├── .gitignore ├── LICENSE ├── README.md ├── conda_macos_dependencies.yml ├── convert-from-old-version.py ├── custom-layers.py ├── custom-train.py ├── dataset.py ├── early-stop.py ├── reuse_model_layer.py ├── save-load.py └── simplified-training.py /.gitignore: -------------------------------------------------------------------------------- 1 | training/ 2 | .idea/ 3 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | Apache License 2 | Version 2.0, January 2004 3 | http://www.apache.org/licenses/ 4 | 5 | TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION 6 | 7 | 1. Definitions. 8 | 9 | "License" shall mean the terms and conditions for use, reproduction, 10 | and distribution as defined by Sections 1 through 9 of this document. 11 | 12 | "Licensor" shall mean the copyright owner or entity authorized by 13 | the copyright owner that is granting the License. 14 | 15 | "Legal Entity" shall mean the union of the acting entity and all 16 | other entities that control, are controlled by, or are under common 17 | control with that entity. For the purposes of this definition, 18 | "control" means (i) the power, direct or indirect, to cause the 19 | direction or management of such entity, whether by contract or 20 | otherwise, or (ii) ownership of fifty percent (50%) or more of the 21 | outstanding shares, or (iii) beneficial ownership of such entity. 22 | 23 | "You" (or "Your") shall mean an individual or Legal Entity 24 | exercising permissions granted by this License. 25 | 26 | "Source" form shall mean the preferred form for making modifications, 27 | including but not limited to software source code, documentation 28 | source, and configuration files. 29 | 30 | "Object" form shall mean any form resulting from mechanical 31 | transformation or translation of a Source form, including but 32 | not limited to compiled object code, generated documentation, 33 | and conversions to other media types. 34 | 35 | "Work" shall mean the work of authorship, whether in Source or 36 | Object form, made available under the License, as indicated by a 37 | copyright notice that is included in or attached to the work 38 | (an example is provided in the Appendix below). 39 | 40 | "Derivative Works" shall mean any work, whether in Source or Object 41 | form, that is based on (or derived from) the Work and for which the 42 | editorial revisions, annotations, elaborations, or other modifications 43 | represent, as a whole, an original work of authorship. For the purposes 44 | of this License, Derivative Works shall not include works that remain 45 | separable from, or merely link (or bind by name) to the interfaces of, 46 | the Work and Derivative Works thereof. 47 | 48 | "Contribution" shall mean any work of authorship, including 49 | the original version of the Work and any modifications or additions 50 | to that Work or Derivative Works thereof, that is intentionally 51 | submitted to Licensor for inclusion in the Work by the copyright owner 52 | or by an individual or Legal Entity authorized to submit on behalf of 53 | the copyright owner. For the purposes of this definition, "submitted" 54 | means any form of electronic, verbal, or written communication sent 55 | to the Licensor or its representatives, including but not limited to 56 | communication on electronic mailing lists, source code control systems, 57 | and issue tracking systems that are managed by, or on behalf of, the 58 | Licensor for the purpose of discussing and improving the Work, but 59 | excluding communication that is conspicuously marked or otherwise 60 | designated in writing by the copyright owner as "Not a Contribution." 