A
accuracy 13
action 391
action-value function 392
ReLU 479
sigmoid 478
tanh 479
ALBERT 214
key intuitions 214
aleatory uncertainty 438
probabilistic neural networks 441, 442
probabilistic neural networks, using 440
AlexNet 95
Android Studio
reference link 589
Application-Specific Integrated Circuit (ASIC) 501
Arduino Nano 33 BLE Sense
reference link 587
Area Under the Curve (AUC) 458, 549
Area Under the Receiver Operating Characteristic Curve (AUC ROC) 458
Arm Cortex-M
reference link 587
Artificial General Intelligence (AGI) 445
artificial neural networks (nets/ANNs) 3
atrous convolution
using, for audio 635
Attention mechanism 182-184, 195-197
full, versus sparse matrices 206
local attention 206
LSH Attention 206
seq2seq model, using with 184-189
Augmented Multiscale Deep InfoMax (AMDIM) 380
convolutional 301-306
deonising 297-301
stacked 301
stacked deonising 367
AutoEncoding (AE) 365
Autograd Module 521
AutoKeras 450
architecture 451
automatic differentiation 495
Automatic Machine Learning (AutoML) 445, 446, 570
achieving 446
automatic data preparation 447
automatic feature engineering 447, 448
automatic model generation 448, 449
pipeline, steps 446
autoregressive (AR) generation 364
GPT-3 365
Image GPT (IPT) 364
PixelRNN 364
WaveNet 366
WaveRNN 366
XLNet 365
autoregressive (AR) models 362
B
backdrop 480
and ConvNets 491
forward propagation 481
purpose 480
backpropagation through time (BPTT) 142, 143, 494
neuron equation, form hidden layer to hidden layer 485-489
neuron equation, form hidden layer to output layer 484, 485
Bahdanau attention (additive) 183
baseline model 569
Batch Gradient Descent (BGD) 491
batch normalization 33
Bayesian Networks (BN) 434-436
Bayesian optimization 450
beginning-of-string (BOS) 172
Bernoulli distribution 428
best practices
for data 564
for model 568
bfloat16 504
Bidirectional Encoder Representations from Transformers (BERT) 133, 207, 366, 367
key intuitions 208
using, as feature extractor 134, 135
bidirectional LSTM (biLSTM) 147
BigBird 211
key intuitions 212
BiLingual Evaluation Understudy (BLEU) 179
binary_crossentropy, objective functions 13
Bootstrap Your Own Latent (BYOL) model 378
bottleneck layer 289
Byte Pair Encoding (BPE) 372
C
Caffe
URL 2
capsule networks (CapsNets) 641
catastrophic forgetting 394
categorical_crossentropy, objective functions 13
causal graphs 434
Central Processing Units (CPUs) 499
chain rule 477
CIFAR-10
performance, improving with data augmentation 84-87
performance, improving with deeper network 82-84
predicting with 87
CIFAR-10 images
recognizing, with deep learning 78-82
classification task
versus regression task 58
CLEVR dataset
reference link 624
CNN architectures 95
AlexNet 95
DenseNets 96
HighwaysNet 96
residual networks 95
CodeSearchNet model 383
using, with TPU 509
collaborative filtering (CF) models 288
colorization 369
concatenative TTS 635
ConceptNet Numberbatch 117
Conditional Probability Table (CPT) 434
content-based attention 183
Context Free Network (CFN) 370
Contextualized Vectors (CoVe) 129
continuous backpropagation
history 473
Continuous Bag of Words (CBOW) 107
Contrastive Divergence (CD) 278
contrastive learning (CL) 207, 362, 373
Deep InfoMax (DIM) 207
instance transformation 376
multimodal models 381
multiview coding 380
Next Sentence Prediction (NSP)
Replaced Token Detection (RTD) 207
Sentence Order Prediction (SOP) 207
training objectives 373
contrastive loss 374
Contrastive Multiview Coding (CMC) 381
convergence 19
ConvNet (1D CNN) 118
ConvNets
summarizing 69
convolutional autoencoder 301
used, for removing noise from images 301-306
Convolutional Neural Network (CNN) 41, 114, 533
classification and localization 614
composing, for complex tasks 613
issue 641
semantic segmentation 615
using, for audio 634
using, for sentiment analysis 632- 634
using, with videos 630
convolution layer configuration, parameters
kernel size 640
padding 640
stride 640
convolution operations 639
depthwise convolution 640
depthwise separable convolution 641
