_categorical_crossentropy: 0.0511 - sparse_categorical_accuracy: 0.9848 - scaled_adversarial_loss: 0.0977 365/1875 [====>.........................] - ETA: 11s - loss: 0.1486 - sparse_categorical_crossentropy: 0.0509 - sparse_categorical_accuracy: 0.9848 - scaled_adversarial_loss: 0.0977 372/1875 [====>.........................] - ETA: 11s - loss: 0.1483 - sparse_categorical_crossentropy: 0.0508 - sparse_categorical_accuracy: 0.9849 - scaled_adversarial_loss: 0.0976 379/1875 [=====>........................] - ETA: 11s - loss: 0.1481 - sparse_categorical_crossentropy: 0.0506 - sparse_categorical_accuracy: 0.9849 - scaled_adversarial_loss: 0.0975 387/1875 [=====>........................] - ETA: 11s - loss: 0.1478 - sparse_categorical_crossentropy: 0.0505 - sparse_categorical_accuracy: 0.9849 - scaled_adversarial_loss: 0.0974 394/1875 [=====>........................] - ETA: 10s - loss: 0.1476 - sparse_categorical_crossentropy: 0.0503 - sparse_categorical_accuracy: 0.9849 - scaled_adversarial_loss: 0.0973 401/1875 [=====>........................] - ETA: 10s - loss: 0.1474 - sparse_categorical_crossentropy: 0.0502 - sparse_categorical_accuracy: 0.9850 - scaled_adversarial_loss: 0.0972 408/1875 [=====>........................] - ETA: 10s - loss: 0.1472 - sparse_categorical_crossentropy: 0.0501 - sparse_categorical_accuracy: 0.9850 - scaled_adversarial_loss: 0.0971 416/1875 [=====>........................] - ETA: 10s - loss: 0.1469 - sparse_categorical_crossentropy: 0.0499 - sparse_categorical_accuracy: 0.9850 - scaled_adversarial_loss: 0.0970 424/1875 [=====>........................] - ETA: 10s - loss: 0.1467 - sparse_categorical_crossentropy: 0.0498 - sparse_categorical_accuracy: 0.9850 - scaled_adversarial_loss: 0.0969 431/1875 [=====>........................] - ETA: 10s - loss: 0.1465 - sparse_categorical_crossentropy: 0.0497 - sparse_categorical_accuracy: 0.9850 - scaled_adversarial_loss: 0.0968 438/1875 [======>.......................] - ETA: 10s - loss: 0.1463 - sparse_categorical_crossentropy: 0.0496 - sparse_categorical_accuracy: 0.9851 - scaled_adversarial_loss: 0.0967 445/1875 [======>.......................] - ETA: 10s - loss: 0.1461 - sparse_categorical_crossentropy: 0.0495 - sparse_categorical_accuracy: 0.9851 - scaled_adversarial_loss: 0.0967 453/1875 [======>.......................] - ETA: 10s - loss: 0.1459 - sparse_categorical_crossentropy: 0.0494 - sparse_categorical_accuracy: 0.9851 - scaled_adversarial_loss: 0.0966 460/1875 [======>.......................] - ETA: 10s - loss: 0.1458 - sparse_categorical_crossentropy: 0.0492 - sparse_categorical_accuracy: 0.9851 - scaled_adversarial_loss: 0.0965 467/1875 [======>.......................] - ETA: 10s - loss: 0.1456 - sparse_categorical_crossentropy: 0.0491 - sparse_categorical_accuracy: 0.9851 - scaled_adversarial_loss: 0.0965 474/1875 [======>.......................] - ETA: 10s - loss: 0.1454 - sparse_categorical_crossentropy: 0.0490 - sparse_categorical_accuracy: 0.9851 - scaled_adversarial_loss: 0.0964 482/1875 [======>.......................] - ETA: 10s - loss: 0.1453 - sparse_categorical_crossentropy: 0.0489 - sparse_categorical_accuracy: 0.9852 - scaled_adversarial_loss: 0.0963 490/1875 [======>.......................] - ETA: 10s - loss: 0.1451 - sparse_categorical_crossentropy: 0.0488 - sparse_categorical_accuracy: 0.9852 - scaled_adversarial_loss: 0.0963 497/1875 [======>.......................] - ETA: 10s - loss: 0.1449 - sparse_categorical_crossentropy: 0.0487 - sparse_categorical_accuracy: 0.9852 - scaled_adversarial_loss: 0.0962 505/1875 [=======>......................] - ETA: 10s - loss: 0.1448 - sparse_categorical_crossentropy: 0.0486 - sparse_categorical_accuracy: 0.9852 - scaled_adversarial_loss: 0.0962 513/1875 [=======>......................] - ETA: 10s - loss: 0.1446 - sparse_categorical_crossentropy: 0.0485 - sparse_categorical_accuracy: 0.9852 - scaled_adversarial_loss: 0.0961 520/1875 [=======>......................] - ETA: 9s - loss: 0.1444 - sparse_categorical_crossentropy: 0.0484 - sparse_categorical_accuracy: 0.9852 - scaled_adversarial_loss: 0.0960 527/1875 [=======>......................] - ETA: 9s - loss: 0.1443 - sparse_categorical_crossentropy: 0.0483 - sparse_categorical_accuracy: 0.9853 - scaled_adversarial_loss: 0.0960 534/1875 [=======>......................] - ETA: 9s - loss: 0.1441 - sparse_categorical_crossentropy: 0.0482 - sparse_categorical_accuracy: 0.9853 - scaled_adversarial_loss: 0.0959 541/1875 [=======>......................] - ETA: 9s - loss: 0.1440 - sparse_categorical_crossentropy: 0.0481 - sparse_categorical_accuracy: 0.9853 - scaled_adversarial_loss: 0.0958 548/1875 [=======>......................] - ETA: 9s - loss: 0.1438 - sparse_categorical_crossentropy: 0.0481 - sparse_categorical_accuracy: 0.9853 - scaled_adversarial_loss: 0.0958 555/1875 [=======>......................] - ETA: 9s - loss: 0.1437 - sparse_categorical_crossentropy: 0.0480 - sparse_categorical_accuracy: 0.9853 - scaled_adversarial_loss: 0.0957 562/1875 [=======>......................] - ETA: 9s - loss: 0.1436 - sparse_categorical_crossentropy: 0.0479 - sparse_categorical_accuracy: 0.9853 - scaled_adversarial_loss: 0.0956 569/1875 [========>.....................] - ETA: 9s - loss: 0.1434 - sparse_categorical_crossentropy: 0.0478 - sparse_categorical_accuracy: 0.9853 - scaled_adversarial_loss: 0.0956 576/1875 [========>.....................] - ETA: 9s - loss: 0.1433 - sparse_categorical_crossentropy: 0.0478 - sparse_categorical_accuracy: 0.9853 - scaled_adversarial_loss: 0.0955 583/1875 [========>.....................] - ETA: 9s - loss: 0.1432 - sparse_categorical_crossentropy: 0.0477 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0955 589/1875 [========>.....................] - ETA: 9s - loss: 0.1431 - sparse_categorical_crossentropy: 0.0477 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0954 595/1875 [========>.....................] - ETA: 9s - loss: 0.1430 - sparse_categorical_crossentropy: 0.0476 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0954 602/1875 [========>.....................] - ETA: 9s - loss: 0.1429 - sparse_categorical_crossentropy: 0.0476 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0953 609/1875 [========>.....................] - ETA: 9s - loss: 0.1428 - sparse_categorical_crossentropy: 0.0475 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0953 616/1875 [========>.....................] - ETA: 9s - loss: 0.1427 - sparse_categorical_crossentropy: 0.0475 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0952 624/1875 [========>.....................] - ETA: 9s - loss: 0.1427 - sparse_categorical_crossentropy: 0.0475 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0952 632/1875 [=========>....................] - ETA: 9s - loss: 0.1426 - sparse_categorical_crossentropy: 0.0474 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0951 639/1875 [=========>....................] - ETA: 9s - loss: 0.1425 - sparse_categorical_crossentropy: 0.0474 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0951 647/1875 [=========>....................] - ETA: 9s - loss: 0.1424 - sparse_categorical_crossentropy: 0.0474 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0950 655/1875 [=========>....................] - ETA: 9s - loss: 0.1423 - sparse_categorical_crossentropy: 0.0473 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0950 663/1875 [=========>....................] - ETA: 8s - loss: 0.1423 - sparse_categorical_crossentropy: 0.0473 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0949 670/1875 [=========>....................] - ETA: 8s - loss: 0.1422 - sparse_categorical_crossentropy: 0.0473 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0949 677/1875 [=========>....................] - ETA: 8s - loss: 0.1421 - sparse_categorical_crossentropy: 0.0473 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0949 685/1875 [=========>....................] - ETA: 8s - loss: 0.1420 - sparse_categorical_crossentropy: 0.0472 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0948 693/1875 [==========>...................] - ETA: 8s - loss: 0.1420 - sparse_categorical_crossentropy: 0.0472 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0948 700/1875 [==========>...................] - ETA: 8s - loss: 0.1419 - sparse_categorical_crossentropy: 0.0472 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0947 708/1875 [==========>...................] - ETA: 8s - loss: 0.1418 - sparse_categorical_crossentropy: 0.0472 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0947 716/1875 [==========>...................] - ETA: 8s - loss: 0.1417 - sparse_categorical_crossentropy: 0.0471 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0946 723/1875 [==========>...................] - ETA: 8s - loss: 0.1417 - sparse_categorical_crossentropy: 0.0471 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0946 730/1875 [==========>...................] - ETA: 8s - loss: 0.1416 - sparse_categorical_crossentropy: 0.0471 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0945 737/1875 [==========>...................] - ETA: 8s - loss: 0.1415 - sparse_categorical_crossentropy: 0.0471 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0945 744/1875 [==========>...................] - ETA: 8s - loss: 0.1415 - sparse_categorical_crossentropy: 0.0470 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0944 751/1875 [===========>..................] - ETA: 8s - loss: 0.1414 - sparse_categorical_crossentropy: 0.0470 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0944 758/1875 [===========>..................] - ETA: 8s - loss: 0.1413 - sparse_categorical_crossentropy: 0.0470 - sparse_categorical_accuracy: 0.9854 - scaled_adversarial_loss: 0.0943 765/1875 [===========>..................] - ETA: 8s - loss: 0.1412 - sparse_categorical_crossentropy: 0.0469 - sparse_categorical_accuracy: 0.9855 - scaled_adversarial_loss: 0.0943 773/1875 [===========>..................] - ETA: 8s - loss: 0.1412 - sparse_categorical_crossentropy: 0.0469 - sparse_categorical_accuracy: 0.9855 - scaled_adversarial_loss: 0.0943 781/1875 [===========>..................] - ETA: 8s - loss: 0.1411 - sparse_categorical_crossentropy: 0.0469 - sparse_categorical_accuracy: 0.9855 - scaled_adversarial_loss: 0.0942 788/1875 [===========>..................] - ETA: 8s - loss: 0.1410 - sparse_categorical_crossentropy: 0.0468 - sparse_categorical_accuracy: 0.9855 - scaled_adversarial_loss: 0.0942 796/1875 [===========>..................] - ETA: 7s - loss: 0.1409 - sparse_categorical_crossentropy: 0.0468 - sparse_categorical_accuracy: 0.9855 - scaled_adversarial_loss: 0.0941 804/1875 [===========>..................] - ETA: 7s - loss: 0.1409 - sparse_categorical_crossentropy: 0.0468 - sparse_categorical_accuracy: 0.9855 - scaled_adversarial_loss: 0.0941 812/1875 [===========>..................] - ETA: 7s - loss: 0.1408 - sparse_categorical_crossentropy: 0.0468 - sparse_categorical_accuracy: 0.9855 - scaled_adversarial_loss: 0.0940 819/1875 [============>.................] - ETA: 7s - loss: 0.1407 - sparse_categorical_crossentropy: 0.0467 - sparse_categorical_accuracy: 0.9855 - scaled_adversarial_loss: 0.0940 826/1875 [============>.................] - ETA: 7s - loss: 0.1406 - sparse_categorical_crossentropy: 0.0467 - sparse_categorical_accuracy: 0.9855 - scaled_adversarial_loss: 0.0939 834/1875 [============>.................] - ETA: 7s - loss: 0.1406 - sparse_categorical_crossentropy: 0.0467 - sparse_categorical_accuracy: 0.9855 - scaled_adversarial_loss: 0.0939 842/1875 [============>.................] - ETA: 7s - loss: 0.1405 - sparse_categorical_crossentropy: 0.0466 - sparse_categorical_accuracy: 0.9855 - scaled_adversarial_loss: 0.0939 850/1875 [============>.................] - ETA: 7s - loss: 0.1404 - sparse_categorical_crossentropy: 0.0466 - sparse_categorical_accuracy: 0.9855 - scaled_adversarial_loss: 0.0938 857/1875 [============>.................] - ETA: 7s - loss: 0.1404 - sparse_categorical_crossentropy: 0.0466 - sparse_categorical_accuracy: 0.9855 - scaled_adversarial_loss: 0.0938 864/1875 [============>.................] - ETA: 7s - loss: 0.1403 - sparse_categorical_crossentropy: 0.0466 - sparse_categorical_accuracy: 0.9855 - scaled_adversarial_loss: 0.0937 871/1875 [============>.................] - ETA: 7s - loss: 0.1402 - sparse_categorical_crossentropy: 0.0465 - sparse_categorical_accuracy: 0.9855 - scaled_adversarial_loss: 0.0937 878/1875 [=============>................] - ETA: 7s - loss: 0.1402 - sparse_categorical_crossentropy: 0.0465 - sparse_categorical_accuracy: 0.9855 - scaled_adversarial_loss: 0.0936 885/1875 [=============>................] - ETA: 7s - loss: 0.1401 - sparse_categorical_crossentropy: 0.0465 - sparse_categorical_accuracy: 0.9856 - scaled_adversarial_loss: 0.0936 892/1875 [=============>................] - ETA: 7s - loss: 0.1400 - sparse_categorical_crossentropy: 0.0465 - sparse_categorical_accuracy: 0.9856 - scaled_adversarial_loss: 0.0936 899/1875 [=============>................] - ETA: 7s - loss: 0.1399 - sparse_categorical_crossentropy: 0.0464 - sparse_categorical_accuracy: 0.9856 - scaled_adversarial_loss: 0.0935 906/1875 [=============>................] - ETA: 7s - loss: 0.1399 - sparse_categorical_crossentropy: 0.0464 - sparse_categorical_accuracy: 0.9856 - scaled_adversarial_loss: 0.0935 912/1875 [=============>................] - ETA: 7s - loss: 0.1398 - sparse_categorical_crossentropy: 0.0464 - sparse_categorical_accuracy: 0.9856 - scaled_adversarial_loss: 0.0934 919/1875 [=============>................] - ETA: 7s - loss: 0.1398 - sparse_categorical_crossentropy: 0.0464 - sparse_categorical_accuracy: 0.9856 - scaled_adversarial_loss: 0.0934 926/1875 [=============>................] - ETA: 6s - loss: 0.1397 - sparse_categorical_crossentropy: 0.0463 - sparse_categorical_accuracy: 0.9856 - scaled_adversarial_loss: 0.0933 934/1875 [=============>................] - ETA: 6s - loss: 0.1396 - sparse_categorical_crossentropy: 0.0463 - sparse_categorical_accuracy: 0.9856 - scaled_adversarial_loss: 0.0933 941/1875 [==============>...............] - ETA: 6s - loss: 0.1395 - sparse_categorical_crossentropy: 0.0463 - sparse_categorical_accuracy: 0.9856 - scaled_adversarial_loss: 0.0933 948/1875 [==============>...............] - ETA: 6s - loss: 0.1395 - sparse_categorical_crossentropy: 0.0463 - sparse_categorical_accuracy: 0.9856 - scaled_adversarial_loss: 0.0932 955/1875 [==============>...............] - ETA: 6s - loss: 0.1394 - sparse_categorical_crossentropy: 0.0462 - sparse_categorical_accuracy: 0.9856 - scaled_adversarial_loss: 0.0932 963/1875 [==============>...............] - ETA: 6s - loss: 0.1393 - sparse_categorical_crossentropy: 0.0462 - sparse_categorical_accuracy: 0.9856 - scaled_adversarial_loss: 0.0931 970/1875 [==============>...............] - ETA: 6s - loss: 0.1393 - sparse_categorical_crossentropy: 0.0462 - sparse_categorical_accuracy: 0.9856 - scaled_adversarial_loss: 0.0931 978/1875 [==============>...............] - ETA: 6s - loss: 0.1392 - sparse_categorical_crossentropy: 0.0462 - sparse_categorical_accuracy: 0.9856 - scaled_adversarial_loss: 0.0930 986/1875 [==============>...............] - ETA: 6s - loss: 0.1391 - sparse_categorical_crossentropy: 0.0461 - sparse_categorical_accuracy: 0.9857 - scaled_adversarial_loss: 0.0930 993/1875 [==============>...............] - ETA: 6s - loss: 0.1390 - sparse_categorical_crossentropy: 0.0461 - sparse_categorical_accuracy: 0.9857 - scaled_adversarial_loss: 0.0929 1000/1875 [===============>..............] - ETA: 6s - loss: 0.1390 - sparse_categorical_crossentropy: 0.0461 - sparse_categorical_accuracy: 0.9857 - scaled_adversarial_loss: 0.0929 1008/1875 [===============>..............] - ETA: 6s - loss: 0.1389 - sparse_categorical_crossentropy: 0.0461 - sparse_categorical_accuracy: 0.9857 - scaled_adversarial_loss: 0.0929 1015/1875 [===============>..............] - ETA: 6s - loss: 0.1389 - sparse_categorical_crossentropy: 0.0460 - sparse_categorical_accuracy: 0.9857 - scaled_adversarial_loss: 0.0928 1022/1875 [===============>..............] - ETA: 6s - loss: 0.1388 - sparse_categorical_crossentropy: 0.0460 - sparse_categorical_accuracy: 0.9857 - scaled_adversarial_loss: 0.0928 1029/1875 [===============>..............] - ETA: 6s - loss: 0.1387 - sparse_categorical_crossentropy: 0.0460 - sparse_categorical_accuracy: 0.9857 - scaled_adversarial_loss: 0.0927 1037/1875 [===============>..............] - ETA: 6s - loss: 0.1387 - sparse_categorical_crossentropy: 0.0460 - sparse_categorical_accuracy: 0.9857 - scaled_adversarial_loss: 0.0927 1044/1875 [===============>..............] - ETA: 6s - loss: 0.1386 - sparse_categorical_crossentropy: 0.0460 - sparse_categorical_accuracy: 0.9857 - scaled_adversarial_loss: 0.0926 1052/1875 [===============>..............] - ETA: 6s - loss: 0.1385 - sparse_categorical_crossentropy: 0.0459 - sparse_categorical_accuracy: 0.9857 - scaled_adversarial_loss: 0.0926 1059/1875 [===============>..............] - ETA: 6s - loss: 0.1385 - sparse_categorical_crossentropy: 0.0459 - sparse_categorical_accuracy: 0.9857 - scaled_adversarial_loss: 0.0926 1066/1875 [================>.............] - ETA: 5s - loss: 0.1384 - sparse_categorical_crossentropy: 0.0459 - sparse_categorical_accuracy: 0.9857 - scaled_adversarial_loss: 0.0925 1073/1875 [================>.............] - ETA: 5s - loss: 0.1384 - sparse_categorical_crossentropy: 0.0459 - sparse_categorical_accuracy: 0.9857 - scaled_adversarial_loss: 0.0925 1080/1875 [================>.............] - ETA: 5s - loss: 0.1383 - sparse_categorical_crossentropy: 0.0459 - sparse_categorical_accuracy: 0.9857 - scaled_adversarial_loss: 0.0924 1087/1875 [================>.............] - ETA: 5s - loss: 0.1382 - sparse_categorical_crossentropy: 0.0458 - sparse_categorical_accuracy: 0.9857 - scaled_adversarial_loss: 0.0924 1094/1875 [================>.............] - ETA: 5s - loss: 0.1382 - sparse_categorical_crossentropy: 0.0458 - sparse_categorical_accuracy: 0.9858 - scaled_adversarial_loss: 0.0924 1101/1875 [================>.............] - ETA: 5s - loss: 0.1381 - sparse_categorical_crossentropy: 0.0458 - sparse_categorical_accuracy: 0.9858 - scaled_adversarial_loss: 0.0923 1109/1875 [================>.............] - ETA: 5s - loss: 0.1381 - sparse_categorical_crossentropy: 0.0458 - sparse_categorical_accuracy: 0.9858 - scaled_adversarial_loss: 0.0923 1116/1875 [================>.............] - ETA: 5s - loss: 0.1380 - sparse_categorical_crossentropy: 0.0458 - sparse_categorical_accuracy: 0.9858 - scaled_adversarial_loss: 0.0923 1123/1875 [================>.............] - ETA: 5s - loss: 0.1379 - sparse_categorical_crossentropy: 0.0457 - sparse_categorical_accuracy: 0.9858 - scaled_adversarial_loss: 0.0922 1131/1875 [=================>............] - ETA: 5s - loss: 0.1379 - sparse_categorical_crossentropy: 0.0457 - sparse_categorical_accuracy: 0.9858 - scaled_adversarial_loss: 0.0922 1139/1875 [=================>............] - ETA: 5s - loss: 0.1378 - sparse_categorical_crossentropy: 0.0457 - sparse_categorical_accuracy: 0.9858 - scaled_adversarial_loss: 0.0921 1146/1875 [=================>............] - ETA: 5s - loss: 0.1378 - sparse_categorical_crossentropy: 0.0457 - sparse_categorical_accuracy: 0.9858 - scaled_adversarial_loss: 0.0921 1153/1875 [=================>............] - ETA: 5s - loss: 0.1377 - sparse_categorical_crossentropy: 0.0457 - sparse_categorical_accuracy: 0.9858 - scaled_adversarial_loss: 0.0921 1161/1875 [=================>............] - ETA: 5s - loss: 0.1376 - sparse_categorical_crossentropy: 0.0456 - sparse_categorical_accuracy: 0.9858 - scaled_adversarial_loss: 0.0920 1169/1875 [=================>............] - ETA: 5s - loss: 0.1376 - sparse_categorical_crossentropy: 0.0456 - sparse_categorical_accuracy: 0.9858 - scaled_adversarial_loss: 0.0920 1176/1875 [=================>............] - ETA: 5s - loss: 0.1375 - sparse_categorical_crossentropy: 0.0456 - sparse_categorical_accuracy: 0.9858 - scaled_adversarial_loss: 0.0919 1183/1875 [=================>............] - ETA: 5s - loss: 0.1375 - sparse_categorical_crossentropy: 0.0456 - sparse_categorical_accuracy: 0.9858 - scaled_adversarial_loss: 0.0919 1191/1875 [==================>...........] - ETA: 5s - loss: 0.1374 - sparse_categorical_crossentropy: 0.0456 - sparse_categorical_accuracy: 0.9858 - scaled_adversarial_loss: 0.0919 1198/1875 [==================>...........] - ETA: 4s - loss: 0.1374 - sparse_categorical_crossentropy: 0.0455 - sparse_categorical_accuracy: 0.9858 - scaled_adversarial_loss: 0.0918 1205/1875 [==================>...........] - ETA: 4s - loss: 0.1373 - sparse_categorical_crossentropy: 0.0455 - sparse_categorical_accuracy: 0.9859 - scaled_adversarial_loss: 0.0918 1212/1875 [==================>...........] - ETA: 4s - loss: 0.1373 - sparse_categorical_crossentropy: 0.0455 - sparse_categorical_accuracy: 0.9859 - scaled_adversarial_loss: 0.0918 1219/1875 [==================>...........] - ETA: 4s - loss: 0.1372 - sparse_categorical_crossentropy: 0.0455 - sparse_categorical_accuracy: 0.9859 - scaled_adversarial_loss: 0.0917 1226/1875 [==================>...........] - ETA: 4s - loss: 0.1372 - sparse_categorical_crossentropy: 0.0455 - sparse_categorical_accuracy: 0.9859 - scaled_adversarial_loss: 0.0917 1234/1875 [==================>...........] - ETA: 4s - loss: 0.1371 - sparse_categorical_crossentropy: 0.0454 - sparse_categorical_accuracy: 0.9859 - scaled_adversarial_loss: 0.0917 1241/1875 [==================>...........] - ETA: 4s - loss: 0.1370 - sparse_categorical_crossentropy: 0.0454 - sparse_categorical_accuracy: 0.9859 - scaled_adversarial_loss: 0.0916 1248/1875 [==================>...........] - ETA: 4s - loss: 0.1370 - sparse_categorical_crossentropy: 0.0454 - sparse_categorical_accuracy: 0.9859 - scaled_adversarial_loss: 0.0916 1256/1875 [===================>..........] - ETA: 4s - loss: 0.1369 - sparse_categorical_crossentropy: 0.0454 - sparse_categorical_accuracy: 0.9859 - scaled_adversarial_loss: 0.0916 1264/1875 [===================>..........] - ETA: 4s - loss: 0.1369 - sparse_categorical_crossentropy: 0.0454 - sparse_categorical_accuracy: 0.9859 - scaled_adversarial_loss: 0.0915 1271/1875 [===================>..........] - ETA: 4s - loss: 0.1368 - sparse_categorical_crossentropy: 0.0453 - sparse_categorical_accuracy: 0.9859 - scaled_adversarial_loss: 0.0915 1279/1875 [===================>..........] - ETA: 4s - loss: 0.1368 - sparse_categorical_crossentropy: 0.0453 - sparse_categorical_accuracy: 0.9859 - scaled_adversarial_loss: 0.0915 1287/1875 [===================>..........] - ETA: 4s - loss: 0.1367 - sparse_categorical_crossentropy: 0.0453 - sparse_categorical_accuracy: 0.9859 - scaled_adversarial_loss: 0.0914 1295/1875 [===================>..........] - ETA: 4s - loss: 0.1367 - sparse_categorical_crossentropy: 0.0453 - sparse_categorical_accuracy: 0.9859 - scaled_adversarial_loss: 0.0914 1302/1875 [===================>..........] - ETA: 4s - loss: 0.1366 - sparse_categorical_crossentropy: 0.0453 - sparse_categorical_accuracy: 0.9859 - scaled_adversarial_loss: 0.0914 1309/1875 [===================>..........] - ETA: 4s - loss: 0.1366 - sparse_categorical_crossentropy: 0.0453 - sparse_categorical_accuracy: 0.9859 - scaled_adversarial_loss: 0.0913 1316/1875 [====================>.........] - ETA: 4s - loss: 0.1365 - sparse_categorical_crossentropy: 0.0452 - sparse_categorical_accuracy: 0.9860 - scaled_adversarial_loss: 0.0913 1324/1875 [====================>.........] - ETA: 4s - loss: 0.1365 - sparse_categorical_crossentropy: 0.0452 - sparse_categorical_accuracy: 0.9860 - scaled_adversarial_loss: 0.0913 1331/1875 [====================>.........] - ETA: 3s - loss: 0.1364 - sparse_categorical_crossentropy: 0.0452 - sparse_categorical_accuracy: 0.9860 - scaled_adversarial_loss: 0.0912 1339/1875 [====================>.........] - ETA: 3s - loss: 0.1364 - sparse_categorical_crossentropy: 0.0452 - sparse_categorical_accuracy: 0.9860 - scaled_adversarial_loss: 0.0912 1346/1875 [====================>.........] - ETA: 3s - loss: 0.1363 - sparse_categorical_crossentropy: 0.0452 - sparse_categorical_accuracy: 0.9860 - scaled_adversarial_loss: 0.0912 1353/1875 [====================>.........] - ETA: 3s - loss: 0.1363 - sparse_categorical_crossentropy: 0.0451 - sparse_categorical_accuracy: 0.9860 - scaled_adversarial_loss: 0.0911 1359/1875 [====================>.........] - ETA: 3s - loss: 0.1362 - sparse_categorical_crossentropy: 0.0451 - sparse_categorical_accuracy: 0.9860 - scaled_adversarial_loss: 0.0911 1366/1875 [====================>.........] - ETA: 3s - loss: 0.1362 - sparse_categorical_crossentropy: 0.0451 - sparse_categorical_accuracy: 0.9860 - scaled_adversarial_loss: 0.0911 1373/1875 [====================>.........] - ETA: 3s - loss: 0.1362 - sparse_categorical_crossentropy: 0.0451 - sparse_categorical_accuracy: 0.9860 - scaled_adversarial_loss: 0.0910 1381/1875 [=====================>........] - ETA: 3s - loss: 0.1361 - sparse_categorical_crossentropy: 0.0451 - sparse_categorical_accuracy: 0.9860 - scaled_adversarial_loss: 0.0910 1388/1875 [=====================>........] - ETA: 3s - loss: 0.1361 - sparse_categorical_crossentropy: 0.0451 - sparse_categorical_accuracy: 0.9860 - scaled_adversarial_loss: 0.0910 1395/1875 [=====================>........] - ETA: 3s - loss: 0.1360 - sparse_categorical_crossentropy: 0.0451 - sparse_categorical_accuracy: 0.9860 - scaled_adversarial_loss: 0.0910 1403/1875 [=====================>........] - ETA: 3s - loss: 0.1360 - sparse_categorical_crossentropy: 0.0450 - sparse_categorical_accuracy: 0.9860 - scaled_adversarial_loss: 0.0909 1410/1875 [=====================>........] - ETA: 3s - loss: 0.1359 - sparse_categorical_crossentropy: 0.0450 - sparse_categorical_accuracy: 0.9860 - scaled_adversarial_loss: 0.0909 1417/1875 [=====================>........] - ETA: 3s - loss: 0.1359 - sparse_categorical_crossentropy: 0.0450 - sparse_categorical_accuracy: 0.9860 - scaled_adversarial_loss: 0.0909 1425/1875 [=====================>........] - ETA: 3s - loss: 0.1358 - sparse_categorical_crossentropy: 0.0450 - sparse_categorical_accuracy: 0.9860 - scaled_adversarial_loss: 0.0908 1433/1875 [=====================>........] - ETA: 3s - loss: 0.1358 - sparse_categorical_crossentropy: 0.0450 - sparse_categorical_accuracy: 0.9860 - scaled_adversarial_loss: 0.0908 1440/1875 [======================>.......] - ETA: 3s - loss: 0.1357 - sparse_categorical_crossentropy: 0.0450 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0908 1447/1875 [======================>.......] - ETA: 3s - loss: 0.1357 - sparse_categorical_crossentropy: 0.0449 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0907 1454/1875 [======================>.......] - ETA: 3s - loss: 0.1356 - sparse_categorical_crossentropy: 0.0449 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0907 1461/1875 [======================>.......] - ETA: 3s - loss: 0.1356 - sparse_categorical_crossentropy: 0.0449 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0907 1468/1875 [======================>.......] - ETA: 2s - loss: 0.1355 - sparse_categorical_crossentropy: 0.0449 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0906 1475/1875 [======================>.......] - ETA: 2s - loss: 0.1355 - sparse_categorical_crossentropy: 0.0449 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0906 1482/1875 [======================>.......] - ETA: 2s - loss: 0.1355 - sparse_categorical_crossentropy: 0.0449 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0906 1489/1875 [======================>.......] - ETA: 2s - loss: 0.1354 - sparse_categorical_crossentropy: 0.0448 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0906 1496/1875 [======================>.......] - ETA: 2s - loss: 0.1354 - sparse_categorical_crossentropy: 0.0448 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0905 1502/1875 [=======================>......] - ETA: 2s - loss: 0.1353 - sparse_categorical_crossentropy: 0.0448 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0905 1508/1875 [=======================>......] - ETA: 2s - loss: 0.1353 - sparse_categorical_crossentropy: 0.0448 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0905 1516/1875 [=======================>......] - ETA: 2s - loss: 0.1352 - sparse_categorical_crossentropy: 0.0448 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0904 1524/1875 [=======================>......] - ETA: 2s - loss: 0.1352 - sparse_categorical_crossentropy: 0.0448 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0904 1531/1875 [=======================>......] - ETA: 2s - loss: 0.1352 - sparse_categorical_crossentropy: 0.0448 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0904 1538/1875 [=======================>......] - ETA: 2s - loss: 0.1351 - sparse_categorical_crossentropy: 0.0447 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0904 1545/1875 [=======================>......] - ETA: 2s - loss: 0.1351 - sparse_categorical_crossentropy: 0.0447 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0903 1552/1875 [=======================>......] - ETA: 2s - loss: 0.1350 - sparse_categorical_crossentropy: 0.0447 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0903 1560/1875 [=======================>......] - ETA: 2s - loss: 0.1350 - sparse_categorical_crossentropy: 0.0447 - sparse_categorical_accuracy: 0.9862 - scaled_adversarial_loss: 0.0903 1567/1875 [========================>.....] - ETA: 2s - loss: 0.1349 - sparse_categorical_crossentropy: 0.0447 - sparse_categorical_accuracy: 0.9862 - scaled_adversarial_loss: 0.0902 1576/1875 [========================>.....] - ETA: 2s - loss: 0.1349 - sparse_categorical_crossentropy: 0.0447 - sparse_categorical_accuracy: 0.9862 - scaled_adversarial_loss: 0.0902 1586/1875 [========================>.....] - ETA: 2s - loss: 0.1348 - sparse_categorical_crossentropy: 0.0446 - sparse_categorical_accuracy: 0.9862 - scaled_adversarial_loss: 0.0902 1595/1875 [========================>.....] - ETA: 2s - loss: 0.1348 - sparse_categorical_crossentropy: 0.0446 - sparse_categorical_accuracy: 0.9862 - scaled_adversarial_loss: 0.0901 1604/1875 [========================>.....] - ETA: 1s - loss: 0.1347 - sparse_categorical_crossentropy: 0.0446 - sparse_categorical_accuracy: 0.9862 - scaled_adversarial_loss: 0.0901 1613/1875 [========================>.....] - ETA: 1s - loss: 0.1346 - sparse_categorical_crossentropy: 0.0446 - sparse_categorical_accuracy: 0.9862 - scaled_adversarial_loss: 0.0901 1623/1875 [========================>.....] - ETA: 1s - loss: 0.1346 - sparse_categorical_crossentropy: 0.0446 - sparse_categorical_accuracy: 0.9862 - scaled_adversarial_loss: 0.0900 1633/1875 [=========================>....] - ETA: 1s - loss: 0.1345 - sparse_categorical_crossentropy: 0.0445 - sparse_categorical_accuracy: 0.9862 - scaled_adversarial_loss: 0.0900 1642/1875 [=========================>....] - ETA: 1s - loss: 0.1344 - sparse_categorical_crossentropy: 0.0445 - sparse_categorical_accuracy: 0.9862 - scaled_adversarial_loss: 0.0899 1652/1875 [=========================>....] - ETA: 1s - loss: 0.1344 - sparse_categorical_crossentropy: 0.0445 - sparse_categorical_accuracy: 0.9862 - scaled_adversarial_loss: 0.0899 1662/1875 [=========================>....] - ETA: 1s - loss: 0.1343 - sparse_categorical_crossentropy: 0.0445 - sparse_categorical_accuracy: 0.9862 - scaled_adversarial_loss: 0.0899 1671/1875 [=========================>....] - ETA: 1s - loss: 0.1343 - sparse_categorical_crossentropy: 0.0445 - sparse_categorical_accuracy: 0.9862 - scaled_adversarial_loss: 0.0898 1680/1875 [=========================>....] - ETA: 1s - loss: 0.1342 - sparse_categorical_crossentropy: 0.0444 - sparse_categorical_accuracy: 0.9862 - scaled_adversarial_loss: 0.0898 1689/1875 [==========================>...] - ETA: 1s - loss: 0.1342 - sparse_categorical_crossentropy: 0.0444 - sparse_categorical_accuracy: 0.9863 - scaled_adversarial_loss: 0.0898 1698/1875 [==========================>...] - ETA: 1s - loss: 0.1341 - sparse_categorical_crossentropy: 0.0444 - sparse_categorical_accuracy: 0.9863 - scaled_adversarial_loss: 0.0897 1708/1875 [==========================>...] - ETA: 1s - loss: 0.1341 - sparse_categorical_crossentropy: 0.0444 - sparse_categorical_accuracy: 0.9863 - scaled_adversarial_loss: 0.0897 1717/1875 [==========================>...] - ETA: 1s - loss: 0.1340 - sparse_categorical_crossentropy: 0.0444 - sparse_categorical_accuracy: 0.9863 - scaled_adversarial_loss: 0.0896 1727/1875 [==========================>...] - ETA: 1s - loss: 0.1340 - sparse_categorical_crossentropy: 0.0443 - sparse_categorical_accuracy: 0.9863 - scaled_adversarial_loss: 0.0896 1737/1875 [==========================>...] - ETA: 0s - loss: 0.1339 - sparse_categorical_crossentropy: 0.0443 - sparse_categorical_accuracy: 0.9863 - scaled_adversarial_loss: 0.0896 1747/1875 [==========================>...] - ETA: 0s - loss: 0.1338 - sparse_categorical_crossentropy: 0.0443 - sparse_categorical_accuracy: 0.9863 - scaled_adversarial_loss: 0.0895 1756/1875 [===========================>..] - ETA: 0s - loss: 0.1338 - sparse_categorical_crossentropy: 0.0443 - sparse_categorical_accuracy: 0.9863 - scaled_adversarial_loss: 0.0895 1766/1875 [===========================>..] - ETA: 0s - loss: 0.1337 - sparse_categorical_crossentropy: 0.0443 - sparse_categorical_accuracy: 0.9863 - scaled_adversarial_loss: 0.0895 1775/1875 [===========================>..] - ETA: 0s - loss: 0.1337 - sparse_categorical_crossentropy: 0.0442 - sparse_categorical_accuracy: 0.9863 - scaled_adversarial_loss: 0.0894 1785/1875 [===========================>..] - ETA: 0s - loss: 0.1336 - sparse_categorical_crossentropy: 0.0442 - sparse_categorical_accuracy: 0.9863 - scaled_adversarial_loss: 0.0894 1794/1875 [===========================>..] - ETA: 0s - loss: 0.1335 - sparse_categorical_crossentropy: 0.0442 - sparse_categorical_accuracy: 0.9863 - scaled_adversarial_loss: 0.0893 1804/1875 [===========================>..] - ETA: 0s - loss: 0.1335 - sparse_categorical_crossentropy: 0.0442 - sparse_categorical_accuracy: 0.9863 - scaled_adversarial_loss: 0.0893 1814/1875 [============================>.] - ETA: 0s - loss: 0.1334 - sparse_categorical_crossentropy: 0.0442 - sparse_categorical_accuracy: 0.9863 - scaled_adversarial_loss: 0.0893 1823/1875 [============================>.] - ETA: 0s - loss: 0.1334 - sparse_categorical_crossentropy: 0.0442 - sparse_categorical_accuracy: 0.9863 - scaled_adversarial_loss: 0.0892 1832/1875 [============================>.] - ETA: 0s - loss: 0.1333 - sparse_categorical_crossentropy: 0.0441 - sparse_categorical_accuracy: 0.9864 - scaled_adversarial_loss: 0.0892 1841/1875 [============================>.] - ETA: 0s - loss: 0.1333 - sparse_categorical_crossentropy: 0.0441 - sparse_categorical_accuracy: 0.9864 - scaled_adversarial_loss: 0.0891 1850/1875 [============================>.] - ETA: 0s - loss: 0.1332 - sparse_categorical_crossentropy: 0.0441 - sparse_categorical_accuracy: 0.9864 - scaled_adversarial_loss: 0.0891 1860/1875 [============================>.] - ETA: 0s - loss: 0.1331 - sparse_categorical_crossentropy: 0.0441 - sparse_categorical_accuracy: 0.9864 - scaled_adversarial_loss: 0.0891 1869/1875 [============================>.] - ETA: 0s - loss: 0.1331 - sparse_categorical_crossentropy: 0.0441 - sparse_categorical_accuracy: 0.9864 - scaled_adversarial_loss: 0.0890 1875/1875 [==============================] - 14s 7ms/step - loss: 0.1330 - sparse_categorical_crossentropy: 0.0440 - sparse_categorical_accuracy: 0.9864 - scaled_adversarial_loss: 0.0890 Epoch 3/5 1/1875 [..............................] - ETA: 13:49 - loss: 0.0412 - sparse_categorical_crossentropy: 0.0024 - sparse_categorical_accuracy: 1.0000 - scaled_adversarial_loss: 0.0388 8/1875 [..............................] - ETA: 14s - loss: 0.0962 - sparse_categorical_crossentropy: 0.0392 - sparse_categorical_accuracy: 0.9873 - scaled_adversarial_loss: 0.0570 15/1875 [..............................] - ETA: 13s - loss: 0.0974 - sparse_categorical_crossentropy: 0.0372 - sparse_categorical_accuracy: 0.9876 - scaled_adversarial_loss: 0.0602 22/1875 [..............................] - ETA: 13s - loss: 0.1002 - sparse_categorical_crossentropy: 0.0364 - sparse_categorical_accuracy: 0.9874 - scaled_adversarial_loss: 0.0638 30/1875 [..............................] - ETA: 13s - loss: 0.1049 - sparse_categorical_crossentropy: 0.0370 - sparse_categorical_accuracy: 0.9867 - scaled_adversarial_loss: 0.0678 37/1875 [..............................] - ETA: 13s - loss: 0.1084 - sparse_categorical_crossentropy: 0.0377 - sparse_categorical_accuracy: 0.9863 - scaled_adversarial_loss: 0.0707 44/1875 [..............................] - ETA: 13s - loss: 0.1101 - sparse_categorical_crossentropy: 0.0378 - sparse_categorical_accuracy: 0.9861 - scaled_adversarial_loss: 0.0723 51/1875 [..............................] - ETA: 13s - loss: 0.1105 - sparse_categorical_crossentropy: 0.0374 - sparse_categorical_accuracy: 0.9863 - scaled_adversarial_loss: 0.0731 58/1875 [..............................] - ETA: 13s - loss: 0.1104 - sparse_categorical_crossentropy: 0.0369 - sparse_categorical_accuracy: 0.9864 - scaled_adversarial_loss: 0.0735 66/1875 [>.............................] - ETA: 13s - loss: 0.1100 - sparse_categorical_crossentropy: 0.0364 - sparse_categorical_accuracy: 0.9866 - scaled_adversarial_loss: 0.0737 74/1875 [>.............................] - ETA: 13s - loss: 0.1098 - sparse_categorical_crossentropy: 0.0359 - sparse_categorical_accuracy: 0.9867 - scaled_adversarial_loss: 0.0739 81/1875 [>.............................] - ETA: 13s - loss: 0.1092 - sparse_categorical_crossentropy: 0.0354 - sparse_categorical_accuracy: 0.9868 - scaled_adversarial_loss: 0.0738 88/1875 [>.............................] - ETA: 13s - loss: 0.1084 - sparse_categorical_crossentropy: 0.0349 - sparse_categorical_accuracy: 0.9870 - scaled_adversarial_loss: 0.0735 96/1875 [>.............................] - ETA: 12s - loss: 0.1076 - sparse_categorical_crossentropy: 0.0345 - sparse_categorical_accuracy: 0.9872 - scaled_adversarial_loss: 0.0731 104/1875 [>.............................] - ETA: 12s - loss: 0.1070 - sparse_categorical_crossentropy: 0.0342 - sparse_categorical_accuracy: 0.9874 - scaled_adversarial_loss: 0.0728 111/1875 [>.............................] - ETA: 12s - loss: 0.1067 - sparse_categorical_crossentropy: 