61 | 62 | "Contributor" shall mean Licensor and any individual or Legal Entity 63 | on behalf of whom a Contribution has been received by Licensor and 64 | subsequently incorporated within the Work. 65 | 66 | 2. Grant of Copyright License. Subject to the terms and conditions of 67 | this License, each Contributor hereby grants to You a perpetual, 68 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable 69 | copyright license to reproduce, prepare Derivative Works of, 70 | publicly display, publicly perform, sublicense, and distribute the 71 | Work and such Derivative Works in Source or Object form. 72 | 73 | 3. Grant of Patent License. Subject to the terms and conditions of 74 | this License, each Contributor hereby grants to You a perpetual, 75 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable 76 | (except as stated in this section) patent license to make, have made, 77 | use, offer to sell, sell, import, and otherwise transfer the Work, 78 | where such license applies only to those patent claims licensable 79 | by such Contributor that are necessarily infringed by their 80 | Contribution(s) alone or by combination of their Contribution(s) 81 | with the Work to which such Contribution(s) was submitted. If You 82 | institute patent litigation against any entity (including a 83 | cross-claim or counterclaim in a lawsuit) alleging that the Work 84 | or a Contribution incorporated within the Work constitutes direct 85 | or contributory patent infringement, then any patent licenses 86 | granted to You under this License for that Work shall terminate 87 | as of the date such litigation is filed. 88 | 89 | 4. Redistribution. You may reproduce and distribute copies of the 90 | Work or Derivative Works thereof in any medium, with or without 91 | modifications, and in Source or Object form, provided that You 92 | meet the following conditions: 93 | 94 | (a) You must give any other recipients of the Work or 95 | Derivative Works a copy of this License; and 96 | 97 | (b) You must cause any modified files to carry prominent notices 98 | stating that You changed the files; and 99 | 100 | (c) You must retain, in the Source form of any Derivative Works 101 | that You distribute, all copyright, patent, trademark, and 102 | attribution notices from the Source form of the Work, 103 | excluding those notices that do not pertain to any part of 104 | the Derivative Works; and 105 | 106 | (d) If the Work includes a "NOTICE" text file as part of its 107 | distribution, then any Derivative Works that You distribute must 108 | include a readable copy of the attribution notices contained 109 | within such NOTICE file, excluding those notices that do not 110 | pertain to any part of the Derivative Works, in at least one 111 | of the following places: within a NOTICE text file distributed 112 | as part of the Derivative Works; within the Source form or 113 | documentation, if provided along with the Derivative Works; or, 114 | within a display generated by the Derivative Works, if and 115 | wherever such third-party notices normally appear. The contents 116 | of the NOTICE file are for informational purposes only and 117 | do not modify the License. You may add Your own attribution 118 | notices within Derivative Works that You distribute, alongside 119 | or as an addendum to the NOTICE text from the Work, provided 120 | that such additional attribution notices cannot be construed 121 | as modifying the License. 