I doubt people will convert your code to C+/Rust for free, so i suggest you check these options,
1. Use PyPy (https://www.pypy.org) which improve your code performance without change single line of your code.
2. Use another compiler such as Numba (https://numba.pydata.org/). I never tried Numba though since you need to modify your code and read the documentation.

Deleting lines and pushing the changes doesn't completely delete them, because they are still accessible in older commits.

You should make a copy of all the repo's files (except for .git/ folder), and then force-push it to the remote which will delete the entire commit history like this:

Well for starters, numpy is written using Cython and only uses a really basic amount of C language. So you can start by reading Cython's tutorials: https://cython.readthedocs.io/en/latest/src/tutorial/index.html.

But if you don't know much C++ then if you wanna port programs to it you're going to have to at least learn C, the subset of C++. Raw C is slightly faster than using C++'s std library anyway and only has a few syntax/library functions you need to learn, the higher-level functions you might need can be swiped from Github and Stack Overflow.

Before you proceed further, you should delete your mail credentials on your code. I suggest you move it to the .env file, and change your current one, assuming you unintentionally put your real password.

I can't read your README since it's all in Russian but for starters you could add Cython to your python project, write your python files in that syntax and then compile them down into a C++ library, with a simple entry point, something like a main() function. Then dynamically load it from a small, simple, C++ program.

Need Advice on a Risk Management Question ?
So when we eventually  find the Private Key using BitCrack for any of these Bitcoin challenge transactions,
What will be the best/safest method to import the private key into your own wallet in the quickest possible time ?
eg : To an Exodus Digital Hot wallet or to a cold storage offline wallet or bitcoin core address

I heard that its possible to use transaction fees to redirect an import of private keys to another wallet
since there will be so many hackers sniffing for transactions/keys from these wallets trying to intercept transfers.
Is this possible or highly unlikely ?

Thanks 

http://www.sensoria.org.uk/wp-content/uploads/2019/07/the-adventures-of-robin-hood-1200-1200-675-675-crop-000000.jpg

Apologize for the late reply to this comment, but @yoyodapro, what "multiplication issue" are you referring to? I've read this entire thread and do not recall an issue like that being mentioned. While I am using a 3090, I was able to compile the current build of bitcrack against CUDA 11.2 but updating the references from CUDA 10.1 to 11.2. While cubitcrack still gives the misaligned address error, clbitcrack seems to work fine, albeit at a slower rate. But curious what issue you are referring to so I can understand how it may or may not affect what I am working on.

Are you talking about the points being chosen by an RNG? I am not totally sure how long does the process of generating random numbers takes because I haven't measured it, but I'd say it could take a noticeable percentage of the time it takes to check a private key.

I still feel like pregenerating a bunch of keyspace ranges beforehand and passing those to bitcrack serially is a better approach at randomization than generating a random number 2^X times per second (X is an arbitrary number proportional to the speed of your GPU).

Anyway it would be nice if you linked that version of Bitcrack you're talking about so we can take a look at it's commit history to see what it changed.

Thanks for the help a ton, and only changed the c_c to 35 in my case. made the solution file work as a charm,
otherwise no issues yet.
is there a verbose mode on this bitcrack as of no others have that, like show where it is now instead of continue file.
changed to 1024 now.
thanks guys

He is probably talking about Pica's version with random feature. Many say it had some bugs in it. I neve used it.

My version generates millions of random starting points, in the bit range you desire, and then counts sequentially, from each of those random starting points. So if you only want to look in the 8F00000000000000 range, it will generate millions of starting points all starting with 8F, and then compute sequentially. It doesn't take long to generate the points, no longer than generating Kangaroos. It's fast at finding keys in any 2^48 range, but above that, It's better to just search sequentially starting with one key. IMO. So if we took #64 puzzle, and assigned people the front 2^16 range, example 8F00, and another person 8F01, and another person 8F02, etc. all the way to 8FFF, then we only need 256 people to check the 8F range quickly. Now, once that range is done, start with another one like 80, 81, 82, 83, C0, C1, etc.
Who's in?! lol 

Very interesting, was looking for a tool like this.
Definitely going to give it a try.