0.0341 - sparse_categorical_accuracy: 0.9875 - scaled_adversarial_loss: 0.0726 119/1875 [>.............................] - ETA: 12s - loss: 0.1063 - sparse_categorical_crossentropy: 0.0340 - sparse_categorical_accuracy: 0.9876 - scaled_adversarial_loss: 0.0724 127/1875 [=>............................] - ETA: 12s - loss: 0.1062 - sparse_categorical_crossentropy: 0.0340 - sparse_categorical_accuracy: 0.9876 - scaled_adversarial_loss: 0.0722 134/1875 [=>............................] - ETA: 12s - loss: 0.1062 - sparse_categorical_crossentropy: 0.0341 - sparse_categorical_accuracy: 0.9875 - scaled_adversarial_loss: 0.0721 141/1875 [=>............................] - ETA: 12s - loss: 0.1062 - sparse_categorical_crossentropy: 0.0341 - sparse_categorical_accuracy: 0.9875 - scaled_adversarial_loss: 0.0720 148/1875 [=>............................] - ETA: 12s - loss: 0.1062 - sparse_categorical_crossentropy: 0.0342 - sparse_categorical_accuracy: 0.9875 - scaled_adversarial_loss: 0.0720 155/1875 [=>............................] - ETA: 12s - loss: 0.1062 - sparse_categorical_crossentropy: 0.0343 - sparse_categorical_accuracy: 0.9875 - scaled_adversarial_loss: 0.0719 163/1875 [=>............................] - ETA: 12s - loss: 0.1062 - sparse_categorical_crossentropy: 0.0344 - sparse_categorical_accuracy: 0.9875 - scaled_adversarial_loss: 0.0719 170/1875 [=>............................] - ETA: 12s - loss: 0.1062 - sparse_categorical_crossentropy: 0.0344 - sparse_categorical_accuracy: 0.9875 - scaled_adversarial_loss: 0.0718 177/1875 [=>............................] - ETA: 12s - loss: 0.1063 - sparse_categorical_crossentropy: 0.0345 - sparse_categorical_accuracy: 0.9875 - scaled_adversarial_loss: 0.0718 184/1875 [=>............................] - ETA: 12s - loss: 0.1063 - sparse_categorical_crossentropy: 0.0345 - sparse_categorical_accuracy: 0.9875 - scaled_adversarial_loss: 0.0718 191/1875 [==>...........................] - ETA: 12s - loss: 0.1062 - sparse_categorical_crossentropy: 0.0345 - sparse_categorical_accuracy: 0.9875 - scaled_adversarial_loss: 0.0717 198/1875 [==>...........................] - ETA: 12s - loss: 0.1061 - sparse_categorical_crossentropy: 0.0345 - sparse_categorical_accuracy: 0.9875 - scaled_adversarial_loss: 0.0716 205/1875 [==>...........................] - ETA: 12s - loss: 0.1060 - sparse_categorical_crossentropy: 0.0344 - sparse_categorical_accuracy: 0.9876 - scaled_adversarial_loss: 0.0716 212/1875 [==>...........................] - ETA: 12s - loss: 0.1059 - sparse_categorical_crossentropy: 0.0344 - sparse_categorical_accuracy: 0.9876 - scaled_adversarial_loss: 0.0715 220/1875 [==>...........................] - ETA: 12s - loss: 0.1058 - sparse_categorical_crossentropy: 0.0343 - sparse_categorical_accuracy: 0.9876 - scaled_adversarial_loss: 0.0715 227/1875 [==>...........................] - ETA: 12s - loss: 0.1057 - sparse_categorical_crossentropy: 0.0343 - sparse_categorical_accuracy: 0.9877 - scaled_adversarial_loss: 0.0714 235/1875 [==>...........................] - ETA: 12s - loss: 0.1055 - sparse_categorical_crossentropy: 0.0342 - sparse_categorical_accuracy: 0.9877 - scaled_adversarial_loss: 0.0714 243/1875 [==>...........................] - ETA: 11s - loss: 0.1054 - sparse_categorical_crossentropy: 0.0341 - sparse_categorical_accuracy: 0.9878 - scaled_adversarial_loss: 0.0713 251/1875 [===>..........................] - ETA: 11s - loss: 0.1052 - sparse_categorical_crossentropy: 0.0340 - sparse_categorical_accuracy: 0.9878 - scaled_adversarial_loss: 0.0712 258/1875 [===>..........................] - ETA: 11s - loss: 0.1051 - sparse_categorical_crossentropy: 0.0339 - sparse_categorical_accuracy: 0.9879 - scaled_adversarial_loss: 0.0711 265/1875 [===>..........................] - ETA: 11s - loss: 0.1049 - sparse_categorical_crossentropy: 0.0338 - sparse_categorical_accuracy: 0.9879 - scaled_adversarial_loss: 0.0711 272/1875 [===>..........................] - ETA: 11s - loss: 0.1047 - sparse_categorical_crossentropy: 0.0337 - sparse_categorical_accuracy: 0.9880 - scaled_adversarial_loss: 0.0710 280/1875 [===>..........................] - ETA: 11s - loss: 0.1046 - sparse_categorical_crossentropy: 0.0337 - sparse_categorical_accuracy: 0.9880 - scaled_adversarial_loss: 0.0709 287/1875 [===>..........................] - ETA: 11s - loss: 0.1045 - sparse_categorical_crossentropy: 0.0336 - sparse_categorical_accuracy: 0.9880 - scaled_adversarial_loss: 0.0709 294/1875 [===>..........................] - ETA: 11s - loss: 0.1043 - sparse_categorical_crossentropy: 0.0335 - sparse_categorical_accuracy: 0.9881 - scaled_adversarial_loss: 0.0708 301/1875 [===>..........................] - ETA: 11s - loss: 0.1042 - sparse_categorical_crossentropy: 0.0334 - sparse_categorical_accuracy: 0.9881 - scaled_adversarial_loss: 0.0708 309/1875 [===>..........................] - ETA: 11s - loss: 0.1042 - sparse_categorical_crossentropy: 0.0334 - sparse_categorical_accuracy: 0.9881 - scaled_adversarial_loss: 0.0708 316/1875 [====>.........................] - ETA: 11s - loss: 0.1041 - sparse_categorical_crossentropy: 0.0333 - sparse_categorical_accuracy: 0.9881 - scaled_adversarial_loss: 0.0707 322/1875 [====>.........................] - ETA: 11s - loss: 0.1040 - sparse_categorical_crossentropy: 0.0333 - sparse_categorical_accuracy: 0.9882 - scaled_adversarial_loss: 0.0707 329/1875 [====>.........................] - ETA: 11s - loss: 0.1039 - sparse_categorical_crossentropy: 0.0333 - sparse_categorical_accuracy: 0.9882 - scaled_adversarial_loss: 0.0706 337/1875 [====>.........................] - ETA: 11s - loss: 0.1038 - sparse_categorical_crossentropy: 0.0332 - sparse_categorical_accuracy: 0.9882 - scaled_adversarial_loss: 0.0706 343/1875 [====>.........................] - ETA: 11s - loss: 0.1037 - sparse_categorical_crossentropy: 0.0332 - sparse_categorical_accuracy: 0.9883 - scaled_adversarial_loss: 0.0705 350/1875 [====>.........................] - ETA: 11s - loss: 0.1036 - sparse_categorical_crossentropy: 0.0331 - sparse_categorical_accuracy: 0.9883 - scaled_adversarial_loss: 0.0705 357/1875 [====>.........................] - ETA: 11s - loss: 0.1035 - sparse_categorical_crossentropy: 0.0331 - sparse_categorical_accuracy: 0.9883 - scaled_adversarial_loss: 0.0704 364/1875 [====>.........................] - ETA: 11s - loss: 0.1034 - sparse_categorical_crossentropy: 0.0331 - sparse_categorical_accuracy: 0.9883 - scaled_adversarial_loss: 0.0704 372/1875 [====>.........................] - ETA: 11s - loss: 0.1033 - sparse_categorical_crossentropy: 0.0330 - sparse_categorical_accuracy: 0.9884 - scaled_adversarial_loss: 0.0703 379/1875 [=====>........................] - ETA: 11s - loss: 0.1032 - sparse_categorical_crossentropy: 0.0330 - sparse_categorical_accuracy: 0.9884 - scaled_adversarial_loss: 0.0703 386/1875 [=====>........................] - ETA: 10s - loss: 0.1031 - sparse_categorical_crossentropy: 0.0329 - sparse_categorical_accuracy: 0.9884 - scaled_adversarial_loss: 0.0702 394/1875 [=====>........................] - ETA: 10s - loss: 0.1030 - sparse_categorical_crossentropy: 0.0328 - sparse_categorical_accuracy: 0.9885 - scaled_adversarial_loss: 0.0701 402/1875 [=====>........................] - ETA: 10s - loss: 0.1029 - sparse_categorical_crossentropy: 0.0328 - sparse_categorical_accuracy: 0.9885 - scaled_adversarial_loss: 0.0701 409/1875 [=====>........................] - ETA: 10s - loss: 0.1027 - sparse_categorical_crossentropy: 0.0327 - sparse_categorical_accuracy: 0.9885 - scaled_adversarial_loss: 0.0700 417/1875 [=====>........................] - ETA: 10s - loss: 0.1026 - sparse_categorical_crossentropy: 0.0327 - sparse_categorical_accuracy: 0.9886 - scaled_adversarial_loss: 0.0699 425/1875 [=====>........................] - ETA: 10s - loss: 0.1025 - sparse_categorical_crossentropy: 0.0326 - sparse_categorical_accuracy: 0.9886 - scaled_adversarial_loss: 0.0699 433/1875 [=====>........................] - ETA: 10s - loss: 0.1023 - sparse_categorical_crossentropy: 0.0325 - sparse_categorical_accuracy: 0.9886 - scaled_adversarial_loss: 0.0698 440/1875 [======>.......................] - ETA: 10s - loss: 0.1022 - sparse_categorical_crossentropy: 0.0325 - sparse_categorical_accuracy: 0.9886 - scaled_adversarial_loss: 0.0697 447/1875 [======>.......................] - ETA: 10s - loss: 0.1021 - sparse_categorical_crossentropy: 0.0324 - sparse_categorical_accuracy: 0.9887 - scaled_adversarial_loss: 0.0697 455/1875 [======>.......................] - ETA: 10s - loss: 0.1019 - sparse_categorical_crossentropy: 0.0323 - sparse_categorical_accuracy: 0.9887 - scaled_adversarial_loss: 0.0696 462/1875 [======>.......................] - ETA: 10s - loss: 0.1018 - sparse_categorical_crossentropy: 0.0323 - sparse_categorical_accuracy: 0.9887 - scaled_adversarial_loss: 0.0695 469/1875 [======>.......................] - ETA: 10s - loss: 0.1018 - sparse_categorical_crossentropy: 0.0323 - sparse_categorical_accuracy: 0.9887 - scaled_adversarial_loss: 0.0695 477/1875 [======>.......................] - ETA: 10s - loss: 0.1016 - sparse_categorical_crossentropy: 0.0322 - sparse_categorical_accuracy: 0.9888 - scaled_adversarial_loss: 0.0694 484/1875 [======>.......................] - ETA: 10s - loss: 0.1016 - sparse_categorical_crossentropy: 0.0322 - sparse_categorical_accuracy: 0.9888 - scaled_adversarial_loss: 0.0694 492/1875 [======>.......................] - ETA: 10s - loss: 0.1015 - sparse_categorical_crossentropy: 0.0322 - sparse_categorical_accuracy: 0.9888 - scaled_adversarial_loss: 0.0693 499/1875 [======>.......................] - ETA: 10s - loss: 0.1014 - sparse_categorical_crossentropy: 0.0321 - sparse_categorical_accuracy: 0.9888 - scaled_adversarial_loss: 0.0693 507/1875 [=======>......................] - ETA: 9s - loss: 0.1013 - sparse_categorical_crossentropy: 0.0321 - sparse_categorical_accuracy: 0.9888 - scaled_adversarial_loss: 0.0692 514/1875 [=======>......................] - ETA: 9s - loss: 0.1013 - sparse_categorical_crossentropy: 0.0321 - sparse_categorical_accuracy: 0.9888 - scaled_adversarial_loss: 0.0692 522/1875 [=======>......................] - ETA: 9s - loss: 0.1012 - sparse_categorical_crossentropy: 0.0321 - sparse_categorical_accuracy: 0.9888 - scaled_adversarial_loss: 0.0691 529/1875 [=======>......................] - ETA: 9s - loss: 0.1011 - sparse_categorical_crossentropy: 0.0320 - sparse_categorical_accuracy: 0.9888 - scaled_adversarial_loss: 0.0691 537/1875 [=======>......................] - ETA: 9s - loss: 0.1010 - sparse_categorical_crossentropy: 0.0320 - sparse_categorical_accuracy: 0.9888 - scaled_adversarial_loss: 0.0690 544/1875 [=======>......................] - ETA: 9s - loss: 0.1010 - sparse_categorical_crossentropy: 0.0320 - sparse_categorical_accuracy: 0.9888 - scaled_adversarial_loss: 0.0690 551/1875 [=======>......................] - ETA: 9s - loss: 0.1009 - sparse_categorical_crossentropy: 0.0319 - sparse_categorical_accuracy: 0.9889 - scaled_adversarial_loss: 0.0690 558/1875 [=======>......................] - ETA: 9s - loss: 0.1008 - sparse_categorical_crossentropy: 0.0319 - sparse_categorical_accuracy: 0.9889 - scaled_adversarial_loss: 0.0689 566/1875 [========>.....................] - ETA: 9s - loss: 0.1008 - sparse_categorical_crossentropy: 0.0319 - sparse_categorical_accuracy: 0.9889 - scaled_adversarial_loss: 0.0689 573/1875 [========>.....................] - ETA: 9s - loss: 0.1007 - sparse_categorical_crossentropy: 0.0318 - sparse_categorical_accuracy: 0.9889 - scaled_adversarial_loss: 0.0689 580/1875 [========>.....................] - ETA: 9s - loss: 0.1006 - sparse_categorical_crossentropy: 0.0318 - sparse_categorical_accuracy: 0.9889 - scaled_adversarial_loss: 0.0688 587/1875 [========>.....................] - ETA: 9s - loss: 0.1006 - sparse_categorical_crossentropy: 0.0318 - sparse_categorical_accuracy: 0.9889 - scaled_adversarial_loss: 0.0688 594/1875 [========>.....................] - ETA: 9s - loss: 0.1005 - sparse_categorical_crossentropy: 0.0317 - sparse_categorical_accuracy: 0.9889 - scaled_adversarial_loss: 0.0688 601/1875 [========>.....................] - ETA: 9s - loss: 0.1005 - sparse_categorical_crossentropy: 0.0317 - sparse_categorical_accuracy: 0.9889 - scaled_adversarial_loss: 0.0688 609/1875 [========>.....................] - ETA: 9s - loss: 0.1004 - sparse_categorical_crossentropy: 0.0317 - sparse_categorical_accuracy: 0.9889 - scaled_adversarial_loss: 0.0687 616/1875 [========>.....................] - ETA: 9s - loss: 0.1003 - sparse_categorical_crossentropy: 0.0317 - sparse_categorical_accuracy: 0.9889 - scaled_adversarial_loss: 0.0687 623/1875 [========>.....................] - ETA: 9s - loss: 0.1003 - sparse_categorical_crossentropy: 0.0316 - sparse_categorical_accuracy: 0.9890 - scaled_adversarial_loss: 0.0687 630/1875 [=========>....................] - ETA: 9s - loss: 0.1002 - sparse_categorical_crossentropy: 0.0316 - sparse_categorical_accuracy: 0.9890 - scaled_adversarial_loss: 0.0686 637/1875 [=========>....................] - ETA: 9s - loss: 0.1001 - sparse_categorical_crossentropy: 0.0316 - sparse_categorical_accuracy: 0.9890 - scaled_adversarial_loss: 0.0686 645/1875 [=========>....................] - ETA: 8s - loss: 0.1001 - sparse_categorical_crossentropy: 0.0315 - sparse_categorical_accuracy: 0.9890 - scaled_adversarial_loss: 0.0685 653/1875 [=========>....................] - ETA: 8s - loss: 0.1000 - sparse_categorical_crossentropy: 0.0315 - sparse_categorical_accuracy: 0.9890 - scaled_adversarial_loss: 0.0685 661/1875 [=========>....................] - ETA: 8s - loss: 0.0999 - sparse_categorical_crossentropy: 0.0315 - sparse_categorical_accuracy: 0.9890 - scaled_adversarial_loss: 0.0685 669/1875 [=========>....................] - ETA: 8s - loss: 0.0998 - sparse_categorical_crossentropy: 0.0314 - sparse_categorical_accuracy: 0.9890 - scaled_adversarial_loss: 0.0684 676/1875 [=========>....................] - ETA: 8s - loss: 0.0998 - sparse_categorical_crossentropy: 0.0314 - sparse_categorical_accuracy: 0.9890 - scaled_adversarial_loss: 0.0684 684/1875 [=========>....................] - ETA: 8s - loss: 0.0997 - sparse_categorical_crossentropy: 0.0314 - sparse_categorical_accuracy: 0.9891 - scaled_adversarial_loss: 0.0683 692/1875 [==========>...................] - ETA: 8s - loss: 0.0996 - sparse_categorical_crossentropy: 0.0313 - sparse_categorical_accuracy: 0.9891 - scaled_adversarial_loss: 0.0683 699/1875 [==========>...................] - ETA: 8s - loss: 0.0995 - sparse_categorical_crossentropy: 0.0313 - sparse_categorical_accuracy: 0.9891 - scaled_adversarial_loss: 0.0682 706/1875 [==========>...................] - ETA: 8s - loss: 0.0995 - sparse_categorical_crossentropy: 0.0313 - sparse_categorical_accuracy: 0.9891 - scaled_adversarial_loss: 0.0682 713/1875 [==========>...................] - ETA: 8s - loss: 0.0994 - sparse_categorical_crossentropy: 0.0312 - sparse_categorical_accuracy: 0.9891 - scaled_adversarial_loss: 0.0682 721/1875 [==========>...................] - ETA: 8s - loss: 0.0993 - sparse_categorical_crossentropy: 0.0312 - sparse_categorical_accuracy: 0.9891 - scaled_adversarial_loss: 0.0681 729/1875 [==========>...................] - ETA: 8s - loss: 0.0992 - sparse_categorical_crossentropy: 0.0312 - sparse_categorical_accuracy: 0.9891 - scaled_adversarial_loss: 0.0681 736/1875 [==========>...................] - ETA: 8s - loss: 0.0992 - sparse_categorical_crossentropy: 0.0311 - sparse_categorical_accuracy: 0.9891 - scaled_adversarial_loss: 0.0680 743/1875 [==========>...................] - ETA: 8s - loss: 0.0991 - sparse_categorical_crossentropy: 0.0311 - sparse_categorical_accuracy: 0.9891 - scaled_adversarial_loss: 0.0680 751/1875 [===========>..................] - ETA: 8s - loss: 0.0990 - sparse_categorical_crossentropy: 0.0311 - sparse_categorical_accuracy: 0.9892 - scaled_adversarial_loss: 0.0680 758/1875 [===========>..................] - ETA: 8s - loss: 0.0990 - sparse_categorical_crossentropy: 0.0310 - sparse_categorical_accuracy: 0.9892 - scaled_adversarial_loss: 0.0679 765/1875 [===========>..................] - ETA: 8s - loss: 0.0989 - sparse_categorical_crossentropy: 0.0310 - sparse_categorical_accuracy: 0.9892 - scaled_adversarial_loss: 0.0679 772/1875 [===========>..................] - ETA: 8s - loss: 0.0988 - sparse_categorical_crossentropy: 0.0310 - sparse_categorical_accuracy: 0.9892 - scaled_adversarial_loss: 0.0679 780/1875 [===========>..................] - ETA: 7s - loss: 0.0988 - sparse_categorical_crossentropy: 0.0310 - sparse_categorical_accuracy: 0.9892 - scaled_adversarial_loss: 0.0678 788/1875 [===========>..................] - ETA: 7s - loss: 0.0987 - sparse_categorical_crossentropy: 0.0309 - sparse_categorical_accuracy: 0.9892 - scaled_adversarial_loss: 0.0678 796/1875 [===========>..................] - ETA: 7s - loss: 0.0986 - sparse_categorical_crossentropy: 0.0309 - sparse_categorical_accuracy: 0.9892 - scaled_adversarial_loss: 0.0677 803/1875 [===========>..................] - ETA: 7s - loss: 0.0986 - sparse_categorical_crossentropy: 0.0309 - sparse_categorical_accuracy: 0.9892 - scaled_adversarial_loss: 0.0677 810/1875 [===========>..................] - ETA: 7s - loss: 0.0985 - sparse_categorical_crossentropy: 0.0309 - sparse_categorical_accuracy: 0.9892 - scaled_adversarial_loss: 0.0677 818/1875 [============>.................] - ETA: 7s - loss: 0.0985 - sparse_categorical_crossentropy: 0.0308 - sparse_categorical_accuracy: 0.9892 - scaled_adversarial_loss: 0.0676 825/1875 [============>.................] - ETA: 7s - loss: 0.0984 - sparse_categorical_crossentropy: 0.0308 - sparse_categorical_accuracy: 0.9892 - scaled_adversarial_loss: 0.0676 832/1875 [============>.................] - ETA: 7s - loss: 0.0984 - sparse_categorical_crossentropy: 0.0308 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0676 840/1875 [============>.................] - ETA: 7s - loss: 0.0983 - sparse_categorical_crossentropy: 0.0308 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0675 847/1875 [============>.................] - ETA: 7s - loss: 0.0983 - sparse_categorical_crossentropy: 0.0308 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0675 854/1875 [============>.................] - ETA: 7s - loss: 0.0982 - sparse_categorical_crossentropy: 0.0308 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0675 861/1875 [============>.................] - ETA: 7s - loss: 0.0982 - sparse_categorical_crossentropy: 0.0308 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0674 869/1875 [============>.................] - ETA: 7s - loss: 0.0982 - sparse_categorical_crossentropy: 0.0308 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0674 876/1875 [=============>................] - ETA: 7s - loss: 0.0981 - sparse_categorical_crossentropy: 0.0307 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0674 884/1875 [=============>................] - ETA: 7s - loss: 0.0981 - sparse_categorical_crossentropy: 0.0307 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0674 891/1875 [=============>................] - ETA: 7s - loss: 0.0980 - sparse_categorical_crossentropy: 0.0307 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0673 899/1875 [=============>................] - ETA: 