Partial Seed Phrase

[leonard01]

Hi, i have 21 out of 24 from my mnemonic - i know the order of the words but dont know where the missing words are in that order. I have xpub. Ive tried running with 'xpub', '--mnenomic-length 24' and the 21 phrase. It wont run as it tells me that there are not enough mistakes permitted. So i reran with same but with same but now '--big-types 3'. I understand there is a huge number of possible phrases to brute force. The issue is that the python script grind to a halt when hitting into the 0.5 billion mark, with RAM usage maxed out (CPU and GPU are low), ive an extremely powerful PC so im a little surprised by this. Two question:1

1. Am i using the righ flags in the correct way?
2. GPU acceleration makes no difference to performance - is this because of RAM bottlenec? ive 64gb ddr5 many thanks!

[jasonhoi]

dont turn on the avoid duplicates feature to avoid memory usage issue

[leonard01]

dont turn on? do you mean not to pass a single --no-dupchecks or to pass it?

[jasonhoi]

my machine is amd 5950x with around 970k password per second, if u want to have others to help u with cracking u can use the tool to generate extract version of the wallet then give it to the helper so they dont gain control of your wallet after cracking successful.

[jasonhoi]

use --no-dupchecks --no-eta should avoid the memory issue

[jasonhoi]

alao if u got more machine u can try to use --worker 1/2 and 2/2 to separate the work to different machine, and --threads to contract mac thread number it may help to not overheating the machine in long run

[leonard01]

ah great, thank you!

[isieditors]

@leonard01 if you don't know the correct location of the 3 missing words, your chances are really low because the speed is not high and total number of permutations is a very big number

[Mrlang10]

忘记3个助记词 但是这个这个软件只能找回两个 ，3个会报错