VIP20
本课程属于：005-Artificial Intelligence and Machine Learning Engineering English Technical Courses · 共 20 门相关课程，组合学习更完整。
查看所属VIP全部课程 →
← 返回所属目录 咨询 / 购买

TensorFlow Neural Network Development and Deployment

30 Chapters
View Course Samples Online
01
TensorFlow ecosystem overview
core components, version landscape, and deployment targets.
TensorFlowecosystem
02
Environment setup and reproducibility
virtual environments, GPU configuration, and deterministic training.
Environmentreproducibility
03
TensorFlow data pipelines
tf.data, dataset transformations, batching, prefetching, and input performance.
TensorFlowpipelines
04
Tensor representation and automatic differentiation
tf.Tensor, variables, GradientTape, and custom gradients.
representationautomatic
05
Building dense neural networks with Keras Sequential and Functional APIs
layers, activations, and model summaries.
Buildingnetworks
06
Training workflows
compile, fit, validation splits, callbacks, and checkpointing.
Trainingworkflows
07
Loss functions and metrics for regression, classification, and multi-label tasks.
Loss functions and metrics for regression, classification, and multi-label tasks.
functionsregression
08
Optimizers and learning rate schedules
SGD, Adam, exponential decay, and warmup strategies.
Optimizersschedules
09
Regularization techniques
dropout, weight decay, batch normalization, and early stopping.
Regularizationtechniques
10
Convolutional neural networks
Conv2D, pooling, padding, and feature map visualization.
Convolutionalnetworks
11
Transfer learning and fine-tuning with pre-trained models
feature extraction and layer unfreezing.
Transferfine-tuning
12
Recurrent neural networks and sequence modeling
SimpleRNN, LSTM, GRU, and masking.
Recurrentnetworks
13
Attention mechanisms and Transformers in TensorFlow
multi-head attention and positional encoding.
Attentionmechanisms
14
Text processing pipelines
tokenization, vectorization, embeddings, and tf.text.
processingpipelines
15
Image augmentation and preprocessing
tf.image, Keras preprocessing layers, and tf.data integration.
augmentationpreprocessing
16
Custom training loops with GradientTape
step functions, distributed strategies, and metric tracking.
trainingGradientTape
17
TensorBoard for experiment tracking
scalars, graphs, histograms, and profiling.
TensorBoardexperiment
18
Hyperparameter tuning with Keras Tuner
search spaces, Bayesian optimization, and trial analysis.
HyperparameterBayesian
19
Model saving and loading formats
SavedModel, HDF5, and Keras v3 format.
SavedModelHDF5
20
TensorFlow Serving
model server setup, REST and gRPC APIs, and versioning.
TensorFlowREST
21
TensorFlow Lite for mobile and embedded deployment
converter, quantization, and interpreter.
TensorFlowembedded
22
TensorFlow.js for browser and Node.js deployment
model conversion and inference.
TensorFlow.jsdeployment
23
TensorFlow Extended (TFX) pipelines
data validation, transformation, training, and evaluation.
TensorFlowExtended
24
Distributed training strategies
MirroredStrategy, MultiWorkerMirroredStrategy, and TPUStrategy.
Distributedtraining
25
Mixed precision training
float16, loss scaling, and performance optimization.
precisiontraining
26
Model optimization techniques
pruning, clustering, and quantization-aware training.
optimizationtechniques
27
Deploying models on cloud platforms
Vertex AI, SageMaker, and Azure ML integration.
Deployingplatforms
28
Monitoring and maintaining deployed models
data drift, model decay, and retraining triggers.
Monitoringmaintaining
29
End-to-end case study
building, training, and deploying an image classification service.
End-to-endbuilding
30
End-to-end case study
developing and deploying a real-time text sentiment analysis API.
End-to-enddeveloping