Wir helfen Unternehmen seit 17 Jahren,
bessere Software zu finden
Über TensorFlow
Flexible Open-Source-Bibliothek für maschinelles Lernen für Forscher im Bereich des maschinellen Lernens.
The software provides mainstream training model, prediction model, mainstream ML framework to accelerate the efficiency of our project development. Low price, suitable for early learning and research.
It is fairly difficult at first, as it brings the whole complexity of working with machine learning. It is very resource-driven and thus the only viable option is using it in the cloud.
Nutzerbewertungen filtern (103)
Nutzung
Sortieren nach
Nutzerbewertungen filtern (103)
TensorFlow is useful, although it requires a healthy time commitment to produce accurate models
Kommentare: The benefits I received from this software is more accurate modeling and an interesting insight into what makes one software better than another. TensorFlow did for me what it says it does - produce high quality models, such as neural networks, with a lot of human capital input.
Vorteile:
TensorFlow is fascinating in seeing how it produces results in a reasonable time frame. It is completely flexible compared to its costly competitors. The software connects well with various data sources and in setting up scripts to run automatically.
Nachteile:
TensorFlow takes a lot of time to become an expert in what it is doing. The programming time-commitment might not be worth it unless you plan on customizing your modeling to work with other software.
In Betracht gezogene Alternativen:
Relatively Straightforward Deep Learning Framework
Kommentare: Human pattern recognization, image recognization. Habits and trends.
Vorteile:
The 2.0 version is easy to set up and there are a lot of APIs that are integrated for using various programming languages to do the same thing. I personally have been using python with this application and have had very little problems getting going. There are a lot of tutorials on getting started, some good data available for free to assist with the learning process. Everything can be run locally which makes it easy to expand on-site. Cloud options are also affordable.
Nachteile:
The learning curve is a bit steep. This isn't specifically an issue because of TensorFlow itself, the idea of neural networks are not simple. TensorFlow has made improvements on 2.0, that make it easier to use compared to previous versions.
Review of Google Cloud ML Engine
Kommentare: My overall experience with Google Cloud ML platform was very good. I used it's machine learning services to integrate those in my web applications.
Vorteile:
The feature of the Google Cloud ML Engine that I most like is the machine learning features that have been provided by this platform. The ML features of this engine provide SOTA results in every task in machine learning and artificial intelligence. The ML features are very handy and easy to use and integrate in other applications as well. I would recommend everyone to use Google Cloud ML Engine for developing AI systems.
Nachteile:
The pricing, when exceeded the free tier of Google Cloud ML platform, is high. The pricing is high compared to other services like Azure Cloud ML platform.
Deep learning Bestfriend!
Vorteile:
Tensorflow helps me build, train and test models in machine learning and Deep learning. With its commpatibilty to create Deep learning neurons for training purpose and having methods to directly apply it makes tensorflow the best to pursue!!
Nachteile:
So far tensorflow helps even beginners to use it easily with a number of tutorials and documentations making it less likely to have any thing not to like or havee any complaints to users like me.
Feedback
Kommentare: good product
Vorteile:
easy to use. good performance .integration with python is easy
Nachteile:
expensive. AI tools need to be more graphically represented
A Machine and Deep Learner must have Library
Vorteile:
This Library is very flexible for doing Matrices and Tensor So building very deep high level but quick and scalable ready to use neural networks is at your finger tips. The added other Anaconda Library and Keras compatibility
Nachteile:
Depreciation of the code is frustrating. To use one form just to throw a Error message.
Very helpful in the new world of machine learning.
Kommentare: You will learn a lot from TensorFlow. It is a good way of entering the machine learning world.
Vorteile:
I used TensorFlow on AWS which was easier with all the infrastructure AWS built. It was a good start to machine learning with all the AI and neural network popularity going on these days. It was challenging and exciting to prepare datasets, train them and see the satisfactory results in dashboard. It is also open source and this gives an advantage to TensorFlow.
Nachteile:
There is a long and challenging learning period. Documentation is rich but it would be so much better to learn and use it with some visual aids.
My Review of TensorFlow
Kommentare: My overall experience is very good using TensorFlow to develop AI models.
Vorteile:
I like that TensorFlow has a version that runs on GPU which is very useful when applying Machine learning. Also, I like that TensorFlow is updated regularly to support different libraries and with new features.I like that TensorFlow supports all the project lifecycle from building and programming to deployment.
