Is there an overview, how the config object for the constructors of the neural network are defined?

This would be interesting:

• Has brain.NeuralNetwork and brain.recurrent.RNN the same config attributes?
• What is the description of the inputSize attribute?

Thank you for the great work with this library, it is realy amazing!

I created a virtual environment using conda and I installed darkflow and it worked pretty well. I later exported the environment to try it on another computer to verify that everything works. But conda doesn't install the darkflow even though it exists on the requirement.txt

首先感谢楼主的分享，对于basic_seq2seq中的代码，运行时产生如下错误，我的tensorflow是最新的1.3.0 gpu版本，执行时所有代码均未改动，想楼主帮看看
Traceback (most recent call last):
File "D:\Workspaces\Eclipse\PythonLearn1\src\seq2seq_init_.py", line 227, in
num_layers)
File "D:\Workspaces\Eclipse\PythonLearn1\src\seq2seq_init_.py", line 189, in seq2seq_model
decoder_input)
File "D:\Workspaces\Eclipse\PythonLearn1\sr

Can you describe what modifications need to be done if I want to replace dynamic_rnn with tf.keras.RNN in many-to-one example as dynamic_rnn is deprecated now.

• Explain in notebook/FAQ what non-maxima suppression is what values to set (threshold on IoU)
• Explain and provide code how to pick a good score threshold (reuse Patrick's plot which was implemented for the drone demo)

Most of the people who start out new don't find a latest feed of community hyped resources on ML and DL topics. It would be pretty good if we add a Reddit section.
If you're fine with this suggestion I'll put up a PR with the update 😄

Can you please add some performance numbers to the main project docs indicating inference latency running some common hardware options e.g. AWS p2, GCP gpu instance, CPU inference, Raspbery pi, etc.

The procedure for downloading the "en_vectors_web_lg" in spacy. by downloading and unzipping the file, and shifting it to the appropriate directory, as illustrated here is l

Since many ML tasks are done on a remote server (and possibly, the cloud), an option to prevent browser from opening should be added in the launch function.
Also , what about an upload option to test new images ?

For further development of the project, it would be great to have unit tests to ensure the project code does not break with further changes/improvements. The tests go under tests folder in 2.0 branch where Python 3+ porting is happening.

At the moment the random initialisation is baked into the UpdateLayer class.

We could replace this with either Variates from System.Random.MWC or Random.

I am not clear about what the iteration means. Could you explain more about it? Also, how is it related to epoch?

Thanks.