Sign in
Categories
Your Saved List Partners Sell in AWS Marketplace Amazon Web Services Home Help

TensorFlow 1.6 Python 3.6 CPU Production

Jetware | 180306-tensorflow_1_6_0-python_3_6_3

Linux/Unix, Amazon Linux 2017.09.1 - 64-bit Amazon Machine Image (AMI)

Reviews from AWS Marketplace

0 AWS reviews
  • 5 star
    0
  • 4 star
    0
  • 3 star
    0
  • 2 star
    0
  • 1 star
    0

External reviews

59 reviews
from G2

External reviews are not included in the AWS star rating for the product.


    Computer Software

Too messy

  • May 28, 2019
  • Review provided by G2

What do you like best?
The static computational graph is a very good concept. The API usually has implementations available for almost everything you would need.
What do you dislike?
The documentation is not very good. The API is too messed up - there are several functions that do the same thing with minor differences and little documentation about the differences. Boilerplate code is also usually long. The API is cumbersome to use overall.
What problems is the product solving and how is that benefiting you?
I have used tensorflow for several projects related to text classification.


    Mining & Metals

Explore ia

  • May 22, 2019
  • Review provided by G2

What do you like best?
It s open source, and very hse full in deeplearning
What do you dislike?
Some complixity it's hard for biginner to use it
What problems is the product solving and how is that benefiting you?
Image classification


    Research

Excellent library for Deep Learning

  • April 28, 2019
  • Review provided by G2

What do you like best?
The intuitive way in which all the methods interact. It is an extremely user friendly wrapper for deep learning projects and research. Enables quick prototyping of the models.
What do you dislike?
There are major changes coming. People might have to relearn some of the ways in which to use the library. With each update there are some really annoying deprecation warnings.
What problems is the product solving and how is that benefiting you?
I have been using TensorFlow extensively for all my projects in my graduate level Deep Learning course. I even use it for my research.
Recommendations to others considering the product:
It is a great tool to prototype models for research and projects. However, with all the deprecation warnings happening, it might be better to learn Keras.


    Computer Software

My experience with Tensorflow as a grad student

  • February 21, 2019
  • Review provided by G2

What do you like best?
Heavy machine learning models can be built on top of the optimized deep learning models integrated by Tensorflow. Compatibility with other frameworks makes it possible to optimize on the production level.
What do you dislike?
The models can be a little heavier to deploy in mobile applications when the question is of limited space. Thus the execution is also slow. If Google can divide them into several modules making space changes depending on the device.
What problems is the product solving and how is that benefiting you?
In our organization, we used variations of the TensorFlow Inception model mainly for image classification and image categorization of randomized data to compare performance with other libraries in order to publish our research.


    Tanmayan P.

Easy implementation of neural networks

  • February 16, 2019
  • Review provided by G2

What do you like best?
Most of the functionality is abstracted to a very high level. For example, I just need to run the optimizer.optimize() command to Backpropagate over my entire network
What do you dislike?
Sometimes, debugging is very difficult, as unless a TF.session is started, we cannot see the variable values
What problems is the product solving and how is that benefiting you?
Neural network design and testing new algorithms


    Computer Software

Machine Learning Trend

  • January 09, 2019
  • Review provided by G2

What do you like best?
1: Broad libraries
2: Support
3: Documentation
What do you dislike?
1: Example that the tutorial use.
2: Tutorial.
What problems is the product solving and how is that benefiting you?
I use TensorFlow for analyze data.


    Semiconductors

Great library for performing deep learning

  • December 02, 2018
  • Review provided by G2

What do you like best?
Great documentation written by Google's Tensorflow Team, and a lot of community engagement. Most answers regarding Tensorflow have answers available from the community.
What do you dislike?
Library interface is a bit lower-level compared to other Python interfaces like Keras, but developer experience has been improving.
What problems is the product solving and how is that benefiting you?
Performing supervised learning on business problems requiring predictions. We have realized cost savings using the results of our predictions.
Recommendations to others considering the product:
Useful for deep learning.


    Lucifer M.

Write Neural Networks in under 50 lines

  • December 02, 2018
  • Review provided by G2

What do you like best?
- Tensorflow is one of the best frameworks to do deep learning, machine learning for huge datasets.
- The dataset can have millions of records can be of terabytes in size
What do you dislike?
- Tensorflow can be slow when working with large files. A GPU might be required to help with the processing
What problems is the product solving and how is that benefiting you?
I created a neural network on CDC sources datasets to run predictions. Once I worked on many diseases and created individual model, I developed a predictive model platform to run several neural networks at once for multiple diseases.


    james t.

very intuitive and powerful platform.

  • December 02, 2018
  • Review provided by G2

What do you like best?
Offers great platform to develop in the AI domain to deliver business value efficiently.
What do you dislike?
None at this time, as we are still in POC.
What problems is the product solving and how is that benefiting you?
AI proof of concept in the legal and compliance space to help formulate AI strategy.


    Banjo O.

TensorFlow Review

  • December 01, 2018
  • Review provided by G2

What do you like best?
Lots of things you can do for creating models
What do you dislike?
Very complex, has there are alot of manual steps
What problems is the product solving and how is that benefiting you?
Machine Learning in the cloud