Nachteile:
I don't like that TensorFlow requires expertise as it is not easy for beginners. Also, TensorFlow has a slow speed which is not good in deploying deep learning models compared to other frameworks.
Best performance for ML tasks
Kommentare: I often work with ML engine, and it appears very complex to me. Because of that I suggest Newbies to start with AutoML first.
Vorteile:
ML and AutoML by google dramatically simplify work of Machine Learning developers, in my opinion. Google provides a complete infrastructure that can import export, train and deploy model within the ML environment. On the other hand AutoML provides even more simplicity with operations.
Nachteile:
It is often difficult to implement ML solution and require time and efforts that are not always available due to certain constraints.
Incredibly powerful
Kommentare: The framework has been amazing for me both for getting into machine learning and for developing more advanced projects.
Vorteile:
The software is not the easiest to grasp but there are myriad amounts of documentation and examples online which can help with most situations. The Github repo is also well maintained with references to any bugs and problems that one may encounter
Nachteile:
Debugging is incredibly difficult with version 1 of the framework (this is meant to be addressed in version 2) and can take a long time to get a handle of the particular concepts. The complete library is exhaustive but to the point of abstracting certain concepts too much.
Strong tool for deep learning
Kommentare: Using TensorFlow has been a powerful but difficult adventure into deep learning
Vorteile:
TensorFlow is great because it handles data well, supports many deep learning models, works smoothly across different devices, and has a cool tool for visualization called TensorBoard.
Nachteile:
Even though it's good, TensorFlow can be hard to learn, uses different terms, builds models in a fixed way, and sometimes its guides are out of date.
Tensorflow is the future of our business, and likely the future of machine learning modeling.
Kommentare: Tensorflow is the future of machine learning modeling. There is no way around that and we as a company are fortunate to bring this technology to the forefront.
Vorteile:
Tensorflow is the easiest way to implement machine learning software into your product/business. The repository is colossal and there is an abundance of support within the community alone. Tensorflow is updating regularly and will continue to grow in the years to come.
Nachteile:
Hardware is a common bottleneck in machine learning software. We have built out dedicated computing space just for our tensorflow models and will have to continue to upgrade and expand that space. It's just the nature of the business.
A powerful deep learning library that will save your time
Kommentare: Tensorflow is a great deep learning library with so many tutorials. I would like to recommend this to anyone who is willing to use deep learning framework.
Vorteile:
I have applied machine learning with tensorflow library for my university research. I was able to train deep neural network models. It is an open source library which provides visual growth of neural network in the tensorboard. There are multiple algorithms supported by tensorflow like classification and regression. Also this supports many languages like python, JavaScript and Ruby. There is a huge support group for tensorflow and I got a great support from them while I was doing my research.
Nachteile:
It takes some considerable time to understand things and learn things. Complexity is my only concern.
ML must have
Kommentare: Definitely my first option when neural networks are involved in my personal and professional research. Furthermore it also has High-Level API (Keras) which make everything easier.
Vorteile:
Mastering tenworflow unlocks in you all the possibilities you can have with the current Machine Learning techniques. Data Science, Computer Vision, Machine Learning, NLP...name any area of research of AI, tensorflow can easily handle it.
Nachteile:
It might be a little too complicated at first, especially if you are a beginner of Neural Networks. But sticking with it will give you the ability to interact with it.
A powerful high-level machine learning library!
Vorteile:
Tensorflow is a high-level machine learning library. I can use it to design neural network structures without writing C++ or CUDA18 code in order to get high efficiency. It supports automatically calculating derivative. Tensorflow is implemented with C++ and it uses C++ to simplify online deployment. In addition to C++ interface, it also provides us with Python, Java and Go interfaces.
Nachteile:
Although Python is very powerful and easy to use, using Python with TensorFlow will still cause some efficiency problems. For example, every mini-batch needs to be fed from Python to the network. During this process, when the data size of mini-batch is small or calculation time of is short, it will cause long latency.
Best ML Library
Kommentare: I'm totally satisfied of the usage and can recommend it to anyone who is doing a project on machine learning and AI.
Vorteile:
Simplicity, that is the most liked feature for me. As a beginner for machine learning it helped me to create my project without any issues and the support I got was enormous. Also the code labs that were given helped me to understand it better with ease.
