Youssef Menjour

After several announcements and postpone, it's finally here !

Two years of hard development we have succeeded in releasing HAIBAL the deep learning toolkit for LabVIEW. 

This toolkit represents more than 4000 VI and we are aware that there are still many things to develop.

Our ambition is to bring a part of the "classic" community of scientific data developers to the LabVIEW language (and why not!).

We still have a lot of things to do in the next weeks, our priority will be the publication of 𝐫𝐞𝐢𝐧𝐟𝐨𝐫𝐜𝐞𝐦𝐞𝐧𝐭 𝐥𝐞𝐚𝐫𝐧𝐢𝐧𝐠 examples and the arrival of more "talking" examples like 𝐘𝐎𝐋𝐎 V5, V6, V7, V8.

𝐊𝐞𝐫𝐚𝐬 and 𝐏𝐲𝐭𝐨𝐫𝐜𝐡 model import tool that will give you access to thousands of models.

The 𝐨𝐩𝐭𝐢𝐦𝐢𝐳𝐚𝐭𝐢𝐨𝐧 𝐨𝐟 𝐨𝐮𝐫 𝐜𝐨𝐝𝐞𝐬 and the acceleration of inferences will also be a top priority to allow you the best Deep Learning experience.

For students we are thinking about a licensing system adapted to your budget (the question is not yet decided internally). 🧑‍🎓

Our goals are ambitious but not impossible and we will make 𝐋𝐚𝐛𝐕𝐈𝐄𝐖 the ultimate deep learning tool.

It is now available for download on the website here

Get started with 𝐇𝐀𝐈𝐁𝐀𝐋, the 𝐋𝐚𝐛𝐕𝐈𝐄𝐖 𝐝𝐞𝐞𝐩 𝐥𝐞𝐚𝐫𝐧𝐢𝐧𝐠 𝐭𝐨𝐨𝐥𝐤𝐢𝐭 by following our new quick beginner tutorial section, to learn how to design, summarize, run, train, size data, and tune parameters with a simple DNN (Dense neural network) model. 🐘

👉🏼 Get started full playlist available on our 𝐲𝐨𝐮𝐭𝐮𝐛𝐞 𝐜𝐡𝐚𝐧𝐧𝐞𝐥
https://www.youtube.com/@graiphic/
👉🏼 Get started with HAIBAL deep learning toolkit (𝐰𝐞𝐛𝐬𝐢𝐭𝐞 𝐝𝐨𝐜𝐮𝐦𝐞𝐧𝐭𝐢𝐨𝐧)
https://lnkd.in/e6cPn4Fq

Thank you for your support, it gives us a boost and makes us want to do even more. 🤠

Thank you

16 hours ago, LogMAN said:

This might be an issue with environment variables. Your application runs with whatever environment variables exist at the time it was launched and cuda is only added to PATH during its installation.

It should work if you specify the path explicitly (or somehow reload environment variables).

How to  reload in CMD environment variable ?

i was 100 % sure you'll try ShaunR 😁

32 minutes ago, ShaunR said:

Snippets don't work very well and the meta information is stripped by some websites and browsers. It is always better to post the code.

code is posted upped (zip file with thewhole project), i put the code vi in attached.

Quote

Probably the "cmd /c where nvcc" isn't finishing for some reason (waiting for input?) and as you have the cmd VI set to wait until completion, it just gets stuck there.

This is not logical, this command works perfectly before installation and it works perfectly after installation and reboot of the exe

Nevertheless, I agree, there is an execution link. 
But then why in the .VI version it works and in the .exe version not at all? 🤔
 

Installer.vi

22 minutes ago, Lipko said:

I think you should show all cases, maybe you would get replies from people you are not able to try your code at the moment (like me). We don't see where/how you set states apart from "4 - Display".

To answers you only need the two states explain upper.

If you want to see all cases you just have drag and drop on your labview diagram the sniped.

Hi everybody,

I have a problem with cmd LabVIEW. I did a state machine and build an exe(So far, so good). 

I did a state machine with a 2000ms timeout. Everytime out i launch a cmd (So far, so good). This command consist to check if nvidia cuda is installed  (All is in the code and So far, so good).

