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
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
|
<!DOCTYPE html>
<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
<head>
<meta charset="utf-8">
<meta http-equiv="X-UA-Compatible" content="IE=edge">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<link rel="canonical" href="http://keras.io/getting-started/functional-api-guide/">
<link rel="shortcut icon" href="../../img/favicon.ico">
<title>Guide to the Functional API - Keras Documentation</title>
<link href='https://fonts.googleapis.com/css?family=Lato:400,700|Source+Sans+Pro:400,700|Inconsolata:400,700' rel='stylesheet' type='text/css'>
<link rel="stylesheet" href="../../css/theme.css" type="text/css" />
<link rel="stylesheet" href="../../css/theme_extra.css" type="text/css" />
<link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/9.12.0/styles/github.min.css">
<script>
// Current page data
var mkdocs_page_name = "Guide to the Functional API";
var mkdocs_page_input_path = "getting-started/functional-api-guide.md";
var mkdocs_page_url = "/getting-started/functional-api-guide/";
</script>
<script src="../../js/jquery-2.1.1.min.js" defer></script>
<script src="../../js/modernizr-2.8.3.min.js" defer></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/9.12.0/highlight.min.js"></script>
<script>hljs.initHighlightingOnLoad();</script>
<script>
(function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){
(i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o),
m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m)
})(window,document,'script','https://www.google-analytics.com/analytics.js','ga');
ga('create', 'UA-61785484-1', 'keras.io');
ga('send', 'pageview');
</script>
</head>
<body class="wy-body-for-nav" role="document">
<div class="wy-grid-for-nav">
<nav data-toggle="wy-nav-shift" class="wy-nav-side stickynav">
<div class="wy-side-scroll">
<a href="">
<div class="keras-logo">
<img src="/img/keras-logo-small.jpg" class="keras-logo-img">
Keras Documentation
</div>
</a>
<div class="wy-side-nav-search">
<div role="search">
<form id ="rtd-search-form" class="wy-form" action="../../search.html" method="get">
<input type="text" name="q" placeholder="Search docs" title="Type search term here" />
</form>
</div>
</div>
<div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="main navigation">
<ul>
<li class="toctree-l1"><a class="reference internal" href="../..">Home</a>
</li>
</ul>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../why-use-keras/">Why use Keras</a>
</li>
</ul>
<p class="caption"><span class="caption-text">Getting started</span></p>
<ul class="current">
<li class="toctree-l1"><a class="reference internal" href="../sequential-model-guide/">Guide to the Sequential model</a>
</li>
<li class="toctree-l1 current"><a class="reference internal current" href="./">Guide to the Functional API</a>
<ul class="current">
<li class="toctree-l2"><a class="reference internal" href="#first-example-a-densely-connected-network">First example: a densely-connected network</a>
</li>
<li class="toctree-l2"><a class="reference internal" href="#all-models-are-callable-just-like-layers">All models are callable, just like layers</a>
</li>
<li class="toctree-l2"><a class="reference internal" href="#multi-input-and-multi-output-models">Multi-input and multi-output models</a>
</li>
<li class="toctree-l2"><a class="reference internal" href="#shared-layers">Shared layers</a>
</li>
<li class="toctree-l2"><a class="reference internal" href="#the-concept-of-layer-node">The concept of layer "node"</a>
</li>
<li class="toctree-l2"><a class="reference internal" href="#more-examples">More examples</a>
<ul>
<li class="toctree-l3"><a class="reference internal" href="#inception-module">Inception module</a>
</li>
<li class="toctree-l3"><a class="reference internal" href="#residual-connection-on-a-convolution-layer">Residual connection on a convolution layer</a>
</li>
<li class="toctree-l3"><a class="reference internal" href="#shared-vision-model">Shared vision model</a>
</li>
<li class="toctree-l3"><a class="reference internal" href="#visual-question-answering-model">Visual question answering model</a>
</li>
<li class="toctree-l3"><a class="reference internal" href="#video-question-answering-model">Video question answering model</a>
</li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../faq/">FAQ</a>
</li>
</ul>
<p class="caption"><span class="caption-text">Models</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../models/about-keras-models/">About Keras models</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../models/sequential/">Sequential</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../models/model/">Model (functional API)</a>
</li>
</ul>