122 | 123 | You may add Your own copyright statement to Your modifications and 124 | may provide additional or different license terms and conditions 125 | for use, reproduction, or distribution of Your modifications, or 126 | for any such Derivative Works as a whole, provided Your use, 127 | reproduction, and distribution of the Work otherwise complies with 128 | the conditions stated in this License. 129 | 130 | 5. Submission of Contributions. Unless You explicitly state otherwise, 131 | any Contribution intentionally submitted for inclusion in the Work 132 | by You to the Licensor shall be under the terms and conditions of 133 | this License, without any additional terms or conditions. 134 | Notwithstanding the above, nothing herein shall supersede or modify 135 | the terms of any separate license agreement you may have executed 136 | with Licensor regarding such Contributions. 137 | 138 | 6. Trademarks. This License does not grant permission to use the trade 139 | names, trademarks, service marks, or product names of the Licensor, 140 | except as required for reasonable and customary use in describing the 141 | origin of the Work and reproducing the content of the NOTICE file. 142 | 143 | 7. Disclaimer of Warranty. Unless required by applicable law or 144 | agreed to in writing, Licensor provides the Work (and each 145 | Contributor provides its Contributions) on an "AS IS" BASIS, 146 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or 147 | implied, including, without limitation, any warranties or conditions 148 | of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A 149 | PARTICULAR PURPOSE. You are solely responsible for determining the 150 | appropriateness of using or redistributing the Work and assume any 151 | risks associated with Your exercise of permissions under this License. 152 | 153 | 8. Limitation of Liability. In no event and under no legal theory, 154 | whether in tort (including negligence), contract, or otherwise, 155 | unless required by applicable law (such as deliberate and grossly 156 | negligent acts) or agreed to in writing, shall any Contributor be 157 | liable to You for damages, including any direct, indirect, special, 158 | incidental, or consequential damages of any character arising as a 159 | result of this License or out of the use or inability to use the 160 | Work (including but not limited to damages for loss of goodwill, 161 | work stoppage, computer failure or malfunction, or any and all 162 | other commercial damages or losses), even if such Contributor 163 | has been advised of the possibility of such damages. 164 | 165 | 9. Accepting Warranty or Additional Liability. While redistributing 166 | the Work or Derivative Works thereof, You may choose to offer, 167 | and charge a fee for, acceptance of support, warranty, indemnity, 168 | or other liability obligations and/or rights consistent with this 169 | License. However, in accepting such obligations, You may act only 170 | on Your own behalf and on Your sole responsibility, not on behalf 171 | of any other Contributor, and only if You agree to indemnify, 172 | defend, and hold each Contributor harmless for any liability 173 | incurred by, or claims asserted against, such Contributor by reason 174 | of your accepting any such warranty or additional liability. 175 | 176 | END OF TERMS AND CONDITIONS 177 | 178 | APPENDIX: How to apply the Apache License to your work. 179 | 180 | To apply the Apache License to your work, attach the following 181 | boilerplate notice, with the fields enclosed by brackets "[]" 182 | replaced with your own identifying information. (Don't include 183 | the brackets!) The text should be enclosed in the appropriate 184 | comment syntax for the file format. We also recommend that a 185 | file or class name and description of purpose be included on the 186 | same "printed page" as the copyright notice for easier 187 | identification within third-party archives. 