Nachteile:
Well I trained my data set using the CPU it took a lot of time by then. Other than that I never faced any issue of using it.
TensorFlow is a efficient tool for Machine learning task
Kommentare: I built my machine learning project using TensorFlow. Initially, I faced some installation issues in windows. Then I installed and used it in Linux. It was compatible with Linux and easy to learn.
Vorteile:
It has a rich visualization facility and frequent updates to add new additional features. Tensorflow is simpler than other libraries like Torch and Theano.
Nachteile:
It’s not speeder than other libraries and it makes problems in the windows operating system. It also doesn’t collaborate with other frameworks like OpenCL.
Tensorflow for Deep Learning Applications
Kommentare: It
Vorteile:
Tensorflow is an open source library for deep learning algorithms. In case I had problems with Tensorflow implementation or deployment, I found wide support from the Tensorflow community. Tensorflow comes with tensorboard, which a great tool to visualize the learning process and to track the progress of your application in terms of the accuracy and the gradients.
Nachteile:
As starting to learn the basic deep learning tools using Tensorflow, I found it not straight forward in terms of the sessions and variables management. It is quite tricky though to debug the code if it has some problems. Also, TesnorFlow does not support dynamic graphs. It was not an issue for me in the beginning, however, it started to be a challenging problem while dealing with dynamic graphs (e.g. text modeling).
TensorFlow for school project
Vorteile:
I did have chance to use couple of tensorFlow features for my university project, i figured it has lots of features and applicability especially those computation graphs within NN
Nachteile:
My problem was that i had a hard time understanding how I'm able to do what im looking for, more examples with a better defined documentation might be helpful, as a learner it was hard for me
Extensive and versatile machine learning library
Kommentare: Convolutional neural networks for multi-dimensional arrays (2 to 5 dimensions)
Vorteile:
Very good documentation present online. Integrated very well with Google Colab. I like that both beginners and experts use this software.
Nachteile:
Very hard to get started initially. I was struggling a lot at first. But when you get used to it, it's not that bad. I also wish that there were less bugs. Sometimes my network doesn't compile although there is nothing wrong with it.
Walking in AI ground walk with tensorflow
Kommentare: We use TF for machine learning implementations and predict sales of the year, we are not experts yet but TF is so popular that every single time we have doubts we get the right tutorial just googling.
Vorteile:
Tensorflow is fairly easy to learn and understand, even though it documentation is not the best, tensorflow is so popular so you have infinite options to learn. Tenaor flow is faster than other libraries like keras or theano.
Nachteile:
Setting up the data takes time and is not that easy like other libraries such as keras or theanos. Sometimes it uses many lines of code for easy actions that can be done in one or two lines with other library
Tensorflow is the best open source library for machine learning framework
Kommentare: Helped me to create various machine learning models
Vorteile:
Some of the pros are as follows: 1. Compatible across various platforms like GPU, CPU and TPU. 2. Better computational power, performance and graphical visualization than Theano, Torch etc. 3. Keras could be used as backend for the same. 4. It could be used on various devices from cell phones to powerful supercomputers.
Nachteile:
Some of the cons are as follows: 1. Lack of Symbolic loops (which is given in Theano and Caffe) 2. Lack of support on Windows.
Gold standard in the machine learning world
Kommentare: Using machine learning in several projects. They all needed tensorflow even when I am not directly using it, but rather using other machine learning packages.
Vorteile:
Developed by Google, Tensorflow is now the go to package for most machine learning projects that need to manage computations with huge amounts of calculations and data. Everyone uses it so you can't go wrong.
Nachteile:
Huge learning curve. It is very difficult to just pick up. Most people have to use other packages built on top of tensorflow.
An extensive open source library for Deep Learning and AI
Kommentare: It is a great tool for folks in research, academics and industry working on the next generation of machine learning applications
Vorteile:
I have used TensorFlow to solve computer vision problems in my research. It offers an extensive open source platform to train datasets and deploy ML models and neural networks.
Nachteile:
It can be challenging as a beginner to get used to writing the code and to understand the tensor concepts
Good Features and Learning
Vorteile:
They have visually appealing components. Support from google is best feature that tensorflow have. It supports wide varierty range of operations.
Nachteile:
current open source implementation does not support distributed computaion and windows support is not present. In future update more guide for new users should be added.