<p class="caption"><span class="caption-text">Layers</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../layers/about-keras-layers/">About Keras layers</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../layers/core/">Core Layers</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../layers/convolutional/">Convolutional Layers</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../layers/pooling/">Pooling Layers</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../layers/local/">Locally-connected Layers</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../layers/recurrent/">Recurrent Layers</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../layers/embeddings/">Embedding Layers</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../layers/merge/">Merge Layers</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../layers/advanced-activations/">Advanced Activations Layers</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../layers/normalization/">Normalization Layers</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../layers/noise/">Noise layers</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../layers/wrappers/">Layer wrappers</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../layers/writing-your-own-keras-layers/">Writing your own Keras layers</a>
</li>
</ul>
<p class="caption"><span class="caption-text">Preprocessing</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../preprocessing/sequence/">Sequence Preprocessing</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../preprocessing/text/">Text Preprocessing</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../preprocessing/image/">Image Preprocessing</a>
</li>
</ul>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../losses/">Losses</a>
</li>
</ul>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../metrics/">Metrics</a>
</li>
</ul>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../optimizers/">Optimizers</a>
</li>
</ul>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../activations/">Activations</a>
</li>
</ul>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../callbacks/">Callbacks</a>
</li>
</ul>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../datasets/">Datasets</a>
</li>
</ul>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../applications/">Applications</a>
</li>
</ul>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../backend/">Backend</a>
</li>
</ul>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../initializers/">Initializers</a>
</li>
</ul>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../regularizers/">Regularizers</a>
</li>
</ul>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../constraints/">Constraints</a>
</li>
</ul>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../visualization/">Visualization</a>
</li>
</ul>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../scikit-learn-api/">Scikit-learn API</a>
</li>
</ul>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../utils/">Utils</a>
</li>
</ul>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../contributing/">Contributing</a>
</li>
</ul>
<p class="caption"><span class="caption-text">Examples</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../../examples/addition_rnn/">Addition RNN</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/antirectifier/">Custom layer - antirectifier</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/babi_rnn/">Baby RNN</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/babi_memnn/">Baby MemNN</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/cifar10_cnn/">CIFAR-10 CNN</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/cifar10_resnet/">CIFAR-10 ResNet</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/conv_filter_visualization/">Convolution filter visualization</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/conv_lstm/">Convolutional LSTM</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/deep_dream/">Deep Dream</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/image_ocr/">Image OCR</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/imdb_bidirectional_lstm/">Bidirectional LSTM</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/imdb_cnn/">1D CNN for text classification</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/imdb_cnn_lstm/">Sentiment classification CNN-LSTM</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/imdb_fasttext/">Fasttext for text classification</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/imdb_lstm/">Sentiment classification LSTM</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/lstm_seq2seq/">Sequence to sequence - training</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/lstm_seq2seq_restore/">Sequence to sequence - prediction</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/lstm_stateful/">Stateful LSTM</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/lstm_text_generation/">LSTM for text generation</a>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../examples/mnist_acgan/">Auxiliary Classifier GAN</a>
