Deep_Learning_Frameworks
0 likes12 views
The document discusses setting up an environment to develop deep learning (DL) models using Keras and TensorFlow in Python. It explains that Python, TensorFlow, and Keras need to be installed. TensorFlow and Keras can be installed using pip. The document also provides an overview of popular DL frameworks, distinguishing between low-level frameworks like TensorFlow and high-level frameworks like Keras, which allows developing full DL models with less code. Keras is highlighted as a good high-level framework to use due to its simplicity and support for other frameworks as a back-end.
Deep_Learning_Frameworks
- 1. Keras and TensorFlow Back End By:Mohamed Essam
- 2. Setting Up the Environment We will be developing DL models with the Keras stack using TensorFlow as a back end in Python. Hence, to get started we need to set up our playground environment by installing Python, a few important Python packages, TensorFlow, and finally Keras. Let’s get started.
- 3. Installing Keras and TensorFlow Back End Now that Python is set up, we need to install TensorFlow and Keras. Installing packages in Python can be done easily using the pip, a package manager for Python. You can install any Python package with the command pip install package-name in the terminal or command prompt. So, let’s install our required packages (i.e., TensorFlow and Keras).
- 4. Decomposing a DL Model In its most basic form, DL models are designed using neural network architecture. A neural network is a hierarchical organization of neurons (similar to the neurons in the brain) with connections to other neurons. These neurons pass a message or signal to other neurons based on the received input and form a complex network that learns with some feedback mechanism
- 5. Decomposing a DL Model In its most basic form, DL models are designed using neural network architecture. A neural network is a hierarchical organization of neurons (similar to the neurons in the brain) with connections to other neurons. These neurons pass a message or signal to other neurons based on the received input and form a complex network that learns with some feedback mechanism
- 7. Exploring the Popular DL Frameworks
- 8. Low-Level DL Frameworks Given the level of abstraction a framework provides, we can classify it as a low-level or high-level DL framework. While this is by no means industry recognized terminology, we can use this segregation for a more intuitive understanding of the frameworks. The following are a few of the popular low-level frameworks for DL. EG: Theano, Torch,PyTorch,TensorFlow
- 9. High-Level DL Frameworks The previously mentioned frameworks can be defined as the first level of abstraction for DL models. You would still need to write fairly long codes and scripts to get your DL model ready, although much less so than using just Python or C++. The advantage of using the first-level abstraction is the flexibility it provides in designing a model. Eg: Keras…
- 10. High-Level DL Frameworks The previously mentioned frameworks can be defined as the first level of abstraction for DL models. You would still need to write fairly long codes and scripts to get your DL model ready, although much less so than using just Python or C++. The advantage of using the first-level abstraction is the flexibility it provides in designing a model. Eg: Keras…
- 11. Keras Keras is a high-level neural network API written in Python and can help you in developing a fully functional DL model with less than 15 lines of code. Since it is written in Python, it has a larger community of users and supporters and is extremely easy to get started with. The simplicity of Keras is that it helps users quickly develop DL models and provides a ton of flexibility while still being a high-level API. This really makes Keras a special framework to work with. Moreover, given that it supports several other frameworks as a back end.
- 12. CREDITS: This presentation template was created by Slidesgo, including icons by Flaticon, and infographics & images by Freepik Thank you