7s - loss: 0.0980 - sparse_categorical_crossentropy: 0.0307 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0673 907/1875 [=============>................] - ETA: 7s - loss: 0.0980 - sparse_categorical_crossentropy: 0.0307 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0673 915/1875 [=============>................] - ETA: 6s - loss: 0.0979 - sparse_categorical_crossentropy: 0.0307 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0672 922/1875 [=============>................] - ETA: 6s - loss: 0.0979 - sparse_categorical_crossentropy: 0.0307 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0672 930/1875 [=============>................] - ETA: 6s - loss: 0.0979 - sparse_categorical_crossentropy: 0.0307 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0672 938/1875 [==============>...............] - ETA: 6s - loss: 0.0978 - sparse_categorical_crossentropy: 0.0307 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0672 945/1875 [==============>...............] - ETA: 6s - loss: 0.0978 - sparse_categorical_crossentropy: 0.0307 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0671 952/1875 [==============>...............] - ETA: 6s - loss: 0.0978 - sparse_categorical_crossentropy: 0.0306 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0671 959/1875 [==============>...............] - ETA: 6s - loss: 0.0977 - sparse_categorical_crossentropy: 0.0306 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0671 967/1875 [==============>...............] - ETA: 6s - loss: 0.0977 - sparse_categorical_crossentropy: 0.0306 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0671 974/1875 [==============>...............] - ETA: 6s - loss: 0.0977 - sparse_categorical_crossentropy: 0.0306 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0670 982/1875 [==============>...............] - ETA: 6s - loss: 0.0976 - sparse_categorical_crossentropy: 0.0306 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0670 990/1875 [==============>...............] - ETA: 6s - loss: 0.0976 - sparse_categorical_crossentropy: 0.0306 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0670 998/1875 [==============>...............] - ETA: 6s - loss: 0.0976 - sparse_categorical_crossentropy: 0.0306 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0670 1006/1875 [===============>..............] - ETA: 6s - loss: 0.0975 - sparse_categorical_crossentropy: 0.0306 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0669 1013/1875 [===============>..............] - ETA: 6s - loss: 0.0975 - sparse_categorical_crossentropy: 0.0306 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0669 1020/1875 [===============>..............] - ETA: 6s - loss: 0.0975 - sparse_categorical_crossentropy: 0.0306 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0669 1028/1875 [===============>..............] - ETA: 6s - loss: 0.0975 - sparse_categorical_crossentropy: 0.0306 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0669 1035/1875 [===============>..............] - ETA: 6s - loss: 0.0974 - sparse_categorical_crossentropy: 0.0306 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0668 1042/1875 [===============>..............] - ETA: 6s - loss: 0.0974 - sparse_categorical_crossentropy: 0.0306 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0668 1050/1875 [===============>..............] - ETA: 5s - loss: 0.0974 - sparse_categorical_crossentropy: 0.0306 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0668 1058/1875 [===============>..............] - ETA: 5s - loss: 0.0973 - sparse_categorical_crossentropy: 0.0306 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0668 1065/1875 [================>.............] - ETA: 5s - loss: 0.0973 - sparse_categorical_crossentropy: 0.0306 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0667 1073/1875 [================>.............] - ETA: 5s - loss: 0.0973 - sparse_categorical_crossentropy: 0.0306 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0667 1081/1875 [================>.............] - ETA: 5s - loss: 0.0972 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0667 1088/1875 [================>.............] - ETA: 5s - loss: 0.0972 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0667 1095/1875 [================>.............] - ETA: 5s - loss: 0.0972 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0667 1102/1875 [================>.............] - ETA: 5s - loss: 0.0972 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0666 1109/1875 [================>.............] - ETA: 5s - loss: 0.0971 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0666 1117/1875 [================>.............] - ETA: 5s - loss: 0.0971 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0666 1124/1875 [================>.............] - ETA: 5s - loss: 0.0971 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0666 1131/1875 [=================>............] - ETA: 5s - loss: 0.0970 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0665 1138/1875 [=================>............] - ETA: 5s - loss: 0.0970 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0665 1145/1875 [=================>............] - ETA: 5s - loss: 0.0970 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0665 1152/1875 [=================>............] - ETA: 5s - loss: 0.0970 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0665 1159/1875 [=================>............] - ETA: 5s - loss: 0.0969 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0665 1166/1875 [=================>............] - ETA: 5s - loss: 0.0969 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0664 1174/1875 [=================>............] - ETA: 5s - loss: 0.0969 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0664 1182/1875 [=================>............] - ETA: 5s - loss: 0.0969 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0664 1189/1875 [==================>...........] - ETA: 4s - loss: 0.0968 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0664 1196/1875 [==================>...........] - ETA: 4s - loss: 0.0968 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0664 1203/1875 [==================>...........] - ETA: 4s - loss: 0.0968 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0663 1211/1875 [==================>...........] - ETA: 4s - loss: 0.0968 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0663 1218/1875 [==================>...........] - ETA: 4s - loss: 0.0968 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0663 1225/1875 [==================>...........] - ETA: 4s - loss: 0.0967 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0663 1232/1875 [==================>...........] - ETA: 4s - loss: 0.0967 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0663 1239/1875 [==================>...........] - ETA: 4s - loss: 0.0967 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0663 1246/1875 [==================>...........] - ETA: 4s - loss: 0.0967 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0663 1253/1875 [===================>..........] - ETA: 4s - loss: 0.0967 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0662 1260/1875 [===================>..........] - ETA: 4s - loss: 0.0966 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0662 1267/1875 [===================>..........] - ETA: 4s - loss: 0.0966 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0662 1274/1875 [===================>..........] - ETA: 4s - loss: 0.0966 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0662 1281/1875 [===================>..........] - ETA: 4s - loss: 0.0966 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0662 1289/1875 [===================>..........] - ETA: 4s - loss: 0.0966 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0662 1297/1875 [===================>..........] - ETA: 4s - loss: 0.0965 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0661 1304/1875 [===================>..........] - ETA: 4s - loss: 0.0965 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0661 1311/1875 [===================>..........] - ETA: 4s - loss: 0.0965 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0661 1319/1875 [====================>.........] - ETA: 4s - loss: 0.0965 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0661 1326/1875 [====================>.........] - ETA: 3s - loss: 0.0964 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0661 1334/1875 [====================>.........] - ETA: 3s - loss: 0.0964 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0661 1341/1875 [====================>.........] - ETA: 3s - loss: 0.0964 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0660 1349/1875 [====================>.........] - ETA: 3s - loss: 0.0964 - sparse_categorical_crossentropy: 0.0304 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0660 1356/1875 [====================>.........] - ETA: 3s - loss: 0.0964 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0660 1363/1875 [====================>.........] - ETA: 3s - loss: 0.0963 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0660 1370/1875 [====================>.........] - ETA: 3s - loss: 0.0963 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0660 1377/1875 [=====================>........] - ETA: 3s - loss: 0.0963 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0660 1385/1875 [=====================>........] - ETA: 3s - loss: 0.0963 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0660 1393/1875 [=====================>........] - ETA: 3s - loss: 0.0963 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0659 1400/1875 [=====================>........] - ETA: 3s - loss: 0.0963 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0659 1408/1875 [=====================>........] - ETA: 3s - loss: 0.0962 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0659 1415/1875 [=====================>........] - ETA: 3s - loss: 0.0962 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0659 1422/1875 [=====================>........] - ETA: 3s - loss: 0.0962 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0659 1429/1875 [=====================>........] - ETA: 3s - loss: 0.0962 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0659 1437/1875 [=====================>........] - ETA: 3s - loss: 0.0962 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0659 1444/1875 [======================>.......] - ETA: 3s - loss: 0.0961 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0658 1451/1875 [======================>.......] - ETA: 3s - loss: 0.0961 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0658 1458/1875 [======================>.......] - ETA: 3s - loss: 0.0961 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0658 1465/1875 [======================>.......] - ETA: 2s - loss: 0.0961 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0658 1473/1875 [======================>.......] - ETA: 2s - loss: 0.0961 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0658 1480/1875 [======================>.......] - ETA: 2s - loss: 0.0961 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0658 1488/1875 [======================>.......] - ETA: 2s - loss: 0.0960 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0658 1496/1875 [======================>.......] - ETA: 2s - loss: 0.0960 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0657 1504/1875 [=======================>......] - ETA: 2s - loss: 0.0960 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0657 1511/1875 [=======================>......] - ETA: 2s - loss: 0.0960 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0657 1518/1875 [=======================>......] - ETA: 2s - loss: 0.0960 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0657 1525/1875 [=======================>......] - ETA: 2s - loss: 0.0959 - sparse_categorical_crossentropy: 0.0303 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0657 1532/1875 [=======================>......] - ETA: 2s - loss: 0.0959 - sparse_categorical_crossentropy: 0.0302 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0657 1539/1875 [=======================>......] - ETA: 2s - loss: 0.0959 - sparse_categorical_crossentropy: 0.0302 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0657 1546/1875 [=======================>......] - ETA: 2s - loss: 0.0959 - sparse_categorical_crossentropy: 0.0302 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0656 1554/1875 [=======================>......] - ETA: 2s - loss: 0.0959 - sparse_categorical_crossentropy: 0.0302 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0656 1562/1875 [=======================>......] - ETA: 2s - loss: 0.0958 - sparse_categorical_crossentropy: 0.0302 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0656 1571/1875 [========================>.....] - ETA: 2s - loss: 0.0958 - sparse_categorical_crossentropy: 0.0302 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0656 1580/1875 [========================>.....] - ETA: 2s - loss: 0.0958 - sparse_categorical_crossentropy: 0.0302 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0656 1590/1875 [========================>.....] - ETA: 2s - loss: 0.0958 - sparse_categorical_crossentropy: 0.0302 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0656 1600/1875 [========================>.....] - ETA: 1s - loss: 0.0957 - sparse_categorical_crossentropy: 0.0302 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0655 1610/1875 [========================>.....] - ETA: 1s - loss: 0.0957 - sparse_categorical_crossentropy: 0.0302 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0655 1619/1875 [========================>.....] - ETA: 1s - loss: 0.0957 - sparse_categorical_crossentropy: 0.0302 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0655 1629/1875 [=========================>....] - ETA: 1s - loss: 0.0957 - sparse_categorical_crossentropy: 0.0302 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0655 1639/1875 [=========================>....] - ETA: 1s - loss: 0.0956 - sparse_categorical_crossentropy: 0.0302 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0655 1649/1875 [=========================>....] - ETA: 1s - loss: 0.0956 - sparse_categorical_crossentropy: 0.0302 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0654 1658/1875 [=========================>....] - ETA: 1s - loss: 0.0956 - sparse_categorical_crossentropy: 0.0302 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0654 1668/1875 [=========================>....] - ETA: 1s - loss: 0.0956 - sparse_categorical_crossentropy: 0.0302 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0654 1677/1875 [=========================>....] - ETA: 1s - loss: 0.0955 - sparse_categorical_crossentropy: 0.0301 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0654 1686/1875 [=========================>....] - ETA: 1s - loss: 0.0955 - sparse_categorical_crossentropy: 0.0301 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0654 1695/1875 [==========================>...] - ETA: 1s - loss: 0.0955 - sparse_categorical_crossentropy: 0.0301 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0653 1705/1875 [==========================>...] - ETA: 1s - loss: 0.0954 - sparse_categorical_crossentropy: 0.0301 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0653 1715/1875 [==========================>...] - ETA: 1s - loss: 0.0954 - sparse_categorical_crossentropy: 0.0301 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0653 1724/1875 [==========================>...] - ETA: 1s - loss: 0.0954 - sparse_categorical_crossentropy: 0.0301 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0653 1733/1875 [==========================>...] - ETA: 1s - loss: 0.0954 - sparse_categorical_crossentropy: 0.0301 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0653 1742/1875 [==========================>...] - ETA: 0s - loss: 0.0953 - sparse_categorical_crossentropy: 0.0301 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0653 1751/1875 [===========================>..] - ETA: 0s - loss: 0.0953 - sparse_categorical_crossentropy: 0.0301 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0652 1761/1875 [===========================>..] - ETA: 0s - loss: 0.0953 - sparse_categorical_crossentropy: 0.0301 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0652 1770/1875 [===========================>..] - ETA: 0s - loss: 0.0953 - sparse_categorical_crossentropy: 0.0301 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0652 1779/1875 [===========================>..] - ETA: 0s - loss: 0.0952 - sparse_categorical_crossentropy: 0.0301 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0652 1788/1875 [===========================>..] - ETA: 0s - loss: 0.0952 - sparse_categorical_crossentropy: 0.0300 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0652 1798/1875 [===========================>..] - ETA: 0s - loss: 0.0952 - sparse_categorical_crossentropy: 0.0300 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0651 1807/1875 [===========================>..] - ETA: 0s - loss: 0.0952 - sparse_categorical_crossentropy: 0.0300 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0651 1816/1875 [============================>.] - ETA: 0s - loss: 0.0951 - sparse_categorical_crossentropy: 0.0300 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0651 1825/1875 [============================>.] - ETA: 0s - loss: 0.0951 - sparse_categorical_crossentropy: 0.0300 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0651 1834/1875 [============================>.] - ETA: 0s - loss: 0.0951 - sparse_categorical_crossentropy: 0.0300 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0651 1843/1875 [============================>.] - ETA: 0s - loss: 0.0950 - sparse_categorical_crossentropy: 0.0300 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0651 1852/1875 [============================>.] - ETA: 0s - loss: 0.0950 - sparse_categorical_crossentropy: 0.0300 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0650 1861/1875 [============================>.] - ETA: 0s - loss: 0.0950 - sparse_categorical_crossentropy: 0.0300 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0650 1870/1875 [============================>.] - ETA: 0s - loss: 0.0950 - sparse_categorical_crossentropy: 0.0300 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0650 1875/1875 [==============================] - 14s 7ms/step - loss: 0.0950 - sparse_categorical_crossentropy: 0.0300 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0650 Epoch 4/5 1/1875 [..............................] - ETA: 12:07 - loss: 0.0602 - sparse_categorical_crossentropy: 0.0067 - sparse_categorical_accuracy: 1.0000 - scaled_adversarial_loss: 0.0535 8/1875 [..............................] - ETA: 14s - loss: 0.0721 - sparse_categorical_crossentropy: 0.0139 - sparse_categorical_accuracy: 0.9941 - scaled_adversarial_loss: 0.0583 15/1875 [..............................] - ETA: 13s - loss: 0.0811 - sparse_categorical_crossentropy: 0.0197 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0614 23/1875 [..............................] - ETA: 13s - loss: 0.0819 - sparse_categorical_crossentropy: 0.0214 - sparse_categorical_accuracy: 0.9888 - scaled_adversarial_loss: 0.0605 31/1875 [..............................] - ETA: 13s - loss: 0.0828 - sparse_categorical_crossentropy: 