@chantelchantel39: It lasted 2 minutes I was back talking

Chantelchantel39
Chantelchantel39
Open In TikTok:
Region: US
Saturday 25 October 2025 01:07:03 GMT
453
37
3
2

Music

Download

Comments

perfectly_her
PerfectlyHer :
Yeaaaa law
2025-10-25 01:44:10
1
bumpbeezie
BUMP BEEZIE :
😂
2025-11-01 23:22:03
0
denisejackson25
Denise Jackson :
🥰🥰🥰
2025-10-28 02:54:30
0
To see more videos from user @chantelchantel39, please go to the Tikwm homepage.

Other Videos

#chatgpt #gpt4 #artificialintelligence I still think it's a great tool, but it has to be used correctly and with caution. I work with Tensorflow and non language models a good amount, so if someone has real world experience with Transformers believes I'm wrong, please let me know. No need to panic YET is all. 1: I mean aren’t you worried chatGPT is gonna take your job? 2: at its current state? No why would I be? 1: well yeah the current state can get a bunch of things wrong, but didn’t you hear about GPT 4? 2: Yeah I heard, what about it? 1: They’re saying it has 100 trillion parameters versus the 175 billion for gpt3, I mean look at this image. 2: well yeah but do you know, do you know what a parameter is? 1: what do you mean? It’s the number of parameters in its model. 2: Yeah but just because it has more parameters doesn’t mean the model does better. 2: this is a neural network and each of these lines is called a weight and then there are also biases and those are the parameters. 2: the bigger the model is, the more parameters it has. 1: ye exactly so it’s bigger which means it’s better. 2: you know dolphins and elephants have brains 10 times the size of our brains right? 2: Here look at this image classification model, you give it an image and it tells you if that image is a pizza or a steak. 2: we have model 1 here that has thirty one thousand parameters, and a validation set accuracy of 78%. 2: and then we have model 2 with fifteen million parameters, but the validation set accuracy is about 54%.  2: in other words you can also toss a coin and guess if the image is a pizza or a steak based on heads or tails, but it has soo many parameters. 1: I guess we’ll see who’s right later, if GPT 4 is smarter I win. 2: but I never said it’s not going to be smarter, I just said you don’t need to make everyone panic before even seeing it.  Person 2: I mean at any second there could be a solar flare that takes millions of lives, why don’t you start being worried about that instead?
#chatgpt #gpt4 #artificialintelligence I still think it's a great tool, but it has to be used correctly and with caution. I work with Tensorflow and non language models a good amount, so if someone has real world experience with Transformers believes I'm wrong, please let me know. No need to panic YET is all. 1: I mean aren’t you worried chatGPT is gonna take your job? 2: at its current state? No why would I be? 1: well yeah the current state can get a bunch of things wrong, but didn’t you hear about GPT 4? 2: Yeah I heard, what about it? 1: They’re saying it has 100 trillion parameters versus the 175 billion for gpt3, I mean look at this image. 2: well yeah but do you know, do you know what a parameter is? 1: what do you mean? It’s the number of parameters in its model. 2: Yeah but just because it has more parameters doesn’t mean the model does better. 2: this is a neural network and each of these lines is called a weight and then there are also biases and those are the parameters. 2: the bigger the model is, the more parameters it has. 1: ye exactly so it’s bigger which means it’s better. 2: you know dolphins and elephants have brains 10 times the size of our brains right? 2: Here look at this image classification model, you give it an image and it tells you if that image is a pizza or a steak. 2: we have model 1 here that has thirty one thousand parameters, and a validation set accuracy of 78%. 2: and then we have model 2 with fifteen million parameters, but the validation set accuracy is about 54%. 2: in other words you can also toss a coin and guess if the image is a pizza or a steak based on heads or tails, but it has soo many parameters. 1: I guess we’ll see who’s right later, if GPT 4 is smarter I win. 2: but I never said it’s not going to be smarter, I just said you don’t need to make everyone panic before even seeing it. Person 2: I mean at any second there could be a solar flare that takes millions of lives, why don’t you start being worried about that instead?

About