188 | 189 | Copyright [yyyy] [name of copyright owner] 190 | 191 | Licensed under the Apache License, Version 2.0 (the "License"); 192 | you may not use this file except in compliance with the License. 193 | You may obtain a copy of the License at 194 | 195 | http://www.apache.org/licenses/LICENSE-2.0 196 | 197 | Unless required by applicable law or agreed to in writing, software 198 | distributed under the License is distributed on an "AS IS" BASIS, 199 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 200 | See the License for the specific language governing permissions and 201 | limitations under the License. 202 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # Tensorflow 2.0 Tutorials 2 | 3 | There are some simple toy examples of the usages of tf2.0. 4 | 5 | The style of Tf2.0 (keras style) is similar with pytorch now, we can easily define a model with many layers. 6 | 7 | To me, the biggest change would be the use case of "session", it has been deprecated in the new version. 8 | 9 | More tutorials can be found in this [official site](https://www.tensorflow.org/tutorials/). -------------------------------------------------------------------------------- /conda_macos_dependencies.yml: -------------------------------------------------------------------------------- 1 | name: tf2 2 | channels: 3 | - defaults 4 | dependencies: 5 | - blas=1.0=mkl 6 | - ca-certificates=2019.10.16=0 7 | - certifi=2019.9.11=py36_0 8 | - cycler=0.10.0=py36hfc81398_0 9 | - freetype=2.9.1=hb4e5f40_0 10 | - intel-openmp=2019.4=233 11 | - jpeg=9b=he5867d9_2 12 | - kiwisolver=1.1.0=py36h0a44026_0 13 | - libcxx=4.0.1=hcfea43d_1 14 | - libcxxabi=4.0.1=hcfea43d_1 15 | - libedit=3.1.20181209=hb402a30_0 16 | - libffi=3.2.1=h475c297_4 17 | - libgfortran=3.0.1=h93005f0_2 18 | - libpng=1.6.37=ha441bb4_0 19 | - libtiff=4.0.10=hcb84e12_2 20 | - matplotlib=3.1.1=py36h54f8f79_0 21 | - mkl=2019.4=233 22 | - mkl-service=2.3.0=py36hfbe908c_0 23 | - mkl_fft=1.0.15=py36h5e564d8_0 24 | - mkl_random=1.1.0=py36ha771720_0 25 | - ncurses=6.1=h0a44026_1 26 | - numpy-base=1.17.3=py36h6575580_0 27 | - olefile=0.46=py36_0 28 | - openssl=1.1.1d=h1de35cc_3 29 | - pillow=6.2.1=py36hb68e598_0 30 | - pip=19.3.1=py36_0 31 | - pyparsing=2.4.2=py_0 32 | - python=3.6.9=h359304d_0 33 | - python-dateutil=2.8.1=py_0 34 | - pytz=2019.3=py_0 35 | - readline=7.0=h1de35cc_5 36 | - setuptools=41.6.0=py36_0 37 | - sqlite=3.30.1=ha441bb4_0 38 | - tk=8.6.8=ha441bb4_0 39 | - tornado=6.0.3=py36h1de35cc_0 40 | - wheel=0.33.6=py36_0 41 | - xz=5.2.4=h1de35cc_4 42 | - zlib=1.2.11=h1de35cc_3 43 | - zstd=1.3.7=h5bba6e5_0 44 | - pip: 45 | - absl-py==0.8.1 46 | - astor==0.8.0 47 | - gast==0.3.2 48 | - google-pasta==0.1.8 49 | - grpcio==1.25.0 50 | - h5py==2.10.0 51 | - keras-applications==1.0.8 52 | - keras-preprocessing==1.1.0 53 | - markdown==3.1.1 54 | - numpy==1.17.3 55 | - protobuf==3.10.0 56 | - six==1.13.0 57 | - tb-nightly==1.14.0a20190603 58 | - tensorflow==2.0.0b0 59 | - termcolor==1.1.0 60 | - tf-estimator-nightly==1.14.0.dev2019060501 61 | - werkzeug==0.16.0 62 | - wrapt==1.11.2 63 | prefix: /anaconda3/envs/tf2 64 | 65 | -------------------------------------------------------------------------------- /convert-from-old-version.py: -------------------------------------------------------------------------------- 1 | import tensorflow as tf 2 | import numpy as np 3 | 4 | data_x = np.random.normal(size=[1000, 1]) 5 | noise = np.random.normal(size=[1000, 1]) * 0.2 6 | data_y = data_x * 3. + 2. + noise 7 | 8 | 9 | class Model: 10 | def __init__(self): 11 | super(Model, self).