def test_pad():
img = _image()
assert img.shape[0] == 480
assert img.shape[1] == 480
assert img.shape[2] == 3
augmentor = Pad(100)
result = augmentor(**{'image': img})
image = result['image']
_show_image(image)
assert image.shape[0] == 680
assert image.shape[1] == 680
assert image.shape[2] == 3
augmentor = Pad((40, 30))
result = augmentor(**{'image': img})
image = result['image']
_show_image(image)
assert image.shape[0] == 480 + 30 * 2
assert image.shape[1] == 480 + 40 * 2
assert image.shape[2] == 3
},
]),
'weight_decay_rate':
0.0001,
}
NETWORK = SmartDict()
NETWORK.OPTIMIZER_CLASS = None
NETWORK.OPTIMIZER_KWARGS = {}
NETWORK.LEARNING_RATE_FUNC = None
NETWORK.LEARNING_RATE_KWARGS = {}
NETWORK.WEIGHT_DECAY_RATE = None
NETWORK.IMAGE_SIZE = IMAGE_SIZE
NETWORK.BATCH_SIZE = BATCH_SIZE
NETWORK.DATA_FORMAT = DATA_FORMAT
NETWORK.ACTIVATION_QUANTIZER = linear_mid_tread_half_quantizer
NETWORK.ACTIVATION_QUANTIZER_KWARGS = {'bit': 2, 'max_value': 2}
NETWORK.WEIGHT_QUANTIZER = binary_mean_scaling_quantizer
NETWORK.WEIGHT_QUANTIZER_KWARGS = {}
# dataset
DATASET = SmartDict()
DATASET.BATCH_SIZE = BATCH_SIZE
DATASET.DATA_FORMAT = DATA_FORMAT
DATASET.PRE_PROCESSOR = PRE_PROCESSOR
DATASET.AUGMENTOR = Sequence([
Pad(2),
Crop(size=IMAGE_SIZE),
FlipLeftRight(),
])
NETWORK = EasyDict()
NETWORK.OPTIMIZER_CLASS = tf.compat.v1.train.MomentumOptimizer
NETWORK.OPTIMIZER_KWARGS = {"momentum": 0.9}
NETWORK.LEARNING_RATE_FUNC = tf.compat.v1.train.piecewise_constant
step_per_epoch = 50000 // 200
NETWORK.LEARNING_RATE_KWARGS = {
"values": [0.01, 0.001, 0.0001, 0.00001],
"boundaries":
[step_per_epoch * 200, step_per_epoch * 300, step_per_epoch * 350],
}
NETWORK.IMAGE_SIZE = IMAGE_SIZE
NETWORK.BATCH_SIZE = BATCH_SIZE
NETWORK.DATA_FORMAT = DATA_FORMAT
NETWORK.WEIGHT_DECAY_RATE = 0.0005
NETWORK.ACTIVATION_QUANTIZER = linear_mid_tread_half_quantizer
NETWORK.ACTIVATION_QUANTIZER_KWARGS = {'bit': 2, 'max_value': 2}
NETWORK.WEIGHT_QUANTIZER = binary_mean_scaling_quantizer
NETWORK.WEIGHT_QUANTIZER_KWARGS = {}
# dataset
DATASET = EasyDict()
DATASET.BATCH_SIZE = BATCH_SIZE
DATASET.DATA_FORMAT = DATA_FORMAT
DATASET.PRE_PROCESSOR = PRE_PROCESSOR
DATASET.AUGMENTOR = Sequence([
Pad(4),
Crop(size=IMAGE_SIZE),
FlipLeftRight(),
])
DATASET.TRAIN_VALIDATION_SAVING_SIZE = 5000