0.0228 - sparse_categorical_accuracy: 0.9887 - scaled_adversarial_loss: 0.0600 38/1875 [..............................] - ETA: 13s - loss: 0.0829 - sparse_categorical_crossentropy: 0.0229 - sparse_categorical_accuracy: 0.9892 - scaled_adversarial_loss: 0.0600 45/1875 [..............................] - ETA: 13s - loss: 0.0829 - sparse_categorical_crossentropy: 0.0230 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0599 52/1875 [..............................] - ETA: 13s - loss: 0.0824 - sparse_categorical_crossentropy: 0.0227 - sparse_categorical_accuracy: 0.9900 - scaled_adversarial_loss: 0.0598 59/1875 [..............................] - ETA: 13s - loss: 0.0819 - sparse_categorical_crossentropy: 0.0223 - sparse_categorical_accuracy: 0.9904 - scaled_adversarial_loss: 0.0596 66/1875 [>.............................] - ETA: 13s - loss: 0.0816 - sparse_categorical_crossentropy: 0.0222 - sparse_categorical_accuracy: 0.9907 - scaled_adversarial_loss: 0.0594 73/1875 [>.............................] - ETA: 13s - loss: 0.0815 - sparse_categorical_crossentropy: 0.0222 - sparse_categorical_accuracy: 0.9908 - scaled_adversarial_loss: 0.0593 81/1875 [>.............................] - ETA: 13s - loss: 0.0815 - sparse_categorical_crossentropy: 0.0223 - sparse_categorical_accuracy: 0.9910 - scaled_adversarial_loss: 0.0592 88/1875 [>.............................] - ETA: 13s - loss: 0.0814 - sparse_categorical_crossentropy: 0.0224 - sparse_categorical_accuracy: 0.9911 - scaled_adversarial_loss: 0.0590 95/1875 [>.............................] - ETA: 13s - loss: 0.0812 - sparse_categorical_crossentropy: 0.0224 - sparse_categorical_accuracy: 0.9912 - scaled_adversarial_loss: 0.0589 102/1875 [>.............................] - ETA: 13s - loss: 0.0811 - sparse_categorical_crossentropy: 0.0224 - sparse_categorical_accuracy: 0.9914 - scaled_adversarial_loss: 0.0587 110/1875 [>.............................] - ETA: 12s - loss: 0.0810 - sparse_categorical_crossentropy: 0.0225 - sparse_categorical_accuracy: 0.9914 - scaled_adversarial_loss: 0.0585 118/1875 [>.............................] - ETA: 12s - loss: 0.0810 - sparse_categorical_crossentropy: 0.0226 - sparse_categorical_accuracy: 0.9915 - scaled_adversarial_loss: 0.0584 125/1875 [=>............................] - ETA: 12s - loss: 0.0810 - sparse_categorical_crossentropy: 0.0227 - sparse_categorical_accuracy: 0.9916 - scaled_adversarial_loss: 0.0583 132/1875 [=>............................] - ETA: 12s - loss: 0.0809 - sparse_categorical_crossentropy: 0.0227 - sparse_categorical_accuracy: 0.9916 - scaled_adversarial_loss: 0.0582 140/1875 [=>............................] - ETA: 12s - loss: 0.0809 - sparse_categorical_crossentropy: 0.0228 - sparse_categorical_accuracy: 0.9917 - scaled_adversarial_loss: 0.0580 148/1875 [=>............................] - ETA: 12s - loss: 0.0808 - sparse_categorical_crossentropy: 0.0229 - sparse_categorical_accuracy: 0.9918 - scaled_adversarial_loss: 0.0579 155/1875 [=>............................] - ETA: 12s - loss: 0.0808 - sparse_categorical_crossentropy: 0.0229 - sparse_categorical_accuracy: 0.9918 - scaled_adversarial_loss: 0.0578 162/1875 [=>............................] - ETA: 12s - loss: 0.0807 - sparse_categorical_crossentropy: 0.0230 - sparse_categorical_accuracy: 0.9919 - scaled_adversarial_loss: 0.0577 169/1875 [=>............................] - ETA: 12s - loss: 0.0807 - sparse_categorical_crossentropy: 0.0230 - sparse_categorical_accuracy: 0.9920 - scaled_adversarial_loss: 0.0577 177/1875 [=>............................] - ETA: 12s - loss: 0.0806 - sparse_categorical_crossentropy: 0.0230 - sparse_categorical_accuracy: 0.9920 - scaled_adversarial_loss: 0.0576 184/1875 [=>............................] - ETA: 12s - loss: 0.0805 - sparse_categorical_crossentropy: 0.0230 - sparse_categorical_accuracy: 0.9920 - scaled_adversarial_loss: 0.0575 191/1875 [==>...........................] - ETA: 12s - loss: 0.0804 - sparse_categorical_crossentropy: 0.0230 - sparse_categorical_accuracy: 0.9921 - scaled_adversarial_loss: 0.0573 198/1875 [==>...........................] - ETA: 12s - loss: 0.0803 - sparse_categorical_crossentropy: 0.0231 - sparse_categorical_accuracy: 0.9921 - scaled_adversarial_loss: 0.0572 206/1875 [==>...........................] - ETA: 12s - loss: 0.0801 - sparse_categorical_crossentropy: 0.0230 - sparse_categorical_accuracy: 0.9922 - scaled_adversarial_loss: 0.0570 213/1875 [==>...........................] - ETA: 12s - loss: 0.0799 - sparse_categorical_crossentropy: 0.0230 - sparse_categorical_accuracy: 0.9922 - scaled_adversarial_loss: 0.0569 220/1875 [==>...........................] - ETA: 11s - loss: 0.0797 - sparse_categorical_crossentropy: 0.0230 - sparse_categorical_accuracy: 0.9922 - scaled_adversarial_loss: 0.0567 227/1875 [==>...........................] - ETA: 11s - loss: 0.0795 - sparse_categorical_crossentropy: 0.0229 - sparse_categorical_accuracy: 0.9923 - scaled_adversarial_loss: 0.0565 235/1875 [==>...........................] - ETA: 11s - loss: 0.0792 - sparse_categorical_crossentropy: 0.0229 - sparse_categorical_accuracy: 0.9923 - scaled_adversarial_loss: 0.0564 242/1875 [==>...........................] - ETA: 11s - loss: 0.0791 - sparse_categorical_crossentropy: 0.0228 - sparse_categorical_accuracy: 0.9924 - scaled_adversarial_loss: 0.0562 249/1875 [==>...........................] - ETA: 11s - loss: 0.0789 - sparse_categorical_crossentropy: 0.0228 - sparse_categorical_accuracy: 0.9924 - scaled_adversarial_loss: 0.0561 256/1875 [===>..........................] - ETA: 11s - loss: 0.0787 - sparse_categorical_crossentropy: 0.0228 - sparse_categorical_accuracy: 0.9924 - scaled_adversarial_loss: 0.0559 263/1875 [===>..........................] - ETA: 11s - loss: 0.0785 - sparse_categorical_crossentropy: 0.0227 - sparse_categorical_accuracy: 0.9924 - scaled_adversarial_loss: 0.0558 270/1875 [===>..........................] - ETA: 11s - loss: 0.0784 - sparse_categorical_crossentropy: 0.0227 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0557 278/1875 [===>..........................] - ETA: 11s - loss: 0.0783 - sparse_categorical_crossentropy: 0.0227 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0556 285/1875 [===>..........................] - ETA: 11s - loss: 0.0781 - sparse_categorical_crossentropy: 0.0227 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0555 293/1875 [===>..........................] - ETA: 11s - loss: 0.0780 - sparse_categorical_crossentropy: 0.0226 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0554 300/1875 [===>..........................] - ETA: 11s - loss: 0.0779 - sparse_categorical_crossentropy: 0.0226 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0553 307/1875 [===>..........................] - ETA: 11s - loss: 0.0779 - sparse_categorical_crossentropy: 0.0226 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0552 314/1875 [====>.........................] - ETA: 11s - loss: 0.0778 - sparse_categorical_crossentropy: 0.0226 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0552 322/1875 [====>.........................] - ETA: 11s - loss: 0.0777 - sparse_categorical_crossentropy: 0.0226 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0551 330/1875 [====>.........................] - ETA: 11s - loss: 0.0777 - sparse_categorical_crossentropy: 0.0226 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0550 337/1875 [====>.........................] - ETA: 11s - loss: 0.0776 - sparse_categorical_crossentropy: 0.0226 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0550 344/1875 [====>.........................] - ETA: 11s - loss: 0.0776 - sparse_categorical_crossentropy: 0.0226 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0549 351/1875 [====>.........................] - ETA: 11s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0227 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0549 358/1875 [====>.........................] - ETA: 11s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0227 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0548 365/1875 [====>.........................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0227 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0548 372/1875 [====>.........................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0228 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0548 379/1875 [=====>........................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0228 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0548 386/1875 [=====>........................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0228 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0547 393/1875 [=====>........................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0228 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0547 400/1875 [=====>........................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0229 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0547 407/1875 [=====>........................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0229 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0547 414/1875 [=====>........................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0229 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0546 421/1875 [=====>........................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0229 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0546 428/1875 [=====>........................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0229 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0546 436/1875 [=====>........................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0230 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0546 444/1875 [======>.......................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0230 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0546 451/1875 [======>.......................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0230 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0545 458/1875 [======>.......................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0230 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0545 465/1875 [======>.......................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0230 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0545 473/1875 [======>.......................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0230 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0545 481/1875 [======>.......................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0230 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0545 488/1875 [======>.......................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0231 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0544 496/1875 [======>.......................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0231 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0544 503/1875 [=======>......................] - ETA: 10s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0231 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0544 510/1875 [=======>......................] - ETA: 9s - loss: 0.0775 - sparse_categorical_crossentropy: 0.0231 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0544 518/1875 [=======>......................] - ETA: 9s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0231 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0544 526/1875 [=======>......................] - ETA: 9s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0231 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0543 533/1875 [=======>......................] - ETA: 9s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0231 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0543 541/1875 [=======>......................] - ETA: 9s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0231 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0543 549/1875 [=======>......................] - ETA: 9s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0231 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0543 557/1875 [=======>......................] - ETA: 9s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0231 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0543 564/1875 [========>.....................] - ETA: 9s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0231 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0542 571/1875 [========>.....................] - ETA: 9s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0231 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0542 578/1875 [========>.....................] - ETA: 9s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0231 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0542 586/1875 [========>.....................] - ETA: 9s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0231 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0542 593/1875 [========>.....................] - ETA: 9s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0231 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0542 600/1875 [========>.....................] - ETA: 9s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0232 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0542 607/1875 [========>.....................] - ETA: 9s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0232 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0542 614/1875 [========>.....................] - ETA: 9s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0232 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0541 621/1875 [========>.....................] - ETA: 9s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0232 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0541 628/1875 [=========>....................] - ETA: 9s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0232 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0541 635/1875 [=========>....................] - ETA: 9s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0232 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0541 643/1875 [=========>....................] - ETA: 8s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0233 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0541 650/1875 [=========>....................] - ETA: 8s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0233 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0541 657/1875 [=========>....................] - ETA: 8s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0233 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0541 664/1875 [=========>....................] - ETA: 8s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0233 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0541 672/1875 [=========>....................] - ETA: 8s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0233 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0541 679/1875 [=========>....................] - ETA: 8s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0233 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0541 686/1875 [=========>....................] - ETA: 8s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0233 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0540 693/1875 [==========>...................] - ETA: 8s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0233 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0540 701/1875 [==========>...................] - ETA: 8s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0233 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0540 709/1875 [==========>...................] - ETA: 8s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0540 717/1875 [==========>...................] - ETA: 8s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0540 724/1875 [==========>...................] - ETA: 8s - loss: 0.0774 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0540 731/1875 [==========>...................] - ETA: 8s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0540 739/1875 [==========>...................] - ETA: 8s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0540 747/1875 [==========>...................] - ETA: 8s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0539 754/1875 [===========>..................] - ETA: 8s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0539 761/1875 [===========>..................] - ETA: 8s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0539 769/1875 [===========>..................] - ETA: 8s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0539 776/1875 [===========>..................] - ETA: 7s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0539 783/1875 [===========>..................] - ETA: 7s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0539 790/1875 [===========>..................] - ETA: 7s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0539 798/1875 [===========>..................] - ETA: 7s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0539 805/1875 [===========>..................] - ETA: 7s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0538 812/1875 [===========>..................] - ETA: 7s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0538 820/1875 [============>.................] - ETA: 7s - loss: 0.0773 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0538 828/1875 [============>.................] - ETA: 7s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0538 835/1875 [============>.................] - ETA: 7s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0538 842/1875 [============>.................] - ETA: 7s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0538 849/1875 [============>.................] - ETA: 7s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0538 856/1875 [============>.................] - ETA: 7s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0538 864/1875 [============>.................] - ETA: 7s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0538 872/1875 [============>.................] - ETA: 7s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0538 879/1875 [=============>................] - ETA: 7s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0537 887/1875 [=============>................] - ETA: 7s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0537 895/1875 [=============>................] - ETA: 7s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0537 902/1875 [=============>................] - ETA: 7s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0537 909/1875 [=============>................] - ETA: 7s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0537 917/1875 [=============>................] - ETA: 6s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0537 925/1875 [=============>................] - ETA: 6s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0537 932/1875 [=============>................] - ETA: 6s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0537 939/1875 [==============>...............] - ETA: 6s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0537 947/1875 [==============>...............] - ETA: 