Me neither… so it isn’t normal? CPU i7-9700k, GPU RTX3070, 32GB RAM.
Don’t realy see how the two can be compared. Can there be semantic without syntax?
EDIT: I have been unable to reproduce the issue with the export times. Since first encountering there have been several PL and Windows updates. HQ vs DeepPrime XD is now actually about 3x faster (instad of 7x slower).
Well as I’ve written. In the case of “AI” its really about syntax, which refers to the rules that govern the structure of a language. It deals with the way words are arranged to form sentences and how those sentences are structured. Syntax is concerned with the grammatical correctness of a sentence, regardless of its meaning. But “AI” has no understanding of what the words mean. Hence it has no understanding that is necessary to engage in semantics - meaning of words, phrases, and sentences. Semantics deals with the interpretation of language and how words and sentences convey meaning.
Can there be semantic without syntax? Obviously, people are capable engaging in both syntax and semantics, but it is possible for people to understand each other even when they make a lot of grammatical errors, typos, weird and unconventional structures of sentences etc. Semantics is concerned with the meaning of a sentence, regardless of its grammatical correctness.
Everyone understands Yoda speak for example or something similar, even if its not correct syntax.
In Yodish…
Semantic without syntax, can there be? Obviously, capable engaging in both syntax and semantics, people are. Possible for people to understand each other even when they make a lot of grammatical errors, but it is, typos, weird and unconventional structures of sentences etc.Concerned with the meaning of a sentence, semantics is, Regardless of its grammatical correctness.
Off course, often both syntax and semantics are prefferable… unless off you make mistakes, but the osther side still ubndsersnds the point you are trseyng to make. Wink, wink.
Otherwise yes, ChatGPT is just a statistical database expressed as a directed acyclic graph plugged in to an (Natural language processing) NLP. I imagine the compiling of data needs much higher specs.
Natural Language Processing (NLP) involves programming computers to process and analyze large amounts of natural language data. OpenAI’s Generative Pre-trained Transformer, known as GPT-3, has demonstrated ability to synthesize text based on
NLP combines computational linguistics—rule-based modeling of human language—with statistical, machine learning, and deep learning models. Together, these technologies enable computers to process human language in the form of text prompts or voice data and to synthesize a response in the form of an answer.
But much like psychopaths can mimic the emotions of other people, even when they don’t feel it themselves, so can ChatGPT mimic emotions that it does not feel, and engage in syntax without semantics. Depends on the programming.
Hq should be very fast with your configuration. Really faster than DeeprimeXD.
There’s something wrong to figure out.
I suppose when comparing all other settings are the same ?
Didn’t want to bump the thread that has veered off topic So had just edited my last post.
I have been unable to reproduce the issue with the export times. Since encountering it there have been several PL and Windows updates. HQ vs DeepPrime XD is now actually about 3x faster (instead of 7x slower).
And yes, when it was an issue, I did use all the same settings and only changed the type of noise reduction. After it happening several days in row, I thought it was normal and didn’t try again until now.