yswnth

manifestation is simple. just write it and do. now mark it done.

same for me. happy to see it.

just one 15-10 match chess. before

i’m so sleepy today. so just sleeping early.

@xnishidh cleanly presented!!!

Before Overfitting. In machine learning and especially in supervised machine learning. Every dataset has a feature vector and a label which is a class, every feature vector belongs to a particular class. we use this feature vector to generate training, validation and testing dataset. Sometimes needs to make the training dataset balance we use RandomOverSampler() such that the more will not dominate. collectively multiple feature vectors are called a feature matrix. We will feed the training feature vectors to our ML model and then will compare the prediction with the actual check for the losses and make the adjustments to the model for the validation. I drew three graphs that depicts the decision boundary line of our ML model between two classes (white & yellow). let's understand overfitting with these graphs down below:

climate!!!

simple walk.

so i build this video using a software build by @sankalpa_02 and it's freaking good. huge shoutout to him. you can also make this type of videos easily link in the comment.

@code_kartik @sankalpa_02 damnnn

@Vixhal dont. nxt time.

@yswnth sowlly :(

Dear YouTube, I don't want the same videos (shuffled) when I hit the refresh button.

so i built it used Step 3.7 Flash to power a startup research agent and gave it a single startup idea. the workflow needed multiple searches, competitor research, pricing analysis, funding research and market positioning before it could produce anything useful. what stood out was how well Step 3.7 Flash kept the entire process together. instead of treating search, reasoning and output as separate steps, it was able to carry context through the workflow and turn all of that information into something actionable. video below ↓

@lochan_twt banger

thank you Google AI

@ravikiran_dev7 pro

iphone 17 pro or iphone pro max

@adityadotdev how can i?

Prove me that you are not ai

@sattyyouneed either one of those leads and the other one follows. right now it's anthropic time

anthropic bf, openai gf

@Misbahtwts both were so good. if they have the depth effect of leaves and all, it would be much more amazing

Which one is better

somebody point me to wispr flow alternativess

@damnGruz so much change. we didnt expected

what are these benchmarks bro i'm genuinely scared 😭😭

actually what if native apple voice typing is just as good as wispr flow?

@byteHumi you want to be in hydration debt or what? 😭😭

I just want to experiment fable 5 (Mythos) + ultra code (xhigh) + fast mode + /workflow + /goal + /loop how fast it can drain my token limits (it will be a wild run lads)