</li>
</ul>
</div>
</div>
</nav>
<section data-toggle="wy-nav-shift" class="wy-nav-content-wrap">
<nav class="wy-nav-top" role="navigation" aria-label="top navigation">
<i data-toggle="wy-nav-top" class="fa fa-bars"></i>
<a href="../..">Keras Documentation</a>
</nav>
<div class="wy-nav-content">
<div class="rst-content">
<div role="navigation" aria-label="breadcrumbs navigation">
<ul class="wy-breadcrumbs">
<li><a href="../..">Docs</a> »</li>
<li>Getting started »</li>
<li>Guide to the Functional API</li>
<li class="wy-breadcrumbs-aside">
<a href="https://github.com/keras-team/keras/tree/master/docs"
class="icon icon-github"> Edit on GitHub</a>
</li>
</ul>
<hr/>
</div>
<div role="main">
<div class="section">
<h1 id="getting-started-with-the-keras-functional-api">Getting started with the Keras functional API</h1>
<p>The Keras functional API is the way to go for defining complex models, such as multi-output models, directed acyclic graphs, or models with shared layers.</p>
<p>This guide assumes that you are already familiar with the <code>Sequential</code> model.</p>
<p>Let's start with something simple.</p>
<hr />
<h2 id="first-example-a-densely-connected-network">First example: a densely-connected network</h2>
<p>The <code>Sequential</code> model is probably a better choice to implement such a network, but it helps to start with something really simple.</p>
<ul>
<li>A layer instance is callable (on a tensor), and it returns a tensor</li>
<li>Input tensor(s) and output tensor(s) can then be used to define a <code>Model</code></li>
<li>Such a model can be trained just like Keras <code>Sequential</code> models.</li>
</ul>
<pre><code class="python">from keras.layers import Input, Dense
from keras.models import Model
# This returns a tensor
inputs = Input(shape=(784,))
# a layer instance is callable on a tensor, and returns a tensor
output_1 = Dense(64, activation='relu')(inputs)
output_2 = Dense(64, activation='relu')(output_1)
predictions = Dense(10, activation='softmax')(output_2)
# This creates a model that includes
# the Input layer and three Dense layers
model = Model(inputs=inputs, outputs=predictions)
model.compile(optimizer='rmsprop',
loss='categorical_crossentropy',
metrics=['accuracy'])
model.fit(data, labels) # starts training
</code></pre>
<hr />
<h2 id="all-models-are-callable-just-like-layers">All models are callable, just like layers</h2>
<p>With the functional API, it is easy to reuse trained models: you can treat any model as if it were a layer, by calling it on a tensor. Note that by calling a model you aren't just reusing the <em>architecture</em> of the model, you are also reusing its weights.</p>
<pre><code class="python">x = Input(shape=(784,))
# This works, and returns the 10-way softmax we defined above.
y = model(x)
</code></pre>
<p>This can allow, for instance, to quickly create models that can process <em>sequences</em> of inputs. You could turn an image classification model into a video classification model, in just one line.</p>
<pre><code class="python">from keras.layers import TimeDistributed
# Input tensor for sequences of 20 timesteps,
# each containing a 784-dimensional vector
input_sequences = Input(shape=(20, 784))
# This applies our previous model to every timestep in the input sequences.
# the output of the previous model was a 10-way softmax,
# so the output of the layer below will be a sequence of 20 vectors of size 10.
processed_sequences = TimeDistributed(model)(input_sequences)
</code></pre>
<hr />
<h2 id="multi-input-and-multi-output-models">Multi-input and multi-output models</h2>
<p>Here's a good use case for the functional API: models with multiple inputs and outputs. The functional API makes it easy to manipulate a large number of intertwined datastreams.</p>
<p>Let's consider the following model. We seek to predict how many retweets and likes a news headline will receive on Twitter. The main input to the model will be the headline itself, as a sequence of words, but to spice things up, our model will also have an auxiliary input, receiving extra data such as the time of day when the headline was posted, etc.
The model will also be supervised via two loss functions. Using the main loss function earlier in a model is a good regularization mechanism for deep models.</p>
<p>Here's what our model looks like:</p>
<p><img src="https://s3.amazonaws.com/keras.io/img/multi-input-multi-output-graph.png" alt="multi-input-multi-output-graph" style="width: 400px;"/></p>
<p>Let's implement it with the functional API.</p>
<p>The main input will receive the headline, as a sequence of integers (each integer encodes a word).
