Scribd
Upload
11 views13 pages

Tensorflow On Cloud: Shilpa Das

TensorFlow is an open-source machine learning framework developed by Google Brain, designed for building, training, and deploying models across various platforms. It supports a wide range of applications, offers an end-to-end platform, and includes features like automatic differentiation and multi-language support. TensorFlow is utilized by major cloud service providers such as AWS, Azure, and Google for various machine learning tasks including natural language processing and computer vision.

Uploaded by

Shopno Banerjee
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
11 views13 pages

Tensorflow On Cloud: Shilpa Das

TensorFlow is an open-source machine learning framework developed by Google Brain, designed for building, training, and deploying models across various platforms. It supports a wide range of applications, offers an end-to-end platform, and includes features like automatic differentiation and multi-language support. TensorFlow is utilized by major cloud service providers such as AWS, Azure, and Google for various machine learning tasks including natural language processing and computer vision.

Uploaded by

Shopno Banerjee
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 13

TensorFlow on Cloud

Shilpa Das
TensorFlow
• TensorFlow is an open-source framework for machine learning (ML)
and artificial intelligence (AI) that was developed by Google Brain. It
was designed to facilitate the development of machine learning
models, particularly deep learning models, by providing tools to easily
build, train, and deploy them across different platforms.

• TensorFlow supports a wide range of applications, from natural


language processing (NLP) and computer vision (CV) to time series
forecasting and reinforcement learning.
Key Features
• End-to-End Platform - It provides tools for every stage of the machine learning
lifecycle, from data preparation to model deployment.

• Flexible - It supports various machine learning tasks, including deep learning,


and can be used across different platforms (servers, cloud, mobile, edge
devices, browsers).

• High-Level APIs - It offers intuitive, high-level APIs for both beginners and
experts to create machine learning models.

• Open-Source & Scalable - It's freely available and community-driven. It can


handle large datasets and complex models.
Key Features
• Versatile: It can be used for a wide range of applications, including natural language
processing, image recognition, and more. Automatic Differentiation (Autograd) - TensorFlow
automatically calculates gradients for all trainable variables in the model, which simplifies the
backpropagation process during training. This is a core feature that enables efficient model
optimization using techniques like gradient descent.

• Multi-language Support - TensorFlow is primarily designed for Python, but it also provides
APIs for other languages like C++, Java, and JavaScript, making it accessible to developers with
different programming backgrounds.

• TensorFlow Serving and TensorFlow Model Optimization - TensorFlow includes tools for
serving machine learning models in production environments and optimizing them for
inference, allowing for lower latency and higher efficiency.
TensorFlow Workflow
• Building a machine learning model in TensorFlow typically involves the following steps:

• Step 1: Define the Model - Create a model by specifying the layers, activation functions, and output layers. Choose
an optimizer, loss function, and evaluation metric.

• Step 2: Prepare the Data - Load, preprocess, and split the dataset into training, validation, and test sets. Data
preprocessing might include normalization, reshaping, and encoding categorical variables.

• Step 3: Train the Model - Feed the training data into the model and perform forward and backward passes. Use
optimization algorithms such as stochastic gradient descent (SGD) or Adam to minimize the loss function.

• Step 4: Evaluate the Model - After training, evaluate the model’s performance on the validation and test sets to
assess its generalization ability.

• Step 5: Deploy the Model - Once the model is trained and evaluated, deploy it to a production environment using
TensorFlow Serving, TensorFlow Lite, or other deployment tools.
Key Components in TensorFlow
• Tensors: Tensors are the fundamental units of data in TensorFlow. They are
multi-dimensional arrays or matrices used for storing data. A tensor can have
one dimension (vector), two dimensions (matrix), or more dimensions.

• Graph: A TensorFlow graph represents a computation as a flow of tensors


through a series of operations. Each operation in the graph performs a
specific mathematical function on the input tensors, such as matrix
multiplication, addition, or activation.

• Session: A session in TensorFlow runs the computation defined by the graph


and evaluates the tensors. This is where the actual execution of the model
happens, enabling the training and inference processes.
Low Level Architecture of TensorFlow
High-level Architecture in TensorFlow
• The first layer of TensorFlow consists of the device layer and the network layer. The device
layer contains the implementation to communicate to the various devices like GPU, CPU,
TPU in the operating system where TensorFlow will run. Whereas the network layer has
implementations to communicate with different machines using different networking
protocols in the Distributable Trainable setting.

• The second layer of TensorFlow contains kernel implementations for applications mostly
used in machine learning.

• The third layer of TensorFlow consists of distributed master and dataflow executors.
Distributed Master has the ability to distribute workloads to different devices on the
system. Whereas data flow executor performs the data flow graph optimally.
High-level Architecture in TensorFlow
• The next layer exposes all the functionalities in the form of API which
is implemented in C language. C language is chosen because it is fast,
reliable, and can run on any operating system.

• The fifth layer provides support for Python and C++ clients.

• And the last layer of TensorFlow contains training and inference


libraries implemented in python and C++.
High-level Architecture in TensorFlow
Applications
• Natural Language Processing (NLP): Building models for tasks like
text translation, sentiment analysis, and chatbots.

• Computer Vision: Developing models for image recognition, object


detection, and image generation.

• Time Series Forecasting: Predicting future values based on historical


data.

• Reinforcement Learning: Training agents to make decisions in


complex environments.
TensorFlow Ecosystem
• TensorFlow Ecosystem -

• TensorFlow Lite: For deploying models on mobile and edge devices.


• TensorFlow.js: For building and deploying machine learning models in JavaScript.
• TensorFlow Federated: For training models on decentralized data.
• TensorFlow GNN: For building neural networks on graph data.
• TensorFlow Datasets: A collection of ready-to-use datasets.

• Languages - TensorFlow can be used with Python, JavaScript, C++, and Java.
TensorFlow by Cloud Service Providers
• AWS TensorFlow Support - Amazon SageMaker, AWS Lambda

• Azure TensorFlow Support - Azure Machine Learning, Azure


DataBricks

• Google TensorFlow Support - Vertex AI, TensorFlow Cloud,


DeepLearning VM

You might also like