You can find Loras of celebrities on Civit AI but I think you cannot use them for any commerical purpose as they have not provided permission. You can just use them for educational purpose or memes and only if they are non-offensive.
How hard would it be for anyone to figure out the number of 'I's' in the word Mississippi when you actually feed the word ? It's like you're literally feeding the word and asking for the spelling, it would make a lot more sense if u were to say it out orally.
LLMs don't think in a way we do, they do not take in alphabets directly and group them into tokens so the word does not go into it letter by letter, so for example the word 'Strawberry' is broken into smaller tokens like str, aw, berry like that and these are like single units for LLMs. So the ability to count each individual alphabet while getting input token is a big step in reasoning.
hello, i have a question how can we create ai images of celebrities ?
You can find Loras of celebrities on Civit AI but I think you cannot use them for any commerical purpose as they have not provided permission. You can just use them for educational purpose or memes and only if they are non-offensive.
@xclbrxtra thankyou for the information
How hard would it be for anyone to figure out the number of 'I's' in the word Mississippi when you actually feed the word ? It's like you're literally feeding the word and asking for the spelling, it would make a lot more sense if u were to say it out orally.
LLMs don't think in a way we do, they do not take in alphabets directly and group them into tokens so the word does not go into it letter by letter, so for example the word 'Strawberry' is broken into smaller tokens like str, aw, berry like that and these are like single units for LLMs. So the ability to count each individual alphabet while getting input token is a big step in reasoning.
@@xclbrxtra ah alright makes sense.