__init__() 12 | self.w = tf.Variable(0.1, dtype=tf.float32) 13 | self.b = tf.Variable(0.1, dtype=tf.float32) 14 | 15 | def call(self, x): 16 | return self.w * x + self.b 17 | 18 | 19 | model = Model() 20 | var_list = [model.w, model.b] 21 | opt = tf.optimizers.SGD(0.1) 22 | 23 | for t in range(100): 24 | with tf.GradientTape() as tape: 25 | y_ = model.call(data_x) 26 | loss = tf.reduce_mean(tf.square(data_y - y_)) 27 | 28 | grad = tape.gradient(loss, var_list) 29 | opt.apply_gradients(zip(grad, var_list)) 30 | if t % 10 == 0: 31 | print("loss={:.2f} | w={:.2f} | b={:.2f}".format( 32 | loss, model.w.numpy(), model.b.numpy()) 33 | ) 34 | -------------------------------------------------------------------------------- /custom-layers.py: -------------------------------------------------------------------------------- 1 | import tensorflow as tf 2 | from tensorflow import keras 3 | import numpy as np 4 | 5 | 6 | data_x = np.random.normal(size=[1000, 1]) 7 | noise = np.random.normal(size=[1000, 1]) * 0.2 8 | data_y = data_x * 3. + 2. + noise 9 | 10 | train_x, train_y = data_x[:900], data_y[:900] 11 | test_x, test_y = data_x[900:], data_y[900:] 12 | 13 | 14 | class MyLayer(keras.layers.Layer): 15 | def __init__(self, num_outputs): 16 | super().__init__() 17 | self.num_outputs = num_outputs 18 | self.w = None 19 | self.b = None 20 | 21 | def build(self, input_shape): 22 | self.w = self.add_weight( 23 | name="w", 24 | shape=[int(input_shape[-1]), self.num_outputs], 25 | dtype=tf.float32, 26 | initializer=keras.initializers.RandomNormal(), 27 | ) 28 | self.b = self.add_weight( 29 | name="b", 30 | shape=[1, self.num_outputs], 31 | dtype=tf.float32, 32 | initializer=keras.initializers.Constant(0.1), 33 | ) 34 | 35 | def call(self, inputs, **kwargs): 36 | return tf.matmul(inputs, self.w) + self.b 37 | 38 | 39 | model = keras.models.Sequential([ 40 | MyLayer(10), 41 | keras.layers.Dense(1), 42 | ]) 43 | 44 | model.compile( 45 | optimizer=keras.optimizers.SGD(0.01), 46 | loss=keras.losses.MeanSquaredError(), 47 | metrics=[keras.metrics.MeanSquaredError()], 48 | ) 49 | 50 | model.fit(train_x, train_y, batch_size=32, epochs=3, validation_split=0.2, shuffle=True) 51 | model.evaluate(test_x, test_y, verbose=1) -------------------------------------------------------------------------------- /custom-train.py: -------------------------------------------------------------------------------- 1 | import tensorflow as tf 2 | from tensorflow import keras 3 | import numpy as np 4 | 5 | data_x = np.random.normal(size=[1000, 1]) 6 | noise = np.random.normal(size=[1000, 1]) * 0.2 7 | data_y = data_x * 3. + 2. + noise 8 | 9 | 10 | class Model(keras.Model): 11 | def __init__(self): 12 | super(Model, self).__init__() 13 | self.l1 = keras.layers.Dense(10, activation=keras.activations.relu) 14 | self.l2 = keras.layers.Dense(1) 15 | 16 | def call(self, x, training=None, mask=None): 17 | x = self.l1(x) 18 | x = self.l2(x) 19 | return x 20 | 21 | 22 | @tf.function 23 | def train_step(x, y): 24 | with tf.GradientTape() as tape: 25 | y_ = model(x) 26 | loss = loss_func(y, y_) 27 | grad = tape.gradient(loss, model.trainable_variables) 28 | opt.apply_gradients(zip(grad, model.trainable_variables)) 29 | return loss 30 | 31 | 32 | model = Model() 33 | opt = tf.optimizers.SGD(0.1) 34 | loss_func = keras.losses.MeanSquaredError() 35 | 36 | 37 | for t in range(100): 38 | loss = train_step(data_x, data_y) 39 | if t % 10 == 0: 40 | print("loss={:.2f}".format(loss.numpy())) -------------------------------------------------------------------------------- /dataset.py: -------------------------------------------------------------------------------- 1 | import tensorflow as tf 2 | from tensorflow import keras 3 | import numpy as np 4 | 5 | data_x = np.random.normal(size=[1000, 1]) 6 | noise = np.random.normal(size=[1000, 1]) * 0.2 7 | data_y = data_x * 3. + 2. + noise 8 | 9 | 10 | train_x, train_y = data_x[:900], data_y[:900] 11 | test_x, test_y = data_x[900:], data_y[900:] 12 | train_ds = tf.data.Dataset.from_tensor_slices((train_x, train_y)).shuffle(1000).batch(32) 13 | test_ds = tf.data.Dataset.from_tensor_slices((test_x, test_y)).shuffle(1000).batch(32) 14 | 15 | model = keras.models.Sequential([ 16 | keras.layers.Dense(10, activation=keras.activations.relu, input_shape=(1, )), 17 | keras.layers.Dense(1), 18 | ]) 19 | model.compile( 20 | optimizer=keras.optimizers.SGD(0.01), 21 | loss=keras.losses.MeanSquaredError(), 22 | metrics=[keras.metrics.MeanSquaredError()], 23 | ) 24 | 25 | for epoch in range(3): 26 | train_losses = [] 27 | for bx, by in train_ds: 28 | out = model.train_on_batch(bx, by) 29 | # out = [loss, metrics] 30 | train_losses.append(out[0]) 31 | 32 | test_losses = [] 33 | for bx, by in test_ds: 34 | loss = model.evaluate(bx, by, verbose=0) 35 | test_losses.append(loss) 36 | print("train loss={:.2f} | test loss={:.2f}".format(np.mean(train_losses), np.mean(test_losses))) 37 | -------------------------------------------------------------------------------- /early-stop.py: -------------------------------------------------------------------------------- 1 | from tensorflow import keras 2 | import numpy as np 3 | import matplotlib.pyplot as plt 4 | 5 | 6 | data_x = np.random.normal(size=[1000, 1]) 7 | noise = np.random.normal(size=[1000, 1]) * 0.2 8 | data_y = data_x * 3. + 2. + noise 9 | 10 | train_x, train_y = data_x[:900], data_y[:900] 11 | test_x, test_y = data_x[900:], data_y[900:] 12 | 13 | model = keras.models.Sequential([ 14 | keras.layers.Dense(10, activation=keras.activations.relu, input_shape=(1, )), 15 | keras.layers.Dense(1), 16 | ]) 17 | model.compile( 18 | optimizer=keras.optimizers.SGD(0.01), 19 | loss=keras.losses.MeanSquaredError(), 20 | metrics=[keras.metrics.MeanSquaredError()], 21 | ) 22 | 23 | early_stop = keras.callbacks.EarlyStopping(monitor="val_loss", patience=4) 24 | history = model.fit( 25 | train_x, train_y, batch_size=32, epochs=100, validation_split=0.2, shuffle=True, 26 | callbacks=[early_stop, ] 27 | ) 28 | 29 | plt.plot(history.history["val_loss"]) 30 | plt.xlabel("epoch") 31 | plt.ylabel("val_loss") 32 | plt.show() -------------------------------------------------------------------------------- /reuse_model_layer.py: -------------------------------------------------------------------------------- 1 | from tensorflow import keras 2 | import numpy as np 3 | 4 | data_x = np.random.normal(size=[1000, 1]) 5 | noise = np.random.normal(size=[1000, 1]) * 0.2 6 | data_y = data_x * 3. + 2. + noise 7 | 8 | train_x, train_y = data_x[:900], data_y[:900] 9 | test_x, test_y = data_x[900:], data_y[900:] 10 | 11 | 12 | # define your reusable layers in here 13 | l1 = keras.layers.Dense(10, activation=keras.activations.relu) 14 | 15 | 16 | class Model(keras.Model): 17 | def __init__(self): 18 | super(Model, self).