dilated convolution 640
separable convolution 640
transposed convolution 640
cost functions 13
critic network 419
cross entropy 489
custom graph dataset 554
multiple graphs, in datasets 557-559
single graphs, in datasets 554-556
implementing, in TensorFlow 348-356
D
Data2Vec model 383
data best practices 564
feature selection 565
data cleansing 447
data generation
diffusion models, using for 358, 359
flow-based models, using for 356-358
data pipelines
data synthesis 447
D-dilated convolution 636
DeBERTa 217
key intuitions 217
decision boundaries 58
decoder pretraining 206
Deep Averaging Network (DAN) 131
deep belief networks (DBNs) 283
DeepCluster 379
deep convolutional GAN (DCGAN) 329, 330
Deep Convolutional Neural Network (DCNN) 66
ConvNets, in TensorFlow 68
example, LeNet 69
for large-scale image recognition 88, 90
local receptive fields 66
pooling layers 68
shared weights and bias 67
Deep Deterministic Policy Gradient (DDPG) 418, 419
DeepDream network
CIFAR-10 images, recognizing with 78-82
importance 77
Deep Q-Networks (DQNs) 406, 407
used, for playing Atari game 412- 415
Deep Reinforcement Learning (DRL)
Deep Reinforcement Learning (DRL) algorithms 393
action selection, by agent 394
balance, maintaining between exploration and exploitation 394
highly correlated input state space, dealing with 394, 395
moving targets issues, dealing with 395
policy-based methods 393
value-based methods 393
DeepWalk 123
denoising autoencoders 297
used, for clearing images 298-301
DenseNets 96
dependent variable 44
depthwise convolution 640
depthwise separable convolution 641
derivatives 474
differentiation rules 477
diffusion models
using, for data generation 358, 359
Dilated Causal Convolutions 635
dilated ConvNet 636
dilated convolution 640
Directed Acyclic Graph (DAG) 434
distributed representations 105, 106
double DQN 416
DQN variants
double DQN 416
Rainbow 418
dropout
used, for improving net in TensorFlow 19-21
E
edge computing 597
Efficient Neural Architecture Search (ENAS) 449
eigen decomposition 262
ElasticNet regularization 32
ELECTRA 216
key intuitions 216
Embedding Projector tool 265
data panel 265
inspector panel 266
projections panel 265
embeddings
creating, with Gensim 110, 111
Embeddings from Language Models (ELMo) 130
embedding space
exploring, with Gensim 111-113
encoder-decoder architecture 200
seq2seq model 172
encoder-decoder pretraining 206
encoder pretraining 206
end-of-sentence (EOS) 174
entity extraction 631
epistemic uncertainty 438
Evolutionary Algorithm (EA) 449
Evolved Transformer 217
experience replay method 394
exploration vs exploitation tradeoff 394
Exponential Linear Unit (ELU) 8
F
Facebook AI Research (FAIR) 129
False Positive Rate (FPR) 458
Fast Attention Via positive Orthogonal Random (FAVOR) 196
fastText 117
feature clustering 378
feature construction 447
feature map
properties 273
feature mapping 448
features 4
feature selection 447
federated core (FC) 599
Federated Learning (FL) 599
architecture 598
issues 598
TensorFlow FL APIs 599
FeedForward Network (FFN) 533
FigureQA dataset
reference link 624
FlatBuffers 586
reference link 586
Floating-Point Unit (FPU) 502
flow-based models
using, for data generation 356-358
Frechet Inception Distance (FID) 358, 372
function approximator 49
fuzzy clustering 270
G
GAN architectures 339
Gated Recurrent Units (GRUs) 494
Gaussian Mixture Model (GMM) 569
Gazebo
reference link 397
General Language Understanding Evaluation (GLUE) 252
components 252
reference link 252
Generative Adversarial Network (GAN) 322, 323, 372
building, with MNIST in TensorFlow 324-329
Generative Pre-Trained (GPT) model 205
Generative Pre-trained Transformer (GPT-3) model 365, 517, 518
key intuitions 209
Generative Pretraining (GPT) 133
Genetic Programming (GP) 449
Gensim
embedding space, exploring with 111-113
installation link 110
used, for creating embeddings 110, 111
Global vectors for word representation (GloVe) 109, 110, 117
download link 110
GLUE 252, 253
Google Cloud AutoML 451