6s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0537 954/1875 [==============>...............] - ETA: 6s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0537 961/1875 [==============>...............] - ETA: 6s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0537 968/1875 [==============>...............] - ETA: 6s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0537 975/1875 [==============>...............] - ETA: 6s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0537 983/1875 [==============>...............] - ETA: 6s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0536 991/1875 [==============>...............] - ETA: 6s - loss: 0.0772 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0536 999/1875 [==============>...............] - ETA: 6s - loss: 0.0771 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0536 1007/1875 [===============>..............] - ETA: 6s - loss: 0.0771 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0536 1015/1875 [===============>..............] - ETA: 6s - loss: 0.0771 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0536 1022/1875 [===============>..............] - ETA: 6s - loss: 0.0771 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0536 1029/1875 [===============>..............] - ETA: 6s - loss: 0.0771 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0536 1037/1875 [===============>..............] - ETA: 6s - loss: 0.0771 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0536 1044/1875 [===============>..............] - ETA: 6s - loss: 0.0771 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0536 1051/1875 [===============>..............] - ETA: 5s - loss: 0.0771 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0536 1059/1875 [===============>..............] - ETA: 5s - loss: 0.0771 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0536 1066/1875 [================>.............] - ETA: 5s - loss: 0.0771 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0535 1074/1875 [================>.............] - ETA: 5s - loss: 0.0771 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0535 1081/1875 [================>.............] - ETA: 5s - loss: 0.0771 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0535 1089/1875 [================>.............] - ETA: 5s - loss: 0.0771 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0535 1097/1875 [================>.............] - ETA: 5s - loss: 0.0771 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0535 1105/1875 [================>.............] - ETA: 5s - loss: 0.0771 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0535 1113/1875 [================>.............] - ETA: 5s - loss: 0.0771 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0535 1120/1875 [================>.............] - ETA: 5s - loss: 0.0770 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0535 1128/1875 [=================>............] - ETA: 5s - loss: 0.0770 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0535 1136/1875 [=================>............] - ETA: 5s - loss: 0.0770 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0534 1143/1875 [=================>............] - ETA: 5s - loss: 0.0770 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0534 1150/1875 [=================>............] - ETA: 5s - loss: 0.0770 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0534 1158/1875 [=================>............] - ETA: 5s - loss: 0.0770 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0534 1166/1875 [=================>............] - ETA: 5s - loss: 0.0770 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0534 1173/1875 [=================>............] - ETA: 5s - loss: 0.0770 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0534 1180/1875 [=================>............] - ETA: 5s - loss: 0.0769 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0534 1188/1875 [==================>...........] - ETA: 4s - loss: 0.0769 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0534 1196/1875 [==================>...........] - ETA: 4s - loss: 0.0769 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0534 1203/1875 [==================>...........] - ETA: 4s - loss: 0.0769 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0533 1211/1875 [==================>...........] - ETA: 4s - loss: 0.0769 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0533 1219/1875 [==================>...........] - ETA: 4s - loss: 0.0769 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0533 1227/1875 [==================>...........] - ETA: 4s - loss: 0.0769 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0533 1234/1875 [==================>...........] - ETA: 4s - loss: 0.0769 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0533 1241/1875 [==================>...........] - ETA: 4s - loss: 0.0769 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0533 1249/1875 [==================>...........] - ETA: 4s - loss: 0.0769 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0533 1257/1875 [===================>..........] - ETA: 4s - loss: 0.0769 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0533 1265/1875 [===================>..........] - ETA: 4s - loss: 0.0769 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0533 1273/1875 [===================>..........] - ETA: 4s - loss: 0.0769 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0533 1280/1875 [===================>..........] - ETA: 4s - loss: 0.0769 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0533 1287/1875 [===================>..........] - ETA: 4s - loss: 0.0768 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0532 1294/1875 [===================>..........] - ETA: 4s - loss: 0.0768 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0532 1302/1875 [===================>..........] - ETA: 4s - loss: 0.0768 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0532 1310/1875 [===================>..........] - ETA: 4s - loss: 0.0768 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0532 1318/1875 [====================>.........] - ETA: 4s - loss: 0.0768 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0532 1325/1875 [====================>.........] - ETA: 3s - loss: 0.0768 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0532 1333/1875 [====================>.........] - ETA: 3s - loss: 0.0768 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0532 1341/1875 [====================>.........] - ETA: 3s - loss: 0.0768 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0532 1348/1875 [====================>.........] - ETA: 3s - loss: 0.0768 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0532 1355/1875 [====================>.........] - ETA: 3s - loss: 0.0768 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0532 1362/1875 [====================>.........] - ETA: 3s - loss: 0.0768 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0532 1369/1875 [====================>.........] - ETA: 3s - loss: 0.0767 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0531 1377/1875 [=====================>........] - ETA: 3s - loss: 0.0767 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0531 1384/1875 [=====================>........] - ETA: 3s - loss: 0.0767 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0531 1391/1875 [=====================>........] - ETA: 3s - loss: 0.0767 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0531 1399/1875 [=====================>........] - ETA: 3s - loss: 0.0767 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0531 1407/1875 [=====================>........] - ETA: 3s - loss: 0.0767 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0531 1414/1875 [=====================>........] - ETA: 3s - loss: 0.0767 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0531 1421/1875 [=====================>........] - ETA: 3s - loss: 0.0767 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0531 1428/1875 [=====================>........] - ETA: 3s - loss: 0.0766 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0531 1435/1875 [=====================>........] - ETA: 3s - loss: 0.0766 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0530 1442/1875 [======================>.......] - ETA: 3s - loss: 0.0766 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0530 1450/1875 [======================>.......] - ETA: 3s - loss: 0.0766 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0530 1457/1875 [======================>.......] - ETA: 3s - loss: 0.0766 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0530 1464/1875 [======================>.......] - ETA: 2s - loss: 0.0766 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0530 1471/1875 [======================>.......] - ETA: 2s - loss: 0.0766 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0530 1478/1875 [======================>.......] - ETA: 2s - loss: 0.0766 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0530 1485/1875 [======================>.......] - ETA: 2s - loss: 0.0765 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0530 1492/1875 [======================>.......] - ETA: 2s - loss: 0.0765 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0530 1499/1875 [======================>.......] - ETA: 2s - loss: 0.0765 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0529 1506/1875 [=======================>......] - ETA: 2s - loss: 0.0765 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0529 1514/1875 [=======================>......] - ETA: 2s - loss: 0.0765 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0529 1522/1875 [=======================>......] - ETA: 2s - loss: 0.0765 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0529 1530/1875 [=======================>......] - ETA: 2s - loss: 0.0765 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0529 1537/1875 [=======================>......] - ETA: 2s - loss: 0.0764 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0529 1544/1875 [=======================>......] - ETA: 2s - loss: 0.0764 - sparse_categorical_crossentropy: 0.0236 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0529 1552/1875 [=======================>......] - ETA: 2s - loss: 0.0764 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0529 1559/1875 [=======================>......] - ETA: 2s - loss: 0.0764 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0529 1566/1875 [========================>.....] - ETA: 2s - loss: 0.0764 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0528 1575/1875 [========================>.....] - ETA: 2s - loss: 0.0764 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0528 1585/1875 [========================>.....] - ETA: 2s - loss: 0.0763 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0528 1594/1875 [========================>.....] - ETA: 2s - loss: 0.0763 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0528 1603/1875 [========================>.....] - ETA: 1s - loss: 0.0763 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0528 1612/1875 [========================>.....] - ETA: 1s - loss: 0.0763 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0528 1622/1875 [========================>.....] - ETA: 1s - loss: 0.0763 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0528 1631/1875 [=========================>....] - ETA: 1s - loss: 0.0763 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0527 1640/1875 [=========================>....] - ETA: 1s - loss: 0.0762 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0527 1649/1875 [=========================>....] - ETA: 1s - loss: 0.0762 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0527 1658/1875 [=========================>....] - ETA: 1s - loss: 0.0762 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0527 1667/1875 [=========================>....] - ETA: 1s - loss: 0.0762 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0527 1676/1875 [=========================>....] - ETA: 1s - loss: 0.0762 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0527 1685/1875 [=========================>....] - ETA: 1s - loss: 0.0761 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0527 1694/1875 [==========================>...] - ETA: 1s - loss: 0.0761 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0526 1703/1875 [==========================>...] - ETA: 1s - loss: 0.0761 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0526 1712/1875 [==========================>...] - ETA: 1s - loss: 0.0761 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0526 1721/1875 [==========================>...] - ETA: 1s - loss: 0.0761 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0526 1730/1875 [==========================>...] - ETA: 1s - loss: 0.0760 - sparse_categorical_crossentropy: 0.0235 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0526 1740/1875 [==========================>...] - ETA: 0s - loss: 0.0760 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0526 1749/1875 [==========================>...] - ETA: 0s - loss: 0.0760 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0526 1758/1875 [===========================>..] - ETA: 0s - loss: 0.0760 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0525 1768/1875 [===========================>..] - ETA: 0s - loss: 0.0760 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0525 1777/1875 [===========================>..] - ETA: 0s - loss: 0.0760 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0525 1786/1875 [===========================>..] - ETA: 0s - loss: 0.0759 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0525 1795/1875 [===========================>..] - ETA: 0s - loss: 0.0759 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0525 1805/1875 [===========================>..] - ETA: 0s - loss: 0.0759 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0525 1815/1875 [============================>.] - ETA: 0s - loss: 0.0759 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0525 1824/1875 [============================>.] - ETA: 0s - loss: 0.0759 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0524 1833/1875 [============================>.] - ETA: 0s - loss: 0.0758 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0524 1842/1875 [============================>.] - ETA: 0s - loss: 0.0758 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0524 1851/1875 [============================>.] - ETA: 0s - loss: 0.0758 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0524 1860/1875 [============================>.] - ETA: 0s - loss: 0.0758 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0524 1869/1875 [============================>.] - ETA: 0s - loss: 0.0758 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0524 1875/1875 [==============================] - 13s 7ms/step - loss: 0.0758 - sparse_categorical_crossentropy: 0.0234 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0524 Epoch 5/5 1/1875 [..............................] - ETA: 12:52 - loss: 0.1396 - sparse_categorical_crossentropy: 0.0366 - sparse_categorical_accuracy: 0.9688 - scaled_adversarial_loss: 0.1030 8/1875 [..............................] - ETA: 13s - loss: 0.0975 - sparse_categorical_crossentropy: 0.0274 - sparse_categorical_accuracy: 0.9822 - scaled_adversarial_loss: 0.0702 15/1875 [..............................] - ETA: 13s - loss: 0.0805 - sparse_categorical_crossentropy: 0.0222 - sparse_categorical_accuracy: 0.9867 - scaled_adversarial_loss: 0.0583 22/1875 [..............................] - ETA: 13s - loss: 0.0708 - sparse_categorical_crossentropy: 0.0188 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0520 29/1875 [..............................] - ETA: 13s - loss: 0.0653 - sparse_categorical_crossentropy: 0.0169 - sparse_categorical_accuracy: 0.9908 - scaled_adversarial_loss: 0.0484 36/1875 [..............................] - ETA: 13s - loss: 0.0626 - sparse_categorical_crossentropy: 0.0163 - sparse_categorical_accuracy: 0.9914 - scaled_adversarial_loss: 0.0463 43/1875 [..............................] - ETA: 13s - loss: 0.0608 - sparse_categorical_crossentropy: 0.0160 - sparse_categorical_accuracy: 0.9918 - scaled_adversarial_loss: 0.0448 50/1875 [..............................] - ETA: 13s - loss: 0.0594 - sparse_categorical_crossentropy: 0.0157 - sparse_categorical_accuracy: 0.9922 - scaled_adversarial_loss: 0.0437 57/1875 [..............................] - ETA: 13s - loss: 0.0587 - sparse_categorical_crossentropy: 0.0158 - sparse_categorical_accuracy: 0.9924 - scaled_adversarial_loss: 0.0429 64/1875 [>.............................] - ETA: 13s - loss: 0.0582 - sparse_categorical_crossentropy: 0.0159 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0423 71/1875 [>.............................] - ETA: 13s - loss: 0.0580 - sparse_categorical_crossentropy: 0.0161 - sparse_categorical_accuracy: 0.9925 - scaled_adversarial_loss: 0.0419 78/1875 [>.............................] - ETA: 13s - loss: 0.0579 - sparse_categorical_crossentropy: 0.0162 - sparse_categorical_accuracy: 0.9926 - scaled_adversarial_loss: 0.0417 85/1875 [>.............................] - ETA: 13s - loss: 0.0579 - sparse_categorical_crossentropy: 0.0163 - sparse_categorical_accuracy: 0.9927 - scaled_adversarial_loss: 0.0416 92/1875 [>.............................] - ETA: 13s - loss: 0.0579 - sparse_categorical_crossentropy: 0.0164 - sparse_categorical_accuracy: 0.9927 - scaled_adversarial_loss: 0.0415 99/1875 [>.............................] - ETA: 13s - loss: 0.0578 - sparse_categorical_crossentropy: 0.0164 - sparse_categorical_accuracy: 0.9928 - scaled_adversarial_loss: 0.0414 106/1875 [>.............................] - ETA: 13s - loss: 0.0579 - sparse_categorical_crossentropy: 0.0165 - sparse_categorical_accuracy: 0.9928 - scaled_adversarial_loss: 0.0414 113/1875 [>.............................] - ETA: 13s - loss: 0.0579 - sparse_categorical_crossentropy: 0.0166 - sparse_categorical_accuracy: 0.9928 - scaled_adversarial_loss: 0.0414 120/1875 [>.............................] - ETA: 13s - loss: 0.0580 - sparse_categorical_crossentropy: 0.0166 - sparse_categorical_accuracy: 0.9929 - scaled_adversarial_loss: 0.0414 127/1875 [=>............................] - ETA: 13s - loss: 0.0580 - sparse_categorical_crossentropy: 0.0166 - sparse_categorical_accuracy: 0.9929 - scaled_adversarial_loss: 0.0413 134/1875 [=>............................] - ETA: 12s - loss: 0.0579 - sparse_categorical_crossentropy: 0.0166 - sparse_categorical_accuracy: 0.9929 - scaled_adversarial_loss: 0.0413 141/1875 [=>............................] - ETA: 12s - loss: 0.0579 - sparse_categorical_crossentropy: 0.0167 - sparse_categorical_accuracy: 0.9930 - scaled_adversarial_loss: 0.0412 148/1875 [=>............................] - ETA: 12s - loss: 0.0579 - sparse_categorical_crossentropy: 0.0167 - sparse_categorical_accuracy: 0.9930 - scaled_adversarial_loss: 0.0412 155/1875 [=>............................] - ETA: 12s - loss: 0.0579 - sparse_categorical_crossentropy: 0.0168 - sparse_categorical_accuracy: 0.9930 - scaled_adversarial_loss: 0.0411 162/1875 [=>............................] - ETA: 12s - loss: 0.0579 - sparse_categorical_crossentropy: 0.0169 - sparse_categorical_accuracy: 0.9930 - scaled_adversarial_loss: 0.0411 169/1875 [=>............................] - ETA: 