The integers will be between 1 and 10,000 (a vocabulary of 10,000 words) and the sequences will be 100 words long.</p>
<pre><code class="python">from keras.layers import Input, Embedding, LSTM, Dense
from keras.models import Model
import numpy as np
np.random.seed(0) # Set a random seed for reproducibility
# Headline input: meant to receive sequences of 100 integers, between 1 and 10000.
# Note that we can name any layer by passing it a "name" argument.
main_input = Input(shape=(100,), dtype='int32', name='main_input')
# This embedding layer will encode the input sequence
# into a sequence of dense 512-dimensional vectors.
x = Embedding(output_dim=512, input_dim=10000, input_length=100)(main_input)
# A LSTM will transform the vector sequence into a single vector,
# containing information about the entire sequence
lstm_out = LSTM(32)(x)
</code></pre>
<p>Here we insert the auxiliary loss, allowing the LSTM and Embedding layer to be trained smoothly even though the main loss will be much higher in the model.</p>
<pre><code class="python">auxiliary_output = Dense(1, activation='sigmoid', name='aux_output')(lstm_out)
</code></pre>
<p>At this point, we feed into the model our auxiliary input data by concatenating it with the LSTM output:</p>
<pre><code class="python">auxiliary_input = Input(shape=(5,), name='aux_input')
x = keras.layers.concatenate([lstm_out, auxiliary_input])
# We stack a deep densely-connected network on top
x = Dense(64, activation='relu')(x)
x = Dense(64, activation='relu')(x)
x = Dense(64, activation='relu')(x)
# And finally we add the main logistic regression layer
main_output = Dense(1, activation='sigmoid', name='main_output')(x)
</code></pre>
<p>This defines a model with two inputs and two outputs:</p>
<pre><code class="python">model = Model(inputs=[main_input, auxiliary_input], outputs=[main_output, auxiliary_output])
</code></pre>
<p>We compile the model and assign a weight of 0.2 to the auxiliary loss.
To specify different <code>loss_weights</code> or <code>loss</code> for each different output, you can use a list or a dictionary.
Here we pass a single loss as the <code>loss</code> argument, so the same loss will be used on all outputs.</p>
<pre><code class="python">model.compile(optimizer='rmsprop', loss='binary_crossentropy',
loss_weights=[1., 0.2])
</code></pre>
<p>We can train the model by passing it lists of input arrays and target arrays:</p>
<pre><code class="python">headline_data = np.round(np.abs(np.random.rand(12, 100) * 100))
additional_data = np.random.randn(12, 5)
headline_labels = np.random.randn(12, 1)
additional_labels = np.random.randn(12, 1)
model.fit([headline_data, additional_data], [headline_labels, additional_labels],
epochs=50, batch_size=32)
</code></pre>
<p>Since our inputs and outputs are named (we passed them a "name" argument),
we could also have compiled the model via:</p>
<pre><code class="python">model.compile(optimizer='rmsprop',
loss={'main_output': 'binary_crossentropy', 'aux_output': 'binary_crossentropy'},
loss_weights={'main_output': 1., 'aux_output': 0.2})
# And trained it via:
model.fit({'main_input': headline_data, 'aux_input': additional_data},
{'main_output': headline_labels, 'aux_output': additional_labels},
epochs=50, batch_size=32)
</code></pre>
<p>To use the model for inferencing, use</p>
<pre><code class="python">model.predict({'main_input': headline_data, 'aux_input': additional_data})
</code></pre>
<p>or alternatively,</p>
<pre><code class="python">pred = model.predict([headline_data, additional_data])
</code></pre>
<hr />
<h2 id="shared-layers">Shared layers</h2>
<p>Another good use for the functional API are models that use shared layers. Let's take a look at shared layers.</p>