reference link 451
tables solution, using 451-462
training cost 470
Google Colab
URL 33
GPT-2
key intuitions 209
Gradle
URL 591
graph
basics 532
customizations 551
machine learning 532
Graph Attention Network (GAT) 535
graph customizations 551
custom layers 551
message-passing mechanism 551-554
Graphic Processing Units (GPUs) 134, 499-500
Graph Isomorphism Network (GIN) 536, 537
graph layers 534
Graph Attention Network (GAT) 535
Graph Convolution Network (GCN) 535
GraphSAGE 536
greedy search 154
grid search 450
gRPC Remote Procedure Calls (gRPC) 509
H
H2O
reference link 523
using, for AutoML 523, 524, 525
H2O AutoML 523
H2O model, explain module 526
model correlation 528
Partial Dependence Plots (PDP) 526
variable importance heatmap 527
handwritten digits
experiments, summarizing 31
net, improving in TensorFlow with dropout 19-21
net, improving in TensorFlow with hidden layers 16-19
neural net, defining in TensorFlow 11-15
number of epochs, increasing 27
number of internal hidden neurons, increasing 28-30
one hot-encoding (OHE) 10
optimizer learning rate, controlling 28
optimizers, testing in TensorFlow 22-27
recognizing 10
reconstructing, with vanilla autoencoders 292-295
simple neural net, defining in TensorFlow 13, 14
size of batch computation, increasing 30
TensorFlow net, running 15, 16
hard clustering 270
hard negatives 374
hard update 395
heterogeneous graphs 560
hidden layers
used, for improving net in TensorFlow 16-19
HighwaysNet 96
autotokenization 244
features 242
model, autoselecting with 244
named entity recognition, performing 245
used, for text generation 242-244
using 242
hybrid self-prediction models 370
Jukebox 371
VQ-GAN 372
VQ-VAE 371
hyperparameters 38
I
identity block 96
Image GPT (IPT) AR model 364
Importance Weight Sampling
reference link 566
Inception V3
independent and dependent variables in Machine Learning
reference link 44
independent variable 44
InfoNCE loss 375
Information Retrieval (IR) 104
innate relationship prediction 369
jigsaw puzzles, solving 370
relative position 369
rotation 370
input features 4
instance transformation 376
Bootstrap Your Own Latent (BYOL) model 378
DeepCluster 379
feature clustering 378
SimCLR model 376
SWapping Assignments between multiple Views (SwAV) model 379
Internet of Things (IoT) 506
Item2Vec embedding model 122
J
Jacobian matrix 494
Java Caffe
URL 88
jigsaw puzzles
solving 370
Jukebox 371
K
Kaggle VQA challenge
reference link 624
Keras 3
Keras initializer, 5
Keras MNIST TPU, end-to-end training 510
kernel 66
k-means clustering 266
implementing, in TensorFlow 268-270
working 266
Kohonen networks 271
Kullback-Leiber (KL) divergence 296
L
L1 regularization 32
L2 regularization 32
label smoothing 326
LaMDA 219
language model-based embedding 132, 133
Language Modeling (LM) 207
LASSO 32
latent loss 315
Latent Semantic Analysis (LSA) 104
latent space 315
leaderboard 523
LeakyReLU 9
learning rate 22
learning with a critic 390
left singular matrix 262
LeNet 69
lifted structured loss 375
linear regression 44
multiple linear regression 48
multivariate linear regression 49
neural networks 49
simple linear regression 45-47
used, for prediction 44
logistic regression 59
applying, on MNIST dataset 60-64
reference link 60
Long Short-Term Memory (LSTM) 130, 194, 494
reference link 13
loss minimization 13
LSTM-based autoencoder
building, to generate sentence vectors 306-314
M
Machine Learning (ML) 532
Magenta 638
Magenta NSynth 637
Malmo
reference link 397
many-to-many topology
many-to-one topology
Markov Decision Process (MDP) 393
Markov property 392
Masked Autoencoder for Distribution Estimation (MADE) 358
masked generation models 366
Bidirectional Encoder Representation from Transformers (BERT) 366, 367
colorization 369
stacked denoising autoencoder (AE) 367
Masked Language Modeling (MLM) 207, 366
mathematical tools
chain rule 477
derivatives and gradients 474-476
differentiation rules 477
gradient descent 476
matrix operations 478
vectors 474
matrix factorization 109