12s - loss: 0.0580 - sparse_categorical_crossentropy: 0.0169 - sparse_categorical_accuracy: 0.9930 - scaled_adversarial_loss: 0.0411 177/1875 [=>............................] - ETA: 12s - loss: 0.0580 - sparse_categorical_crossentropy: 0.0170 - sparse_categorical_accuracy: 0.9930 - scaled_adversarial_loss: 0.0411 184/1875 [=>............................] - ETA: 12s - loss: 0.0581 - sparse_categorical_crossentropy: 0.0170 - sparse_categorical_accuracy: 0.9930 - scaled_adversarial_loss: 0.0411 191/1875 [==>...........................] - ETA: 12s - loss: 0.0582 - sparse_categorical_crossentropy: 0.0171 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0411 198/1875 [==>...........................] - ETA: 12s - loss: 0.0584 - sparse_categorical_crossentropy: 0.0172 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0412 205/1875 [==>...........................] - ETA: 12s - loss: 0.0585 - sparse_categorical_crossentropy: 0.0173 - sparse_categorical_accuracy: 0.9930 - scaled_adversarial_loss: 0.0412 213/1875 [==>...........................] - ETA: 12s - loss: 0.0586 - sparse_categorical_crossentropy: 0.0174 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0412 220/1875 [==>...........................] - ETA: 12s - loss: 0.0587 - sparse_categorical_crossentropy: 0.0175 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 227/1875 [==>...........................] - ETA: 12s - loss: 0.0588 - sparse_categorical_crossentropy: 0.0175 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 234/1875 [==>...........................] - ETA: 12s - loss: 0.0589 - sparse_categorical_crossentropy: 0.0176 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 242/1875 [==>...........................] - ETA: 12s - loss: 0.0589 - sparse_categorical_crossentropy: 0.0177 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 249/1875 [==>...........................] - ETA: 12s - loss: 0.0590 - sparse_categorical_crossentropy: 0.0177 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 256/1875 [===>..........................] - ETA: 12s - loss: 0.0590 - sparse_categorical_crossentropy: 0.0177 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 263/1875 [===>..........................] - ETA: 12s - loss: 0.0590 - sparse_categorical_crossentropy: 0.0178 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 270/1875 [===>..........................] - ETA: 11s - loss: 0.0590 - sparse_categorical_crossentropy: 0.0178 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 277/1875 [===>..........................] - ETA: 11s - loss: 0.0591 - sparse_categorical_crossentropy: 0.0178 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 284/1875 [===>..........................] - ETA: 11s - loss: 0.0591 - sparse_categorical_crossentropy: 0.0178 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 291/1875 [===>..........................] - ETA: 11s - loss: 0.0592 - sparse_categorical_crossentropy: 0.0179 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 298/1875 [===>..........................] - ETA: 11s - loss: 0.0592 - sparse_categorical_crossentropy: 0.0179 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 305/1875 [===>..........................] - ETA: 11s - loss: 0.0593 - sparse_categorical_crossentropy: 0.0179 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 312/1875 [===>..........................] - ETA: 11s - loss: 0.0593 - sparse_categorical_crossentropy: 0.0180 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 319/1875 [====>.........................] - ETA: 11s - loss: 0.0594 - sparse_categorical_crossentropy: 0.0180 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0414 326/1875 [====>.........................] - ETA: 11s - loss: 0.0594 - sparse_categorical_crossentropy: 0.0180 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0414 334/1875 [====>.........................] - ETA: 11s - loss: 0.0595 - sparse_categorical_crossentropy: 0.0181 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0414 341/1875 [====>.........................] - ETA: 11s - loss: 0.0595 - sparse_categorical_crossentropy: 0.0181 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0414 348/1875 [====>.........................] - ETA: 11s - loss: 0.0595 - sparse_categorical_crossentropy: 0.0181 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0414 355/1875 [====>.........................] - ETA: 11s - loss: 0.0595 - sparse_categorical_crossentropy: 0.0181 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0414 362/1875 [====>.........................] - ETA: 11s - loss: 0.0595 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0414 369/1875 [====>.........................] - ETA: 11s - loss: 0.0595 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0414 376/1875 [=====>........................] - ETA: 11s - loss: 0.0595 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 383/1875 [=====>........................] - ETA: 11s - loss: 0.0595 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 390/1875 [=====>........................] - ETA: 11s - loss: 0.0595 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 397/1875 [=====>........................] - ETA: 11s - loss: 0.0595 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0413 404/1875 [=====>........................] - ETA: 11s - loss: 0.0595 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9932 - scaled_adversarial_loss: 0.0413 411/1875 [=====>........................] - ETA: 10s - loss: 0.0595 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9932 - scaled_adversarial_loss: 0.0413 418/1875 [=====>........................] - ETA: 10s - loss: 0.0594 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9932 - scaled_adversarial_loss: 0.0413 425/1875 [=====>........................] - ETA: 10s - loss: 0.0594 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9932 - scaled_adversarial_loss: 0.0413 432/1875 [=====>........................] - ETA: 10s - loss: 0.0594 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9932 - scaled_adversarial_loss: 0.0413 439/1875 [======>.......................] - ETA: 10s - loss: 0.0594 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9932 - scaled_adversarial_loss: 0.0413 446/1875 [======>.......................] - ETA: 10s - loss: 0.0594 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9932 - scaled_adversarial_loss: 0.0413 453/1875 [======>.......................] - ETA: 10s - loss: 0.0594 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9932 - scaled_adversarial_loss: 0.0413 460/1875 [======>.......................] - ETA: 10s - loss: 0.0595 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9932 - scaled_adversarial_loss: 0.0413 467/1875 [======>.......................] - ETA: 10s - loss: 0.0595 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9932 - scaled_adversarial_loss: 0.0413 475/1875 [======>.......................] - ETA: 10s - loss: 0.0595 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9932 - scaled_adversarial_loss: 0.0413 482/1875 [======>.......................] - ETA: 10s - loss: 0.0595 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9932 - scaled_adversarial_loss: 0.0413 489/1875 [======>.......................] - ETA: 10s - loss: 0.0596 - sparse_categorical_crossentropy: 0.0182 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0413 496/1875 [======>.......................] - ETA: 10s - loss: 0.0596 - sparse_categorical_crossentropy: 0.0183 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0413 503/1875 [=======>......................] - ETA: 10s - loss: 0.0596 - sparse_categorical_crossentropy: 0.0183 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0413 510/1875 [=======>......................] - ETA: 10s - loss: 0.0596 - sparse_categorical_crossentropy: 0.0183 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0414 517/1875 [=======>......................] - ETA: 10s - loss: 0.0597 - sparse_categorical_crossentropy: 0.0183 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0414 524/1875 [=======>......................] - ETA: 10s - loss: 0.0597 - sparse_categorical_crossentropy: 0.0183 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0414 531/1875 [=======>......................] - ETA: 10s - loss: 0.0597 - sparse_categorical_crossentropy: 0.0183 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0414 538/1875 [=======>......................] - ETA: 10s - loss: 0.0598 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0414 545/1875 [=======>......................] - ETA: 9s - loss: 0.0598 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0414 552/1875 [=======>......................] - ETA: 9s - loss: 0.0598 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0414 559/1875 [=======>......................] - ETA: 9s - loss: 0.0599 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0415 566/1875 [========>.....................] - ETA: 9s - loss: 0.0599 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0415 573/1875 [========>.....................] - ETA: 9s - loss: 0.0600 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0415 580/1875 [========>.....................] - ETA: 9s - loss: 0.0600 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0415 588/1875 [========>.....................] - ETA: 9s - loss: 0.0600 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0415 595/1875 [========>.....................] - ETA: 9s - loss: 0.0601 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0415 602/1875 [========>.....................] - ETA: 9s - loss: 0.0601 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0416 609/1875 [========>.....................] - ETA: 9s - loss: 0.0601 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0416 617/1875 [========>.....................] - ETA: 9s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0416 624/1875 [========>.....................] - ETA: 9s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0416 631/1875 [=========>....................] - ETA: 9s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0416 639/1875 [=========>....................] - ETA: 9s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0416 646/1875 [=========>....................] - ETA: 9s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9933 - scaled_adversarial_loss: 0.0416 653/1875 [=========>....................] - ETA: 9s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0416 660/1875 [=========>....................] - ETA: 9s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 667/1875 [=========>....................] - ETA: 9s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 674/1875 [=========>....................] - ETA: 8s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 681/1875 [=========>....................] - ETA: 8s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 688/1875 [==========>...................] - ETA: 8s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 695/1875 [==========>...................] - ETA: 8s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 703/1875 [==========>...................] - ETA: 8s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 710/1875 [==========>...................] - ETA: 8s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 717/1875 [==========>...................] - ETA: 8s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 724/1875 [==========>...................] - ETA: 8s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 731/1875 [==========>...................] - ETA: 8s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 738/1875 [==========>...................] - ETA: 8s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 745/1875 [==========>...................] - ETA: 8s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 752/1875 [===========>..................] - ETA: 8s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 760/1875 [===========>..................] - ETA: 8s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 768/1875 [===========>..................] - ETA: 8s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 775/1875 [===========>..................] - ETA: 8s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 782/1875 [===========>..................] - ETA: 8s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 790/1875 [===========>..................] - ETA: 8s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 797/1875 [===========>..................] - ETA: 8s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 804/1875 [===========>..................] - ETA: 7s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 811/1875 [===========>..................] - ETA: 7s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 818/1875 [============>.................] - ETA: 7s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 825/1875 [============>.................] - ETA: 7s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 832/1875 [============>.................] - ETA: 7s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9934 - scaled_adversarial_loss: 0.0417 839/1875 [============>.................] - ETA: 7s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0417 847/1875 [============>.................] - ETA: 7s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0417 854/1875 [============>.................] - ETA: 7s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0417 861/1875 [============>.................] - ETA: 7s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0417 868/1875 [============>.................] - ETA: 7s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0417 875/1875 [=============>................] - ETA: 7s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0417 882/1875 [=============>................] - ETA: 7s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0417 889/1875 [=============>................] - ETA: 7s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0417 896/1875 [=============>................] - ETA: 7s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0417 903/1875 [=============>................] - ETA: 7s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0417 911/1875 [=============>................] - ETA: 7s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0417 918/1875 [=============>................] - ETA: 7s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 925/1875 [=============>................] - ETA: 7s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 933/1875 [=============>................] - ETA: 7s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 940/1875 [==============>...............] - ETA: 6s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 947/1875 [==============>...............] - ETA: 6s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 954/1875 [==============>...............] - ETA: 6s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 961/1875 [==============>...............] - ETA: 6s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 968/1875 [==============>...............] - ETA: 6s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 975/1875 [==============>...............] - ETA: 6s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 982/1875 [==============>...............] - ETA: 6s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 989/1875 [==============>...............] - ETA: 6s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 996/1875 [==============>...............] - ETA: 6s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 1003/1875 [===============>..............] - ETA: 6s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 1010/1875 [===============>..............] - ETA: 6s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 1017/1875 [===============>..............] - ETA: 6s - loss: 0.0605 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 1024/1875 [===============>..............] - ETA: 6s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 1031/1875 [===============>..............] - ETA: 6s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 1038/1875 [===============>..............] - ETA: 6s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 1045/1875 [===============>..............] - ETA: 6s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 1052/1875 [===============>..............] - ETA: 6s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 1059/1875 [===============>..............] - ETA: 6s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 1066/1875 [================>.............] - ETA: 6s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 1073/1875 [================>.............] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 1080/1875 [================>.............] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 1087/1875 [================>.............] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 1094/1875 [================>.............] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0418 1101/1875 [================>.............] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1109/1875 [================>.............] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1116/1875 [================>.............] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1123/1875 [================>.............] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1130/1875 [=================>............] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1137/1875 [=================>............] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0187 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1144/1875 [=================>............] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1151/1875 [=================>............] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1158/1875 [=================>............] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1165/1875 [=================>............] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1173/1875 [=================>............] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1181/1875 [=================>............] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1189/1875 [==================>...........] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1197/1875 [==================>...........] - ETA: 5s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1204/1875 [==================>...........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1211/1875 [==================>...........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1218/1875 [==================>...........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1226/1875 [==================>...........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1233/1875 [==================>...........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1240/1875 [==================>...........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1247/1875 [==================>...........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1254/1875 [===================>..........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1261/1875 [===================>..........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1268/1875 [===================>..........