__init__() 19 | self.l1 = l1 # this is a reusable layer 20 | self.l2 = keras.layers.Dense(1) # this is NOT a reusable layer 21 | 22 | def call(self, x, training=None, mask=None): 23 | x = self.l1(x) 24 | x = self.l2(x) 25 | return x 26 | 27 | 28 | model1 = Model() 29 | model2 = Model() 30 | 31 | model1.build((None, 1)) 32 | model2.build((None, 1)) 33 | 34 | model1.compile( 35 | optimizer=keras.optimizers.SGD(0.01), 36 | loss=keras.losses.MeanSquaredError(), 37 | metrics=[keras.metrics.MeanSquaredError()], 38 | ) 39 | 40 | # train model1 for a while 41 | model1.fit(train_x, train_y, batch_size=32, epochs=3, validation_split=0.2, shuffle=True) 42 | print("l1 is reused: ", np.all(model1.l1.get_weights()[0] == model2.l1.get_weights()[0])) 43 | print("l2 is reused: ", np.all(model1.l2.get_weights()[0] == model2.l2.get_weights()[0])) -------------------------------------------------------------------------------- /save-load.py: -------------------------------------------------------------------------------- 1 | import tensorflow as tf 2 | from tensorflow import keras 3 | import numpy as np 4 | import os 5 | 6 | 7 | data_x = np.random.normal(size=[1000, 1]) 8 | noise = np.random.normal(size=[1000, 1]) * 0.2 9 | data_y = data_x * 3. + 2. + noise 10 | 11 | train_x, train_y = data_x[:900], data_y[:900] 12 | test_x, test_y = data_x[900:], data_y[900:] 13 | 14 | 15 | def create_model(): 16 | return keras.models.Sequential([ 17 | keras.layers.Dense(10, activation=keras.activations.relu, input_shape=(1, )), 18 | keras.layers.Dense(1), 19 | ]) 20 | 21 | 22 | model = create_model() 23 | model.compile( 24 | optimizer=keras.optimizers.SGD(0.01), 25 | loss=keras.losses.MeanSquaredError(), 26 | ) 27 | 28 | checkpoint_path = "training/cp-{epoch:04d}.ckpt" 29 | ckpt_dir = os.path.dirname(checkpoint_path) 30 | os.makedirs(ckpt_dir, exist_ok=True) 31 | 32 | cp_callback = tf.keras.callbacks.ModelCheckpoint( 33 | filepath=checkpoint_path, 34 | save_weights_only=True, 35 | verbose=1, 36 | period=5 37 | ) 38 | 39 | # save ckpt 40 | model.save_weights(checkpoint_path.format(epoch=0)) 41 | 42 | 43 | history = model.fit( 44 | train_x, train_y, batch_size=32, epochs=10, validation_data=(test_x, test_y), shuffle=True, 45 | callbacks=[cp_callback, ] # save when callback 46 | ) 47 | 48 | # restore ckpt 49 | latest_model = tf.train.latest_checkpoint(ckpt_dir) 50 | model2 = create_model() 51 | model2.load_weights(latest_model) 52 | model2.compile( 53 | optimizer=keras.optimizers.SGD(0.01), 54 | loss=keras.losses.MeanSquaredError(), 55 | ) 56 | loss = model2.evaluate(test_x, test_y, verbose=2) 57 | print("Restored ckpt model, loss: {:.2f}".format(loss)) 58 | 59 | 60 | # save pb 61 | final_path = "training/final_model" 62 | model.save(final_path) 63 | 64 | # restore 65 | model3 = keras.models.load_model(final_path) 66 | loss = model3.evaluate(test_x, test_y, verbose=2) 67 | print("Restored pb model, loss: {:.2f}".format(loss)) -------------------------------------------------------------------------------- /simplified-training.py: -------------------------------------------------------------------------------- 1 | from tensorflow import keras 2 | import numpy as np 3 | 4 | data_x = np.random.normal(size=[1000, 1]) 5 | noise = np.random.normal(size=[1000, 1]) * 0.2 6 | data_y = data_x * 3. + 2. + noise 7 | 8 | train_x, train_y = data_x[:900], data_y[:900] 9 | test_x, test_y = data_x[900:], data_y[900:] 10 | 11 | model = keras.models.Sequential([ 12 | keras.layers.Dense(10, activation=keras.activations.relu), 13 | keras.layers.Dense(1), 14 | ]) 15 | model.compile( 16 | optimizer=keras.optimizers.SGD(0.01), 17 | loss=keras.losses.MeanSquaredError(), 18 | metrics=[keras.metrics.MeanSquaredError()], 19 | ) 20 | 21 | model.fit(train_x, train_y, batch_size=32, epochs=3, validation_split=0.2, shuffle=True) 22 | model.evaluate(test_x, test_y, verbose=1) --------------------------------------------------------------------------------