matrix operations 478
max pooling operator 68
Mean Opinion Score (MOS) 635
mean squared error (MSE) 288
message function 551
message-passing mechanism 535-554
Message Passing Neural Network (MPNN) 551
method of least squares 45
metrics 14
Mini-Batch Gradient Descent (MBGD) 491
MLOps
reference link 257
MNIST
used, for building GAN in TensorFlow 324-329
MNIST dataset
logistic regression, applying 60-64
mobile neural architecture search (MNAS) 593
mobile optimized interpreter 586
model best practices
AutoML 570
baseline model 569
evaluation and validation 570
model evaluation and validation
model deltas, using 570
patterns, searching in measured errors 571
unseen data, testing 571
user experience techniques 570
utilitarian power 570
model evaluation approaches
few-shot learning 210
one-shot learning 210
zero-shot learning 210
model-free reinforcement learning 392
model generation 448
model improvements
data drift 571
training-serving skew 571, 572
model of the environment 392
mse, objective functions 13
multi-head (self-)attention 198
multi-layer perceptron (MLP) 5, 288
activation functions 9
example 5
Exponential Linear Unit (ELU) 8
LeakyReLU 9
perceptron and solution, training problems 6
ReLU activation function 7
sigmoid activation function 7
tanh activation function 7
multimodal models 381
CodeSearchNet 383
Data2Vec 383
multiple linear regression 48
exploring, with TensorFlow Keras API 53-58
multiplicative( Luong's) attention 183, 184
Multitask Unified Model (MUM) 216
reference link 216
multivariate linear regression 49
exploring, with TensorFlow Keras API 53-58
multivariate normal distribution 433, 434
multiview coding 380
Augmented Multiscale Deep InfoMax (AMDIM) 380
Contrastive Multiview Coding (CMC) 381
reference link 638
MXNet
URL 2
N
Named Entity Recognition (NER) 245
Natural Language Generation (NLG) 205
Natural Language Processing (NLP) 104, 150, 563
negative sampling 108
network
inspections, performing 629, 630
Neural Architecture Search (NAS) 217
neural embeddings 122
Item2Vec 122
Neural Machine Translation (NMT) 195
defining, in TensorFlow 11, 12
for linear regression 49
neurons 3
Next Sentence Prediction (NSP) 366
NLP-progress 255
reference link 255
node2vec embedding model 123-128
Node.js
TensorFlow.js, using with 610
nodes 532
Noise Contrastive Estimation (NCE) loss 375
Non-linear Independent Components Estimation (NICE) 358
normal distribution 431
multivariate normal distribution 433, 434
univariate normal distribution 431-433
normalization
batch normalization 33
N-pair loss 374
NSynth 637
O
objective function 13
one-dimensional Convolutional Neural Network (1D CNN) 118
one hot-encoding (OHE) 10
one-to-many topology
OpenAI 517
OpenAI GPT-3 API 517
examples, reference link 36
reference link 517
tasks 518
random agent, playing Breakout game 401-403
supported environments 398
Open Neural Network Exchange (ONNX) 522
optimizer learning rate
controlling 28
optimizers 12
reference link 13
Optim Module 522
out of vocabulary (OOV) 308
output
predicting 39
overfitting 31
P
Paragraph Vectors - Distributed Bag of Words (PV-DBOW) 132
Paragraph Vectors - Distributed Memory (PV-DM) 132
parametric TTS 635
paraphrase database (PPDB) 117
Partial Dependence Plots (PDP) 526
Part-of-Speech (POS) 150
analysis 631
Pathways Language Model (PaLM) 223
peephole LSTM 147
Permuted Language Modeling (PLM) 207
Pixel Neural Core 506
PixelRNN AR model 364
policy 392
pooling layers 68
average pooling 69
max pooling 68
reference link 69
positional encoding 195
posterior probabilities 437
pre-built deep learning models
recycling, for feature extraction 91, 92
precision 13
prediction 58
linear regression, using with 44
pretext tasks 384
pretrained models 570
pretrained models, TensorFlow Lite 591, 592
audio speech synthesis 591
large language models 596
mobile GPUs 596
pose estimation 594
question and answer 591
segmentation 594
segmentations 591
smart reply 594
style transfer 594
style transfers 591
text embedding 591
pretrained TPU models
