Anaconda vs. TensorFlow

I have asked all my juniors to work with Anaconda and Pycharm only, as this is the best combination for now. Coming to use cases: 1. When you have multiple applications using multiple Python variants, it is a really good tool instead of Venv (I never like it). 2. If you have to work on multiple tools and you are someone who needs to work on data analytics, development, and machine learning, this is good. 3. If you have to work with both R and Python, then also this is a good tool, and it provides support for both.

Incentivized

1. Whenever the problem has the demand for a neural networks based solution, Tensorflow (TF) is a great fit.
2. The tf.dataset API makes it really simple to create complex data pipelines in a few lines of code.
3. tf.estimators API abstracts all the complex computation graph creation logic making it very simple to get started.
4. Eager execution makes it simple to develop a TF graph as debugging the code would be like any other imperative Python program.
5. TF abstracts all the complexities of scaling it to multiple machines. It has various code and data distribution algorithms ready to use.
6. Projects like TensorBoard make monitoring the training process really easy. It also gives the ability to view embeddings without any extra code. Their What-If is extremely useful for poking and understanding a black box model. It also has tools to visualize data to quickly check for anomalies.
7. TF Autograph aims to covert any normal Python code into a distributed program which is quite handy to scale an existing code base.

Incentivized

• Installing packages is very easy with Anaconda. Anaconda comes with 'anaconda navigator', a terminal-like utility from which you can easily install R packages and python libraries.
• Launching R and python IDEs as well as Jupyter notebooks from anaconda navigator is simple, and Anaconda makes it very easy to keep these packages up-to-date.
• I really like the fact that if you don't want to install the full version of Anaconda, you can opt to install a lightweight version (called Miniconda) that includes less python libraries and only core conda. I've installed it when I didn't want to take up as much disk space as Anaconda requires, but it works just the same.

Incentivized

• Data pipeline implementation is quite good, loading large amounts of data and pre-process it in an efficient way is no more issue for us
• It supports all major DL algorithms and network layouts such as ConvNets, RNN, LSTMs, Word2Vec, and even the latest transformer architecture
• The abstraction for the device is perfectly done and its support seamlessly for multiple GPU and even TPU will bring a lot of performance gain for enterprise scoped solution while still keep the flexibility
• The TensorBoard is amazing. I haven't seen a similar thing in other frameworks on the market. It allows us to quickly understand and debug the model with the info visualization which makes understanding much better
• A very supportive community, which is the key for sharing the ideas and find the quick and best solutions

Incentivized

• More graphics need in Spyder book. If you work for couple of years then you will be bored with the graphics.
• Extra tools are required for making it secure. We uses extra tools for adding Username /Password to Jupyter.
• R Studio Hangs a lot when open from Anaconda Navigator.

Incentivized

• It would be much better if they could provide good documentation and easy ways to understand concepts.
• It is difficult to understand the concept behind for example, Tensor Graph, which takes a lot of time.
• As you have to write everything, it is time consuming to write the implementation of whole neural network. It would be better if they can provide some wrapper library to make things easier.

Incentivized

It's really good at data processing, but needs to grow more in publishing in a way that a non-programmer can interact with. It also introduces confusion for programmers that are familiar with normal Python processes which are slightly different in Anaconda such as virtualenvs.

Incentivized

I am giving this rating because I have been using this tool since 2017, and I was in college at that time. Initially, I hesitated to use it as I was not very aware of the workings of Python and how difficult it is to manage its dependency from project to project. Anaconda really helped me with that. The first machine-learning model that I deployed on the Live server was with Anaconda only. It was so managed that I only installed libraries from the requirement.txt file, and it started working. There was no need to manually install cuda or tensor flow as it was a very difficult job at that time. Graphical data modeling also provides tools for it, and they can be easily saved to the system and used anywhere.

Incentivized

Support of multiple components and ease of development.

Incentivized

Anaconda provides fast support, and a large number of users moderate its online community. This enables any questions you may have to be answered in a timely fashion, regardless of the topic. The fact that it is based in a Python environment only adds to the size of the online community.

Incentivized

Community support for TensorFlow is great. There's a huge community that truly loves the platform and there are many examples of development in TensorFlow. Often, when a new good technique is published, there will be a TensorFlow implementation not long after. This makes it quick to ally the latest techniques from academia straight to production-grade systems. Tooling around TensorFlow is also good. TensorBoard has been such a useful tool, I can't imagine how hard it would be to debug a deep neural network gone wrong without TensorBoard.

Incentivized

Use of cloud for better execution power is recommended.

Incentivized

One of the main competitors to Anaconda can be Google products such as Colab. Colab gives you the flexibility to handle large datasets gives it an edge over Anaconda. But again, the ease of access and usability of Anaconda stacks up against Colab. Besides, Anaconda relies more on your machine which makes it safe to use.

Incentivized

Can't seem to choose any deep learning platform in the above, so I'll list it here: 1. Apache MXNet: this has been used for one of our main algorithms for search as an end-to-end pipeline. We chose this because of the Scala bindings, which makes it easier to integrate with out JVM backend. MXNet seems comparable to TensorFlow, although community support is not as good as TensorFlow, and there are issues with memory leaks that are being worked on. TensorFlow in general is easier to use, but MXNet isn't too far behind. 2. Keras: still a favorite. Often I use this when paired with TensorFlow. TensorFlow 2.0 will make it even easier. 3. PyTorch: only used it a little, so it's hard to provide a good opinion. 4. DL4J: used it initially in an early days project because it has good JVM support. Harder to used not because of poor API design, but because community support is lacking and features don't come out as fast as TensorFlow.

Incentivized

• Positive impact - Multiple options for data presenting , visualizing and sharing. (Eg: R-Markdown).
• Positive impact - Ease of access to build complex machine learning models. (I work in NLP, it has multiple built in models to analyze the various contexts).
• Positive impact - Conda package let's to deal with external packages which can be used in Jupyter.

Incentivized

• Positive Impact- As I mentioned before its open source. Very easy to learn for average programmer/ developer. We were able to design a POC model for understanding the patient appointment cancellation snd reasons behind it in 3 week time frame.
• Negative Impact- If you are using tensor flow for small project it works fine. If you are trying to build a model for face recognition it will be hard to program and train the system. It needs data to be processed before hand cannot learn on the go.