<p>Let's consider a dataset of tweets. We want to build a model that can tell whether two tweets are from the same person or not (this can allow us to compare users by the similarity of their tweets, for instance).</p>
<p>One way to achieve this is to build a model that encodes two tweets into two vectors, concatenates the vectors and then adds a logistic regression; this outputs a probability that the two tweets share the same author. The model would then be trained on positive tweet pairs and negative tweet pairs.</p>
<p>Because the problem is symmetric, the mechanism that encodes the first tweet should be reused (weights and all) to encode the second tweet. Here we use a shared LSTM layer to encode the tweets.</p>
<p>Let's build this with the functional API. We will take as input for a tweet a binary matrix of shape <code>(280, 256)</code>, i.e. a sequence of 280 vectors of size 256, where each dimension in the 256-dimensional vector encodes the presence/absence of a character (out of an alphabet of 256 frequent characters).</p>
<pre><code class="python">import keras
from keras.layers import Input, LSTM, Dense
from keras.models import Model
tweet_a = Input(shape=(280, 256))
tweet_b = Input(shape=(280, 256))
</code></pre>
<p>To share a layer across different inputs, simply instantiate the layer once, then call it on as many inputs as you want:</p>
<pre><code class="python"># This layer can take as input a matrix
# and will return a vector of size 64
shared_lstm = LSTM(64)
# When we reuse the same layer instance
# multiple times, the weights of the layer
# are also being reused
# (it is effectively *the same* layer)
encoded_a = shared_lstm(tweet_a)
encoded_b = shared_lstm(tweet_b)
# We can then concatenate the two vectors:
merged_vector = keras.layers.concatenate([encoded_a, encoded_b], axis=-1)
# And add a logistic regression on top
predictions = Dense(1, activation='sigmoid')(merged_vector)
# We define a trainable model linking the
# tweet inputs to the predictions
model = Model(inputs=[tweet_a, tweet_b], outputs=predictions)
model.compile(optimizer='rmsprop',
loss='binary_crossentropy',
metrics=['accuracy'])
model.fit([data_a, data_b], labels, epochs=10)
</code></pre>
<p>Let's pause to take a look at how to read the shared layer's output or output shape.</p>
<hr />
<h2 id="the-concept-of-layer-node">The concept of layer "node"</h2>
<p>Whenever you are calling a layer on some input, you are creating a new tensor (the output of the layer), and you are adding a "node" to the layer, linking the input tensor to the output tensor. When you are calling the same layer multiple times, that layer owns multiple nodes indexed as 0, 1, 2...</p>
<p>In previous versions of Keras, you could obtain the output tensor of a layer instance via <code>layer.get_output()</code>, or its output shape via <code>layer.output_shape</code>. You still can (except <code>get_output()</code> has been replaced by the property <code>output</code>). But what if a layer is connected to multiple inputs?</p>
<p>As long as a layer is only connected to one input, there is no confusion, and <code>.output</code> will return the one output of the layer:</p>
<pre><code class="python">a = Input(shape=(280, 256))
lstm = LSTM(32)
encoded_a = lstm(a)
assert lstm.output == encoded_a
</code></pre>
<p>Not so if the layer has multiple inputs:</p>
<pre><code class="python">a = Input(shape=(280, 256))
b = Input(shape=(280, 256))
lstm = LSTM(32)
encoded_a = lstm(a)
encoded_b = lstm(b)
lstm.output
</code></pre>
<pre><code>>> AttributeError: Layer lstm_1 has multiple inbound nodes,
hence the notion of "layer output" is ill-defined.
Use `get_output_at(node_index)` instead.