In my state machine i have a state which launch an external installer which will be installed cuda (So far, so good). 

My problem start when cuda is installed during the execution of this state machine (build in exe) and my timeout state does no work (the cmd which check if cuda is installed does not work).

When i restart the exe this state work ! and detect cuda with no problem 🤔

Why do i have to restart my exe to make it work ? (i put the state machine in snip copy and work well in LV version - Not work well when build exe)

CMD inside the exe are blocked since cuda installer was launched, why ?

If you want to do the test, just run the exe in the project, install cuda and when cuda is installed just see that it is not detected.
Restart the exe and there are no worries cuda is detected. 😭

I searched and I still haven't found out why   😭  😭  😭

217806318_SourceInstaller.zip

solution here

https://forums.vipm.io/topic/7076-error-1041/

Hi community,

I try my luck here maybe someone will answer me

With VIPM when i try to package this example with a purchase url we have a proxy error. 

I tested with no proxy --> error, use system proxy settings --> error

Someone know about proxy parameters or what i have to fill to avoid this error ?

Does someone know where is "the licence file to test directory where LabVIEW expects add-on licences" ?

4 minutes ago, hooovahh said:

Probably here somewhere: C:\ProgramData\JKI\VIPM

Found :: many thanks !! 

C:\ProgramData\JKI\VIPM\cache

Is someone know where are stored all these file localy ? 

I was not able to find them

The problem is not solved in the latest version of VIPM !

version 2022.0 --> build 2371

This problem is known since 2 years ! Please wake up JKI ! This functionality is important for who want to make business.

JKI.rar

Ni just modify my resume and i have a point with them today.

10 hours ago, ShaunR said:

You don't need that. That is the API that you could create TPLAT with.

This is the basic procedure

1. Create a product definition (Using LFEdit or via the Softwarekey web page).
2. Create a licence file for the product (Using LFEdit).
3. Apply the licence to the lvlib (Using TPLAT in advanced mode. This will give you a Licenced  version of your code.)
4. Point VIPM to the licence when you build the distribution (If you have the full VIPM, it can also apply it instead of using TPLAT above).

This is the full licencing process (excluding the VIPM).

You can also buy the licensing server from Softwarekey for installation on your own network, if you want everything in-house rather than using theirs (or NI's)

Ok let's do it today.

On 11/16/2022 at 10:35 AM, Antoine Chalons said:

Glad I could help.
I don't think the "app store" like model is something NI is willing to invest effort on, they seem happy to let the community deal with it, from what I could see NI did the bare minimum - which is already not too bad - to give the community a reasonably good solution to handle licenses.

Again I've not used it myself, but I've seen people use, I hope it works out well for you.

I strongly encourage you to contact NI, and more specifically Nancy Henson or Eric Reffett (look them up on linkedin to see their role at NI) to give them feedback about the developer center on community.ni.com.

As soon as i have time i'll will contact them and make feedback. The problem is there is too much disparate document and all are not updated and miss a lot of information.

I download Protection PLUS 5 SDK LabVIEW Edition. This is a labview project. But when i try to use it an error occur. (screen shot : LabVIEW 1,2)
I dont understood how it works.

(download link here)

HAIBAL project started in June 2021.
After multiple postpone we are proud to share our first release note.

Yes already read this document.
If i apply new to solo server you will create an account on softwarekey.ni.com --> only for case 2.

The best way is to use Protection PLUS 5 SDK LabVIEW Edition to generate *.if but for now it does not work and softwarekey are aware on it. I'm waiting there feedback.

Let’s clarify all of these information !

To licence you have two licencing tools (why why whyyyyyyyyy 😆)
If you choose TPLAT you'll have to generate a file *.if then protect your library then construct it. Then you'll be able to manage it 😲

It's a MESS

To resume ... it's a mess 😆

If you want to go through NI information are disparate and hard to have, document date from 2012, 2015, 2018 ... a mess

If you want to do it by yourselves it's worst because Protection PLUS 5 SDK LabVIEW Edition which is useful to generate *.if is not working (download link here) to manage licence have to create an account https://secure.softwarekey.com/

I'm actually in contact with softwarekey to solve issues ...

