input ‘1zS’ string 3 characters // character is encoded 58
output ‘65’ line 2 characters // character is encoded 58
conditionally there are 10,000 lines
1zS; 65 # 1zS input; 65 output
Ad9; 1S # Ad9 input; 1S output
task we load => ‘tdS’ predict ‘Gh’ for example
As it is something very specific to that repository I suggest you to open a ticket directly there for support.
Ok, so I think that you need to be a little bit more specific.
Do you have a few lines of a dummy example to expose your issue?
this pict show not good generate
examle — predict on 10 base number = 61 ==> to 58 base == 13 or 14
need preditc 10 base 1 - 58
Can you explain what is your goal? What problem do you want to solve with TF?
theoretical possibility of a reverse function from the BTC public key
I don’t know any reasearch baseline on this topic and I have many doubts that you can approach on this with ML.
I assume that if the correct data is loaded, the model will produce the desired result
thus the task should be split into a generator of correct data
and training the model on this data
probably need a model and data evaluator
it’s may be good? but i’m not understand how use in my examle
It was not related to use that in your example, it is an academic example with code on how Deep learning could be used to mitigate some specific attacks in public-key cryptosystems.
As I’ve told you my opinion is that you cannot use directly ML for your original idea for the random component in public-key cryptosystems.
Read the paper of the code I’ve already metioned:
I don’t know anything specific for BTC.
More in general if you are interested to be introduced in these research topics (not strictly in cryptocurrencies) Elie Bursztein at Google has some nice blog posts recently like this one:
i’m try create model
pls check if possible