01-10 08:57
Notice
Recent Posts
Recent Comments
Link
«   2025/01   »
일 월 화 수 목 금 토
1 2 3 4
5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21 22 23 24 25
26 27 28 29 30 31
Tags more
Archives
Today
Total
관리 메뉴

너와나의 관심사

tensorflow lite 로 personalization on-devcie 본문

카테고리 없음

tensorflow lite 로 personalization on-devcie

벤치마킹 2021. 4. 12. 01:03
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
import tensorflow as tf
from tensorflow.keras import layers
from tensorflow.keras.regularizers import l2
from tfltransfer import bases
from tfltransfer import heads
from tfltransfer import optimizers
from tfltransfer.tflite_transfer_converter import TFLiteTransferConverter
import tensorflow.keras.backend as K
 
#tf.compat.v1.disable_v2_behavior()
#tf.compat.v1.disable_eager_execution()
#tf.compat.v1.disable_eager_execution()
#tf.compat.v1.disable_v2_behavior()
#tf.compat.v1.enable_eager_execution()
#tf.disable_v2_behavior()
 
 
DEFAULT_BATCH_SIZE = 32
 
input_size = 224
output_size = 5
 
DEFAULT_INPUT_SIZE = 64
DEFAULT_BATCH_SIZE = 128
LEARNING_RATE = 0.001
 
'''
batch_size = DEFAULT_BATCH_SIZE
base = bases.MobileNetV2Base(image_size=input_size)
head = heads.SoftmaxClassifierHead(batch_size, base.bottleneck_shape(),
                                   output_size)
optimizer = optimizers.SGD(LEARNING_RATE)
converter = TFLiteTransferConverter(
    output_size, base, head, optimizer, batch_size)
models = converter.convert_and_save('custom_keras_model')
#parameter_shapes = [(7 * 7 * 1280, output_size), (output_size,)]
'''
 
 
 
def test_mobilenet_v2_saved_model():
    input_size = DEFAULT_INPUT_SIZE
    output_size = 4
    base = bases.MobileNetV2Base(image_size=input_size)
 
    head_model = tf.keras.Sequential([
        layers.Flatten(input_shape=(7, 7, 1280)),
        layers.Dropout(0.25),
        layers.Dense(
            units=32,
            activation='relu',
            kernel_regularizer=l2(0.01),
            bias_regularizer=l2(0.01)),
        layers.Dropout(0.25),
        layers.Dense(
            units=32,
            activation='relu',
            kernel_regularizer=l2(0.01),
            bias_regularizer=l2(0.01)),
        layers.Dense(
            units=4,
            activation='softmax',
            kernel_regularizer=l2(0.01),
            bias_regularizer=l2(0.01)),
    ])
 
    head_model.compile(loss="categorical_crossentropy", optimizer="sgd")
 
    #head_model.compile(loss='categorical_crossentropy', optimizer='sgd')
    converter = TFLiteTransferConverter(
        output_size, base, heads.KerasModelHead(head_model),
        optimizers.SGD(LEARNING_RATE), DEFAULT_BATCH_SIZE)
 
    models = converter.convert_and_save('custom_keras_model_second')
    models = converter._convert()
 
test_mobilenet_v2_saved_model()
 
 
cs
저작자표시
Comments