18 hours ago, ShaunR said:

 

NI have a SoftwareKey Server that they use for Tools Network offerings but you can buy and install your own from SoftwareKey.

 

Thank you ShaunR for the help 

Yes but if you use NI software key you'll have to pay fee to NI.
As i understood there is two different way of thinking this part

Yes, but if you use the NI software key, you will have to pay a fee to NI.
From what I understand, there are two different ways to think about this part

1 - Get started with NI 
Advantage --> Your package will be on ni.com, NI helps you directly and they manage your licensing policy.

Disadvantage: you have to pay a fee to NI (30%). I feel it exessive for long term, they are too greedy and i dont want to sell my package to expensive i will lose customers.

2 - Do it yourself

Advantage: you are free to do what you want. 
Disadvantage --> more complicated to do, not well documented.

18 hours ago, Antoine Chalons said:

I'm not sure it's all up to date, but from the link I gave in my first response you should find all the existing doc & tutorials.

https://forums.ni.com/t5/Developer-Center-Resources/VI-Package-Manager-Licensing-a-Package/ta-p/3491276

I've never had to go through this process, but I know many people who have.

Good luck to you!

Merci Antoine, it was usefull.

I finally found the solution here . Now Third Party Licensing & Activation by NI is distributed on VIPM.

The thing is documentation is not update and can be misleading.

I think NI or JKI need to update there documentation on it. There is a lot of information spread out all over the place and it's not an ease ...

Package is installed and it's not present (after reboot)

Your capture is labview 2013 ...

Thank for your feed back.

After reading and took the time on all these point our conclusion is we will do 2 versions. Lite for free and Pro with activation licence.

So now I would like to add an activation code or an installation key to my installer for deployment.

I found a link to the Third Party Licensing and Activation Toolkit

I followed the instruction, Installed JKI VI Package Manager (VIPM) before installing the Third Party Licensing and Activation Toolkit.

However, I could not find where to start the Third Party Licensing and Activation Toolkit, and could not figure out how to use it. 😰

I'm not the only one with this same problem and on NI forum no answers ...

Is someone already use it and can help me ?

As i understood i have to use it to format my project then i'll will need to link to softwarekey solution to manage the generation and sell part of key.

Thank you for your help.

thanks again for all of these informations. Let me check all of these.

Hi everybody,

Is someone know what possibility exist to distribute a library on LabVIEW. I mean, is there only VIPM tool ?

I would also like to know where I can find a complete guide on licensing and licensing policy. If i want to put licence on my product what are my option ? How it works (licensing tools) ? What are the different types of licenses possible ?

I would like to make several versions of my HAIBAL project with several licensing policies (student, lab and pro) and didn't find a document explaining all of this part.

As this is our first software sale on the NI store (or other platform) and I don't have a lawyer on my team, I have a lot of questions: is it complicated to sell my product online ?

What are the obligations attached to each type of license? Does JKI offer a solution for all this? If so, how much does it cost? Is it only an acquisition cost, or do you then take a royalty on sales?

Is someone can help me on this subject ?

Thanks !

Youssef

On 8/20/2022 at 4:22 PM, Antoine Chalons said:

Do you mean you want to attract python people so they can use ML in LabVIEW?

The idea is to propose not only a simple library on LabVIEW but also to propose an alternative to Keras and Pytorch.

HAIBAL is a powerful modular library and more practical than python.  Be able to run any deep learning graph on cuda or Xilinx FPGA platform with no particular syntaxes is really good for Ai users. 

We are convinced that doing artificial intelligence with LabVIEW is the way of the future. The graphics language is perfect for this.

The promotional video is not aimed at LabVIEW users. It is for those who want to do deep learning in graphical language. We are addressing engineering students, professors, researchers and R&D departments of companies.

LabVIEW users already know what drag and drop and data flow are. LabVIEW users do not need this kind of video.   

14 hours ago, ShaunR said:

Is that some marketing guy's idea of what it would probably look like to use or do you actually have that interface?

This video is for the vulgarisation of our product. We took the inspiration from NI : 

It's marketing and vulgarisation because we address the world of deeplearning which does not formally know the graphic language