So if I read it correctly, you feel like nothing you have tried had worked in terms of forming meaningful relationships. A lot of people are struggleing to form friendships nowdays. It is hard for me to make helpful suggestions about this, because I’m in a similar position, but I will try. I suggest searching and joining activities where you can meet people who have the same interests as you. I’m sure you will meet people who like you this way. If this is not feasible because of where you live, I suggest changing this. I think, relationships are a key to a happy life for most people. Last but not least I suggest talking about your hobbies for those who ask and never stop trying, Who knows where this will lead you.

I wish ypu the best :).

That sound genuiently sad, but it shouldn’t be. Your hobbies sound totally fine, so I guess something different bothers you…

May I ask what?

Sorry for the question, I feel like holding a conversation today.

Now I’m curious

What are your hobbies?

I agree, but its better than nothing. I’m sorry I mislead you.

Gpt4all is avalible on linux, it is an open source software that can run LLMs locally.

Were you thinking about something like this?

**Tip:** The default source can be referred as `@DEFAULT_SOURCE@` in commands, for example: `$ pactl set-source-mute @DEFAULT_SOURCE@ toggle`

Default sink:

@DEFAULT_SINK@

Ok, that makes sense, thank you.

Why could ram usage be a waste? I thought only the allocation is the performance heavy part, allocated ram does not cost extra performance.

Niceee

I think this should work (I haven’t tested it):

make a new file called default (.) pa in the pulseaudio congfig folder (.config/pulse)

this will mute source_name (replace source_name with your source name):

#include this config file
.include /etc/pulse/default.pa
set-source-mute source_name true

run this command to list source names:

pactl list short sources

Most translators aren’t perfect, they generally can not understand context

you can export to pdf and the text is searchable (in firefox with ctrl f)

I can recommend lmms too

I did manage to write a back-propogation algorithm, at this point I don’t fully understand the math behind back-propogation. Generally back-propogation algorithms take the activation, calculate the delta(?) with the activation and the target output (only on last layer). I don’t know where tokens come in. From your comment it sounds like it has to do something in a unsupervised learning network. I am also not a professional. Sorry if I didn’t really understand your comment.

I have experience in creating supervised learning networks. (not large language models) I don’t know what tokens are, I assume they are output nodes. In that case I think increasing the output nodes don’t make the Ai a lot more intelligent. You could measure confidence with the output nodes if they are designed accordingly (1 node corresponds to 1 word, confidence can be measured with the output strength). Ai-s are popular because they can overcome unknown circumstances (most of the cases), like when you input a question slightly different way.

I agree with you on that Ai has a problem understanding the meaning of the words. The Ai’s correct answers happened to be correct because the order of the words (output) happened to match with the order of the correct answer’s words. I think “hallucinations” happen when there is no sufficient answers to the given problem, the Ai gives an answer from a few random contexts pieced together in the most likely order. I think you have mostly good understanding on how Ai-s work.