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1275/1875 [===================>..........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1282/1875 [===================>..........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1289/1875 [===================>..........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1296/1875 [===================>..........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1303/1875 [===================>..........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1311/1875 [===================>..........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1319/1875 [====================>.........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1326/1875 [====================>.........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1333/1875 [====================>.........] - ETA: 4s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1340/1875 [====================>.........] - ETA: 3s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1347/1875 [====================>.........] - ETA: 3s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1354/1875 [====================>.........] - ETA: 3s - loss: 0.0604 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1361/1875 [====================>.........] - ETA: 3s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1369/1875 [====================>.........] - ETA: 3s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1376/1875 [=====================>........] - ETA: 3s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1383/1875 [=====================>........] - ETA: 3s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1390/1875 [=====================>........] - ETA: 3s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1397/1875 [=====================>........] - ETA: 3s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1404/1875 [=====================>........] - ETA: 3s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1412/1875 [=====================>........] - ETA: 3s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0186 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1419/1875 [=====================>........] - ETA: 3s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1426/1875 [=====================>........] - ETA: 3s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9936 - scaled_adversarial_loss: 0.0418 1434/1875 [=====================>........] - ETA: 3s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0418 1441/1875 [======================>.......] - ETA: 3s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0418 1448/1875 [======================>.......] - ETA: 3s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0418 1455/1875 [======================>.......] - ETA: 3s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0418 1462/1875 [======================>.......] - ETA: 3s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0418 1469/1875 [======================>.......] - ETA: 3s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1476/1875 [======================>.......] - ETA: 2s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1483/1875 [======================>.......] - ETA: 2s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1490/1875 [======================>.......] - ETA: 2s - loss: 0.0603 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1497/1875 [======================>.......] - ETA: 2s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1504/1875 [=======================>......] - ETA: 2s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1511/1875 [=======================>......] - ETA: 2s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1518/1875 [=======================>......] - ETA: 2s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1525/1875 [=======================>......] - ETA: 2s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1532/1875 [=======================>......] - ETA: 2s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1539/1875 [=======================>......] - ETA: 2s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1547/1875 [=======================>......] - ETA: 2s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1554/1875 [=======================>......] - ETA: 2s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1561/1875 [=======================>......] - ETA: 2s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1569/1875 [========================>.....] - ETA: 2s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1578/1875 [========================>.....] - ETA: 2s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1587/1875 [========================>.....] - ETA: 2s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1596/1875 [========================>.....] - ETA: 2s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1605/1875 [========================>.....] - ETA: 1s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1615/1875 [========================>.....] - ETA: 1s - loss: 0.0602 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1624/1875 [========================>.....] - ETA: 1s - loss: 0.0601 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1633/1875 [=========================>....] - ETA: 1s - loss: 0.0601 - sparse_categorical_crossentropy: 0.0185 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1642/1875 [=========================>....] - ETA: 1s - loss: 0.0601 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1651/1875 [=========================>....] - ETA: 1s - loss: 0.0601 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1660/1875 [=========================>....] - ETA: 1s - loss: 0.0601 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1669/1875 [=========================>....] - ETA: 1s - loss: 0.0601 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1678/1875 [=========================>....] - ETA: 1s - loss: 0.0601 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1688/1875 [==========================>...] - ETA: 1s - loss: 0.0601 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1698/1875 [==========================>...] - ETA: 1s - loss: 0.0601 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1708/1875 [==========================>...] - ETA: 1s - loss: 0.0601 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1717/1875 [==========================>...] - ETA: 1s - loss: 0.0601 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0417 1726/1875 [==========================>...] - ETA: 1s - loss: 0.0601 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0416 1735/1875 [==========================>...] - ETA: 1s - loss: 0.0600 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0416 1744/1875 [==========================>...] - ETA: 0s - loss: 0.0600 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0416 1753/1875 [===========================>..] - ETA: 0s - loss: 0.0600 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0416 1762/1875 [===========================>..] - ETA: 0s - loss: 0.0600 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0416 1771/1875 [===========================>..] - ETA: 0s - loss: 0.0600 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0416 1780/1875 [===========================>..] - ETA: 0s - loss: 0.0600 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9937 - scaled_adversarial_loss: 0.0416 1789/1875 [===========================>..] - ETA: 0s - loss: 0.0600 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9938 - scaled_adversarial_loss: 0.0416 1798/1875 [===========================>..] - ETA: 0s - loss: 0.0600 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9938 - scaled_adversarial_loss: 0.0416 1808/1875 [===========================>..] - ETA: 0s - loss: 0.0600 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9938 - scaled_adversarial_loss: 0.0416 1818/1875 [============================>.] - ETA: 0s - loss: 0.0600 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9938 - scaled_adversarial_loss: 0.0416 1827/1875 [============================>.] - ETA: 0s - loss: 0.0600 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9938 - scaled_adversarial_loss: 0.0416 1836/1875 [============================>.] - ETA: 0s - loss: 0.0600 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9938 - scaled_adversarial_loss: 0.0416 1845/1875 [============================>.] - ETA: 0s - loss: 0.0600 - sparse_categorical_crossentropy: 0.0184 - sparse_categorical_accuracy: 0.9938 - scaled_adversarial_loss: 0.0416 1855/1875 [============================>.] - ETA: 0s - loss: 0.0599 - sparse_categorical_crossentropy: 0.0183 - sparse_categorical_accuracy: 0.9938 - scaled_adversarial_loss: 0.0416 1864/1875 [============================>.] - ETA: 0s - loss: 0.0599 - sparse_categorical_crossentropy: 0.0183 - sparse_categorical_accuracy: 0.9938 - scaled_adversarial_loss: 0.0416 1874/1875 [============================>.] - ETA: 0s - loss: 0.0599 - sparse_categorical_crossentropy: 0.0183 - sparse_categorical_accuracy: 0.9938 - scaled_adversarial_loss: 0.0416 1875/1875 [==============================] - 14s 7ms/step - loss: 0.0599 - sparse_categorical_crossentropy: 0.0183 - sparse_categorical_accuracy: 0.9938 - scaled_adversarial_loss: 0.0416 <tensorflow.python.keras.callbacks.History at 0x7fbfe500d588>
results = adv_model.evaluate(test_set_for_adv_model)
named_results = dict(zip(adv_model.metrics_names, results))
print('\naccuracy:', named_results['sparse_categorical_accuracy'])
WARNING:tensorflow:The dtype of the watched tensor must be floating (e.g. tf.float32), got tf.int64 WARNING:tensorflow:The dtype of the watched tensor must be floating (e.g. tf.float32), got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 1/313 [..............................] - ETA: 1:40 - loss: 0.0810 - sparse_categorical_crossentropy: 0.0415 - sparse_categorical_accuracy: 0.9688 - scaled_adversarial_loss: 0.0395 12/313 [>.............................] - ETA: 1s - loss: 0.0582 - sparse_categorical_crossentropy: 0.0267 - sparse_categorical_accuracy: 0.9896 - scaled_adversarial_loss: 0.0315 24/313 [=>............................] - ETA: 1s - loss: 0.0646 - sparse_categorical_crossentropy: 0.0283 - sparse_categorical_accuracy: 0.9909 - scaled_adversarial_loss: 0.0363 36/313 [==>...........................] - ETA: 1s - loss: 0.0658 - sparse_categorical_crossentropy: 0.0250 - sparse_categorical_accuracy: 0.9922 - scaled_adversarial_loss: 0.0408 48/313 [===>..........................] - ETA: 1s - loss: 0.0616 - sparse_categorical_crossentropy: 0.0221 - sparse_categorical_accuracy: 0.9935 - scaled_adversarial_loss: 0.0396 59/313 [====>.........................] - ETA: 1s - loss: 0.0639 - sparse_categorical_crossentropy: 0.0217 - sparse_categorical_accuracy: 0.9931 - scaled_adversarial_loss: 0.0422 70/313 [=====>........................] - ETA: 1s - loss: 0.0718 - sparse_categorical_crossentropy: 0.0277 - sparse_categorical_accuracy: 0.9906 - scaled_adversarial_loss: 0.0440 82/313 [======>.......................] - ETA: 1s - loss: 0.0714 - sparse_categorical_crossentropy: 0.0278 - sparse_categorical_accuracy: 0.9905 - scaled_adversarial_loss: 0.0436 94/313 [========>.....................] - ETA: 0s - loss: 0.0760 - sparse_categorical_crossentropy: 0.0305 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0454 106/313 [=========>....................] - ETA: 0s - loss: 0.0750 - sparse_categorical_crossentropy: 0.0297 - sparse_categorical_accuracy: 0.9900 - scaled_adversarial_loss: 0.0453 118/313 [==========>...................] - ETA: 0s - loss: 0.0787 - sparse_categorical_crossentropy: 0.0326 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0461 131/313 [===========>..................] - ETA: 0s - loss: 0.0759 - sparse_categorical_crossentropy: 0.0307 - sparse_categorical_accuracy: 0.9897 - scaled_adversarial_loss: 0.0452 144/313 [============>.................] - ETA: 0s - loss: 0.0757 - sparse_categorical_crossentropy: 0.0301 - sparse_categorical_accuracy: 0.9902 - scaled_adversarial_loss: 0.0456 156/313 [=============>................] - ETA: 0s - loss: 0.0763 - sparse_categorical_crossentropy: 0.0312 - sparse_categorical_accuracy: 0.9902 - scaled_adversarial_loss: 0.0451 168/313 [===============>..............] - ETA: 0s - loss: 0.0762 - sparse_categorical_crossentropy: 0.0315 - sparse_categorical_accuracy: 0.9903 - scaled_adversarial_loss: 0.0447 180/313 [================>.............] - ETA: 0s - loss: 0.0754 - sparse_categorical_crossentropy: 0.0310 - sparse_categorical_accuracy: 0.9901 - scaled_adversarial_loss: 0.0444 192/313 [=================>............] - ETA: 0s - loss: 0.0764 - sparse_categorical_crossentropy: 0.0313 - sparse_categorical_accuracy: 0.9899 - scaled_adversarial_loss: 0.0451 203/313 [==================>...........] - ETA: 0s - loss: 0.0759 - sparse_categorical_crossentropy: 0.0308 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0450 214/313 [===================>..........] - ETA: 0s - loss: 0.0769 - sparse_categorical_crossentropy: 0.0321 - sparse_categorical_accuracy: 0.9898 - scaled_adversarial_loss: 0.0448 226/313 [====================>.........] - ETA: 0s - loss: 0.0781 - sparse_categorical_crossentropy: 0.0329 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0452 238/313 [=====================>........] - ETA: 0s - loss: 0.0812 - sparse_categorical_crossentropy: 0.0351 - sparse_categorical_accuracy: 0.9891 - scaled_adversarial_loss: 0.0460 250/313 [======================>.......] - ETA: 0s - loss: 0.0807 - sparse_categorical_crossentropy: 0.0349 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0458 261/313 [========================>.....] - ETA: 0s - loss: 0.0823 - sparse_categorical_crossentropy: 0.0361 - sparse_categorical_accuracy: 0.9892 - scaled_adversarial_loss: 0.0461 273/313 [=========================>....] - ETA: 0s - loss: 0.0819 - sparse_categorical_crossentropy: 0.0355 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0464 285/313 [==========================>...] - ETA: 0s - loss: 0.0819 - sparse_categorical_crossentropy: 0.0358 - sparse_categorical_accuracy: 0.9893 - scaled_adversarial_loss: 0.0461 297/313 [===========================>..] - ETA: 0s - loss: 0.0813 - sparse_categorical_crossentropy: 0.0348 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0465 308/313 [============================>.] - ETA: 0s - loss: 0.0809 - sparse_categorical_crossentropy: 0.0345 - sparse_categorical_accuracy: 0.9894 - scaled_adversarial_loss: 0.0463 313/313 [==============================] - 2s 4ms/step - loss: 0.0810 - sparse_categorical_crossentropy: 0.0345 - sparse_categorical_accuracy: 0.9895 - scaled_adversarial_loss: 0.0465 accuracy: 0.9894999861717224
적대적 정규화 모델도 테스트세트에서 매우 우수한 것을 볼 수 있습니다(99% 정확성).
적대적 교란에서의 견고성
이제 적대적 교란에서의 견고성을 위해 기본 모델과 적대적 정규화 모델을 비교합니다.
AdversarialRegularization.perturb_on_batch 함수를 사용하여 적대적으로 교란된 예를 생성합니다. 그리고 기본 모델을 기반으로 한 세대를 사용합니다. 이를 위해 기본 모델을 AdversarialRegularization으로 래핑합니다. 훈련(Model.fit)을 호출하지 않는 한, 모델에서 학습된 변수는 변경되지 않으며, 모델은 여전히 기본 모델 섹션에서와 같습니다.
reference_model = nsl.keras.AdversarialRegularization(
base_model,
label_keys=[LABEL_INPUT_NAME],
adv_config=adv_config)
reference_model.compile(
optimizer='adam',
loss='sparse_categorical_crossentropy',
metrics=['acc'])
평가할 모델을 사전에 수집하고 각 모델에 대한 메트릭 객체를 만듭니다.
기본 모델과 같은 입력 형식(레이블 정보 필요 없음)을 갖기 위해 adv_model.base_model을 사용합니다. adv_model.base_model에서 학습된 변수는 adv_model에서의 변수와 같습니다.
models_to_eval = {
'base': base_model,
'adv-regularized': adv_model.base_model
}
metrics = {
name: tf.keras.metrics.SparseCategoricalAccuracy()
for name in models_to_eval.keys()
}
다음은 교란된 예를 생성하고 이를 사용하여 모델을 평가하는 루프입니다. 다음 섹션에서 시각화를 위해 교란된 이미지, 레이블 및 예측값을 저장합니다.
perturbed_images, labels, predictions = [], [], []
for batch in test_set_for_adv_model:
perturbed_batch = reference_model.perturb_on_batch(batch)
# Clipping makes perturbed examples have the same range as regular ones.
perturbed_batch[IMAGE_INPUT_NAME] = tf.clip_by_value(
perturbed_batch[IMAGE_INPUT_NAME], 0.0, 1.0)
y_true = perturbed_batch.pop(LABEL_INPUT_NAME)
perturbed_images.append(perturbed_batch[IMAGE_INPUT_NAME].numpy())
labels.append(y_true.numpy())
predictions.append({})
for name, model in models_to_eval.items():
y_pred = model(perturbed_batch)
metrics[name](y_true, y_pred)
predictions[-1][name] = tf.argmax(y_pred, axis=-1).numpy()
for name, metric in metrics.items():
print('%s model accuracy: %f' % (name, metric.result().numpy()))
WARNING:tensorflow:The dtype of the watched tensor must be floating (e.g. tf.float32), got tf.int64 WARNING:tensorflow:The dtype of the watched tensor must be floating (e.g. tf.float32), got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:absl:Cannot perturb features dict_keys(['label']) WARNING:tensorflow:The dtype of the watched tensor must be floating (e.g. tf.float32), got tf.int64 WARNING:tensorflow:The dtype of the watched tensor must be floating (e.g. tf.float32), got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the watched tensor must be floating (e.g. tf.float32), got tf.int64 WARNING:tensorflow:The dtype of the watched tensor must be floating (e.g. tf.float32), got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the watched tensor must be floating (e.g. tf.float32), got tf.int64 WARNING:tensorflow:The dtype of the watched tensor must be floating (e.g. tf.float32), got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the watched tensor must be floating (e.g. tf.float32), got tf.int64 WARNING:tensorflow:The dtype of the watched tensor must be floating (e.g. tf.float32), got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 WARNING:tensorflow:The dtype of the source tensor must be floating (e.g. tf.float32) when calling GradientTape.gradient, got tf.int64 base model accuracy: 0.570600 adv-regularized model accuracy: 0.957200
입력이 적대적으로 교란될 때 기본 모델의 정확성이 급격히 떨어지는 것을 볼 수 있습니다(99%에서 약 50%로). 반면에 적대적 정규화 모델의 정확성은 약간만 저하됩니다(99%에서 95%로). 이는 모델의 견고성 향상에 대한 적대적 학습의 효과를 보여줍니다.
적대적으로 교란된 이미지의 예
다음에서 적대적으로 교란된 이미지를 살펴봅니다. 교란된 이미지는 여전히 사람이 인식할 수 있는 숫자를 보여주지만, 기본 모델을 속일 수 있음을 알 수 있습니다.
batch_index = 0
batch_image = perturbed_images[batch_index]
batch_label = labels[batch_index]
batch_pred = predictions[batch_index]
batch_size = HPARAMS.batch_size
n_col = 4
n_row = (batch_size + n_col - 1) / n_col
print('accuracy in batch %d:' % batch_index)
for name, pred in batch_pred.items():
print('%s model: %d / %d' % (name, np.sum(batch_label == pred), batch_size))
plt.figure(figsize=(15, 15))
for i, (image, y) in enumerate(zip(batch_image, batch_label)):
y_base = batch_pred['base'][i]
y_adv = batch_pred['adv-regularized'][i]
plt.subplot(n_row, n_col, i+1)
plt.title('true: %d, base: %d, adv: %d' % (y, y_base, y_adv))
plt.imshow(tf.keras.preprocessing.image.array_to_img(image), cmap='gray')
plt.axis('off')
plt.show()
accuracy in batch 0: base model: 20 / 32 adv-regularized model: 31 / 32 /tmpfs/src/tf_docs_env/lib/python3.6/site-packages/ipykernel_launcher.py:19: MatplotlibDeprecationWarning: Passing non-integers as three-element position specification is deprecated since 3.3 and will be removed two minor releases later.
결론
Neural Structured Learning(NSL) 프레임워크를 사용하여 이미지 분류를 위한 적대적 학습을 사용하는 방법을 시연했습니다. 사용자가 다양한 적대적 설정(하이퍼 매개변수에서)을 실험하고 모델 견고성에 어떻게 영향을 미치는지 확인할 것을 권장합니다.