</code></pre>
<p>Okay then. The following works:</p>
<pre><code class="python">assert lstm.get_output_at(0) == encoded_a
assert lstm.get_output_at(1) == encoded_b
</code></pre>
<p>Simple enough, right?</p>
<p>The same is true for the properties <code>input_shape</code> and <code>output_shape</code>: as long as the layer has only one node, or as long as all nodes have the same input/output shape, then the notion of "layer output/input shape" is well defined, and that one shape will be returned by <code>layer.output_shape</code>/<code>layer.input_shape</code>. But if, for instance, you apply the same <code>Conv2D</code> layer to an input of shape <code>(32, 32, 3)</code>, and then to an input of shape <code>(64, 64, 3)</code>, the layer will have multiple input/output shapes, and you will have to fetch them by specifying the index of the node they belong to:</p>
<pre><code class="python">a = Input(shape=(32, 32, 3))
b = Input(shape=(64, 64, 3))
conv = Conv2D(16, (3, 3), padding='same')
conved_a = conv(a)
# Only one input so far, the following will work:
assert conv.input_shape == (None, 32, 32, 3)
conved_b = conv(b)
# now the `.input_shape` property wouldn't work, but this does:
assert conv.get_input_shape_at(0) == (None, 32, 32, 3)
assert conv.get_input_shape_at(1) == (None, 64, 64, 3)
</code></pre>
<hr />
<h2 id="more-examples">More examples</h2>
<p>Code examples are still the best way to get started, so here are a few more.</p>
<h3 id="inception-module">Inception module</h3>
<p>For more information about the Inception architecture, see <a href="http://arxiv.org/abs/1409.4842">Going Deeper with Convolutions</a>.</p>
<pre><code class="python">from keras.layers import Conv2D, MaxPooling2D, Input
input_img = Input(shape=(256, 256, 3))
tower_1 = Conv2D(64, (1, 1), padding='same', activation='relu')(input_img)
tower_1 = Conv2D(64, (3, 3), padding='same', activation='relu')(tower_1)
tower_2 = Conv2D(64, (1, 1), padding='same', activation='relu')(input_img)
tower_2 = Conv2D(64, (5, 5), padding='same', activation='relu')(tower_2)
tower_3 = MaxPooling2D((3, 3), strides=(1, 1), padding='same')(input_img)
tower_3 = Conv2D(64, (1, 1), padding='same', activation='relu')(tower_3)
output = keras.layers.concatenate([tower_1, tower_2, tower_3], axis=1)
</code></pre>
<h3 id="residual-connection-on-a-convolution-layer">Residual connection on a convolution layer</h3>
<p>For more information about residual networks, see <a href="http://arxiv.org/abs/1512.03385">Deep Residual Learning for Image Recognition</a>.</p>
<pre><code class="python">from keras.layers import Conv2D, Input
# input tensor for a 3-channel 256x256 image
x = Input(shape=(256, 256, 3))
# 3x3 conv with 3 output channels (same as input channels)
y = Conv2D(3, (3, 3), padding='same')(x)
# this returns x + y.
z = keras.layers.add([x, y])
</code></pre>
<h3 id="shared-vision-model">Shared vision model</h3>
<p>This model reuses the same image-processing module on two inputs, to classify whether two MNIST digits are the same digit or different digits.</p>
<pre><code class="python">from keras.layers import Conv2D, MaxPooling2D, Input, Dense, Flatten
from keras.models import Model
# First, define the vision modules
digit_input = Input(shape=(27, 27, 1))
x = Conv2D(64, (3, 3))(digit_input)
x = Conv2D(64, (3, 3))(x)
x = MaxPooling2D((2, 2))(x)
out = Flatten()(x)
vision_model = Model(digit_input, out)
# Then define the tell-digits-apart model
digit_a = Input(shape=(27, 27, 1))
digit_b = Input(shape=(27, 27, 1))
# The vision model will be shared, weights and all
out_a = vision_model(digit_a)
out_b = vision_model(digit_b)
concatenated = keras.layers.concatenate([out_a, out_b])
out = Dense(1, activation='sigmoid')(concatenated)
classification_model = Model([digit_a, digit_b], out)
</code></pre>
<h3 id="visual-question-answering-model">Visual question answering model</h3>
<p>This model can select the correct one-word answer when asked a natural-language question about a picture.</p>
<p>It works by encoding the question into a vector, encoding the image into a vector, concatenating the two, and training on top a logistic regression over some vocabulary of potential answers.</p>
