Shulou(Shulou.com)06/01 Report--

TensorLayer principle and installation is what, I believe many inexperienced people are helpless about this, this article summarizes the causes and solutions of the problem, through this article I hope you can solve this problem.

Tensorlayer Introduction Deep Learning Framework Usage Questions

For deep learning developers, deep learning systems are becoming increasingly complex. So much so that we need lightweight tools to manage the process from start to finish, in order to build more continuous learning into the neural network, this requires us to build more elastic datasets that can be iteratively enhanced and more dynamic models.

Deep learning developers have to spend a lot of time integrating components of neural networks, managing model lifecycles, organizing data, and tuning system parallelism, among other things. As new training samples are used, human insights into neural network models, the ability to update models and track changes in them become essential. To this end, a team at Imperial College London developed a python library to manage the complex iterative process of interdisciplinary developer projects.

TensorLayer was born

To better manage the development process, the team developed TensorLayer, a generic Python library. This library integrates many of the methods involved in the development process, including (operations, model lifecycle, parallel computation, failure), abstracted in modular form. These modules include the following functions:

Used to manage neural network layers

Used to manage models and their lifecycles

for managing the set of data

Workflow modules for fault tolerance.

Keras and TFLearn Disadvantages

While tools like Keras and TFLearn are useful today, they don't have the same complex or even unlimited iterations that networks can have as they grow. They provide the necessary abstraction to lower the barriers to tool use, but they obscure much of the underlying algorithms. While beneficial to developers, the underlying technology becomes relatively difficult to tweak and modify, which is necessary to solve many real-world problems.

TensorLayer Features

Compared to Keras and TFLearn, TensorLayer not only provides high-level abstractions, but also provides an end-to-end workflow that includes data preprocessing, training, post-training processing, and service modules and database management, all of which developers need to build a complete model.

TensorLayer advocates a more flexible and combinable paradigm: neural network libraries can be swapped with native engines. This allows developers to easily leverage pre-built modules without compromising visibility. This non-invasiveness also makes it possible to merge with other TF wrappers such as TF-Slim and Keras. And the development team believes flexibility doesn't impact performance.

Tensorlayer environment build (CPU version)

conda 4.4.10

python 3.5

opencv 3.2.0

tensorflow 1.7.0

tensorlayer source code installation

Installation steps View conda environment conda env list

#Or

conda info -e

conda create -n tensorlayer python=3.5 ipykernel

Enter the tensorlayer environment source activate tensorlayer

Install tensoflow 1.7.0

Install using Tsinghua image source (CPU version)

Website: mirrors.tuna.tsinghua.edu.cn/help/tensorflow/

pip install \

-i https://pypi.tuna.tsinghua.edu.cn/simple/ \

https://mirrors.tuna.tsinghua.edu.cn/tensorflow/linux/cpu/tensorflow-1.7.0-cp35-cp35m-linux_x86_64.whl

install opencv3conda install -c https://conda.anaconda.org/menpo opencv3

Download and install tensorlayergit clone https://github.com/tensorlayer/tensorlayer.git

cd tensorlayer

pip install -e .

Jupyter notebook Load Virtual Environment #Make sure it's in Virtual Environment!!

source activate tensorlayer

python -m ipykernel install --user --name tensorlayer --display-name "Python (tensorlayer)"

Testing Python

>>> import tensorflow

>>> import tensorlayer

After reading the above content, do you understand the TensorLayer principle and how to install it? If you still want to learn more skills or want to know more related content, welcome to pay attention to the industry information channel, thank you for reading!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.