decoder pretraining 206
encoder-decoder pretraining 206
encoder pretraining 206
principal component analysis (PCA) 261, 288
implementing, on MNIST dataset 262-264
probabilistic neural networks
for aleatory uncertainty 441, 442
using, for aleatory uncertainty 440
prompt engineering 365
PyTorch 520
modules 520
URL 2
PyTorch, modules
Autograd Module 521
NN Module 520
Optim Module 522
Q
quantization 585
post-training quantization 585
quantization-aware training 586
R
Rainbow 418
random search 450
ReAding Comprehension dataset from Examinations (RACE) 255
Real-valued Non-Volume Preserving (RealNVP) 357
recall 13
Recurrent Neural Network (RNN) 194, 362, 364 533
reduce function 551
Reformer model 210
key intuitions 210
Region-based CNN (R-CNN) 617
Region Of Interest (ROI) 617
Region Proposal Network (RPN) 618
regression model
building, with TensorFlow 439, 440
regression task
versus classification task 58
regularization
adopting, to avoid overfitting 31, 32
used, in machine learning 32
regularizers
reference link 33
reinforcement learning (RL) 389, 390
goal 389
interaction, with environment 389
simulation environments 396
trial and error 389
relative position prediction 369
ReLU (REctified Linear Unit) 7
derivative 479
LeakyReLU 9
Remote Procedure Call (RPC) 509
residual block 96
residual networks 95
REST API
reference link 461
Restricted Boltzmann Machines (RBM) 278, 362
backward pass operation 278
deep belief networks (DBNs) 283
forward pass operation 278
hidden layer 278
images, reconstructing with 279-283
visible layer 278
Retrieval Database (DB) 222
Retrieval-Enhanced Transformer (RETRO) 222
key intuitions 222
return 392
reward 391
Ridge 32
right singular matrix 262
backpropagation through time (BPTT) 142, 143
gradients, exploding 144
gradients, vanishing 143
RNN cell variants 144
gated recurrent unit (GRU) 146
long short-term memory (LSTM) 144-146
peephole LSTM 147
RNN variants 147
stateful RNNs 148
RoBERTa 213
key intuitions 213
Robot Operating System (ROS) 397
rotation
using, as self-supervision signal 370
RotNet model 370
S
SavedModel 587
scaled dot-product attention 184
Scheduled Sampling 178
scikit-learn 114
reference link 114
self-attention mechanism 198
Self-Driving Car (SDC) 391
self-organizing maps (SOMs) 271, 272
used, for color mapping 273- 278
self-prediction 363
autoregressive (AR) generation 364
hybrid self-prediction models 370
innate relationship prediction 369
masked generation models 366
self-supervised learning 363
advantages 363
semi-supervised learning 287
Sensibleness, Specificity, and Interestingness (SSI) 220
hyperparameters, tuning 38, 39
separable convolution 640
seq2seq model 172
using, with Attention mechanism for machine translation 184-189
Sequential() model 4
shallow neural networks 278
Short Message Service (SMS) 114
sigmoid function 7
derivative 479
SimCLR model 376
simple linear regression 45-47
building, with TensorFlow Keras 49-53
simulation environments, for RL
Blender learning environment 397
Gazebo 397
Malmo 397
OpenAI Gym 397
Unity ML-Agents SDK 397
Single Shot Detectors (SSD) 619
singular value decomposition (SVD) 262
Skip-Gram with Negative Sampling (SGNS) model 109
skip-thought vectors 131
soft clustering 270
soft nearest neighbors loss 376
soft update 395
SparkFun Edge
reference link 587
stacked autoencoder 301
stacked denoising autoencoder (AE) 367
Stanford Question Answering Dataset (SQuAD) 254
state 391
stateful RNNs 148
state-value function 392
static embeddings 106
STM32F746 Discovery kit
reference link 587
Stochastic Gradient Descent (SGD) 14, 23, 109, 473, 491
StructBERT 214
content distance 100
style distance 101
sum of squared error (SSE) distance 269
Super Resolution GANs (SRGANs) 339, 340
supervised learning 10
support-vector machines (SVMs)
reference link 617
SWapping Assignments between multiple Views (SwAV) model 379
Switch Transformer 221
key intuitions 222
synthetic dataset
T
tanh function 7
derivative 479
taxonomy
pretraining 207
Teacher Forcing 178
techniques, for augmenting speech data