<pre><code class="python">from keras.layers import Conv2D, MaxPooling2D, Flatten
from keras.layers import Input, LSTM, Embedding, Dense
from keras.models import Model, Sequential
# First, let's define a vision model using a Sequential model.
# This model will encode an image into a vector.
vision_model = Sequential()
vision_model.add(Conv2D(64, (3, 3), activation='relu', padding='same', input_shape=(224, 224, 3)))
vision_model.add(Conv2D(64, (3, 3), activation='relu'))
vision_model.add(MaxPooling2D((2, 2)))
vision_model.add(Conv2D(128, (3, 3), activation='relu', padding='same'))
vision_model.add(Conv2D(128, (3, 3), activation='relu'))
vision_model.add(MaxPooling2D((2, 2)))
vision_model.add(Conv2D(256, (3, 3), activation='relu', padding='same'))
vision_model.add(Conv2D(256, (3, 3), activation='relu'))
vision_model.add(Conv2D(256, (3, 3), activation='relu'))
vision_model.add(MaxPooling2D((2, 2)))
vision_model.add(Flatten())
# Now let's get a tensor with the output of our vision model:
image_input = Input(shape=(224, 224, 3))
encoded_image = vision_model(image_input)
# Next, let's define a language model to encode the question into a vector.
# Each question will be at most 100 words long,
# and we will index words as integers from 1 to 9999.
question_input = Input(shape=(100,), dtype='int32')
embedded_question = Embedding(input_dim=10000, output_dim=256, input_length=100)(question_input)
encoded_question = LSTM(256)(embedded_question)
# Let's concatenate the question vector and the image vector:
merged = keras.layers.concatenate([encoded_question, encoded_image])
# And let's train a logistic regression over 1000 words on top:
output = Dense(1000, activation='softmax')(merged)
# This is our final model:
vqa_model = Model(inputs=[image_input, question_input], outputs=output)
# The next stage would be training this model on actual data.
</code></pre>
<h3 id="video-question-answering-model">Video question answering model</h3>
<p>Now that we have trained our image QA model, we can quickly turn it into a video QA model. With appropriate training, you will be able to show it a short video (e.g. 100-frame human action) and ask a natural language question about the video (e.g. "what sport is the boy playing?" -> "football").</p>
<pre><code class="python">from keras.layers import TimeDistributed
video_input = Input(shape=(100, 224, 224, 3))
# This is our video encoded via the previously trained vision_model (weights are reused)
encoded_frame_sequence = TimeDistributed(vision_model)(video_input) # the output will be a sequence of vectors
encoded_video = LSTM(256)(encoded_frame_sequence) # the output will be a vector
# This is a model-level representation of the question encoder, reusing the same weights as before:
question_encoder = Model(inputs=question_input, outputs=encoded_question)
# Let's use it to encode the question:
video_question_input = Input(shape=(100,), dtype='int32')
encoded_video_question = question_encoder(video_question_input)
# And this is our video question answering model:
merged = keras.layers.concatenate([encoded_video, encoded_video_question])
output = Dense(1000, activation='softmax')(merged)
video_qa_model = Model(inputs=[video_input, video_question_input], outputs=output)
</code></pre>
</div>
</div>
<footer>
<div class="rst-footer-buttons" role="navigation" aria-label="footer navigation">
<a href="../faq/" class="btn btn-neutral float-right" title="FAQ">Next <span class="icon icon-circle-arrow-right"></span></a>
<a href="../sequential-model-guide/" class="btn btn-neutral" title="Guide to the Sequential model"><span class="icon icon-circle-arrow-left"></span> Previous</a>
</div>
<hr/>
<div role="contentinfo">
<!-- Copyright etc -->
</div>
Built with <a href="https://www.mkdocs.org/">MkDocs</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>.
</footer>
</div>
</div>
</section>
</div>
<div class="rst-versions" role="note" aria-label="versions">
<span class="rst-current-version" data-toggle="rst-current-version">
<a href="http://github.com/keras-team/keras/" class="fa fa-github" style="float: left; color: #fcfcfc"> GitHub</a>
<span><a href="../sequential-model-guide/" style="color: #fcfcfc;">« Previous</a></span>
<span style="margin-left: 15px"><a href="../faq/" style="color: #fcfcfc">Next »</a></span>
</span>
</div>
<script>var base_url = '../..';</script>
<script src="../../js/theme.js" defer></script>
<script src="../../search/main.js" defer></script>
<script type="text/javascript" defer>
window.onload = function () {
SphinxRtdTheme.Navigation.enable(true);
};
</script>
</body>
</html>
|