Frequency Masking 568
Time Masking 568
time warping 568
techniques, for augmenting textual data
back translation 567
synonym replacement 567
TensorFlow (TF) 495
ConvNets 68
CycleGAN, implementing 348-356
features 2
GAN, building with MNIST 324-329
used, for building regression model 439, 440
TensorFlow Datasets (TFDS) 580
data pipelines, building with 583-585
TensorFlow Embedding API 264-266
TensorFlow Federated (TTF)
datasets 600
Federated core (FC) 599
Federated learning (FL) 599
TensorFlow FL APIs
builders 600
models 599
pretrained models, using for inference 577- 580
reference link 621
TensorFlow.js 600
models, converting 607
using, with Node.js 610
TensorFlow Keras
used, for building simple linear regression 49-53
used, for exploring multiple linear regression 53-58
used, for exploring multivariate linear regression 53-58
TensorFlow Keras layers
custom layers, defining 290, 291
TensorFlow Lite 585
architecture 587
FlatBuffers 586
GPUs and accelerators, using 589
mobile converter 586
mobile optimized interpreter 586
pretrained models 591
quantization 585
supported platforms 587
using 588
TensorFlow Probability (TFP) 423-427
distributions 427
used, for handling uncertainty in predictions 437
TensorFlow Probability (TFP) distributions 427
coin-flip example 428
normal distribution 431
using 428
Tensor Processing Unit (TPU) 134, 500
availability, checking 509, 510
fourth generation 506
generations 501
second generation 504
using, with Colab 509
Term Frequency-Inverse Document Frequency (TF-IDF) 104
Text-to-Speech (TTS) 635
Text-to-Text Transfer Transformer (T5)
key intuitions 215
textual data
TFDS dataset
TFHub 250
reference link 250
tf.Keras built-in VGG16 net module
tf.keras.datasets
reference link 11
TFLite Converter 587
TFLite FlatBuffer 587
TFLite interpreter 587
thought vector 131
topic categorization 631
training objectives, CL models 373
contrastive loss 374
InfoNCE loss 375
lifted structured loss 375
Noise Contrastive Estimation (NCE) loss 375
N-pair loss 374
soft nearest neighbors loss 376
triplet loss 374
transfer learning
transformer categories
decoder or autoregressive 205
encoder or autoencoding 205
multimodal 205
retrieval 205
seq2seq 205
transformers
architectures 204
attention mechanism 205
categories 204
cost of serving 257
evaluating 252
future 259
implementations 223
normalization layer 200
optimization 257
pitfalls 259
quality, measuring 252
reference implementation 224-242
residual layers 200
training, via semi-supervised learning 204
transformers optimization 257
knowledge distillation 258
quantization 257
weight pruning 257
Transformer-XL 212
key intuitions 212
transposed convolution 640
triplet loss 374
True Positive Rate (TPR) 458
U
UCI ML repository
reference link 53
uncertainty, in predictions
aleatory uncertainty 438
epistemic uncertainty 438
handling, with TensorFlow Probability 437
synthetic dataset, creating 438, 439
underfitting 32
U-Net
reference link 615
univariate normal distribution 431-433
Universal Language Model Fine-Tuning (ULMFiT) model 133
update function 551
V
value function 392
handwritten digits, reconstructing with 292-295
TensorFlow Keras layers 290, 291
variational autoencoders 314-319, 371
vectorization 104
Vector Processing Unit (VPU) 504
Vector Quantized Variational AutoEncoder (VQ-VAE) 371
vectors 474
Vertex AI 451
VGG16 net
cats, recognizing with 90
videos
classifying, with pretrained nets 630
vision transformers (ViTs) 259
Visual Question Answering (VQA) 622-625
reference link 622
URL 622
vocabulary 158
VQ-GAN 372
W
reference link 637
WaveRNN 366
weight pruning
reference link 258
Winner-Take-All Units (WTUs) 271
algorithm 384
reference link 109
skip-gram architecture 107, 108
word embedding, used for spam detection 114
data, obtaining 115
embedding matrix, building 117, 118
model, evaluating 120
model, training 120
spam classifier, defining 118
spam detector, running 121, 122
wrappers 403
X
key intuitions 213
Y
You Only Look Once (YOLO)
reference link 619