#optimizationproblem resultados da pesquisa

Chinmay Kak

13 de set. de

New Blog Post! Covers a lot of stuff I find interesting about optimisers and how they affect model training both in terms of resources and quality of the model. Link Below, please do check it out and RT/QT for the reach:)

ChinmayKak's tweet image. New Blog Post!
Covers a lot of stuff I find interesting about optimisers and how they affect model training both in terms of resources and quality of the model.
Link Below, please do check it out and RT/QT for the reach:)

Mario Nawfal

@MarioNawfal

27 de dez. de

🇺🇸 ELON: STOP TRYING TO OPTIMIZE SOMETHING THAT SHOULDN’T EXIST “Everyone’s been trained in high school and college that you answer the question — convergent logic. You can’t tell the professor 'your question is dumb' or you’ll get a bad grade, you have to answer the question.…

Rishabh

@Rixhabh__

17 h

🚨 Workflows are broken. Teams waste hours every week fighting bottlenecks, miscommunication, and outdated tools. We fixed that. Meet OrgFlow — an AI-powered workflow optimizer built on @rocketdotnew. It finds inefficiency before it costs you time. 👇

Rixhabh__'s tweet image. 🚨 Workflows are broken.

Teams waste hours every week fighting bottlenecks, miscommunication, and outdated tools.

We fixed that.

Meet OrgFlow — an AI-powered workflow optimizer built on @rocketdotnew.

It finds inefficiency before it costs you time. 👇

ueaj

@_ueaj

13 h

New blog post! This one is a purely theoretical one attempting identifying the central reason why LLMs suffer from mode collapse in RL and fail to generate novel or truly diverse outputs. It's actually a way more complicated problem than you think! Naively encouraging…

_ueaj's tweet image. New blog post! This one is a purely theoretical one attempting identifying the central reason why LLMs suffer from mode collapse in RL and fail to generate novel or truly diverse outputs. It's actually a way more complicated problem than you think!

Naively encouraging…

actual poweruser

@actualpoweruser

7 h

Anthropomorphic stuff aside there really is just no good reason to optimize for this sort of behavior @GoogleDeepMind

actualpoweruser's tweet image. Anthropomorphic stuff aside there really is just no good reason to optimize for this sort of behavior @GoogleDeepMind

ℏεsam

@Hesamation

25 de dez. de

OOM error is a pain in the GPU. here this is a fresh article on the visualization of GPU memory and how to predict the required memory of a model

Hesamation's tweet image. OOM error is a pain in the GPU. here this is a fresh article on the visualization of GPU memory and how to predict the required memory of a model

TyrannoSaurav

@TyrannoSaurav

17 de jan. de 2023

Ahh I see the problem is that there are too many if statements. Here's a solution implementing binary search in O(log n)

effectfully

@effectfully

1 de abr. de

Just implemented an obvious optimization that I've been meaning to implement for years. It made our code 40% slower.

Michael Cohen

@Michael05156007

11 de out. de 2024

New paper! Over-optimization in RL is well-known, but it even occurs when KL(policy || base model) is constrained fairly tightly. Why? And can we fix it? 🧵

Michael05156007's tweet image. New paper! Over-optimization in RL is well-known, but it even occurs when KL(policy || base model) is constrained fairly tightly. Why? And can we fix it? 🧵

Jonathan Müller

@foonathan

20 de set. de 2023

C++ optimization in a nutshell: stackoverflow.com/questions/5043… #cpp #cplusplus

WPOptimizers

@WPOptimizers

18 min

Ayer subí una imagen y noté que el rendimiento de mi web WordPress empeoró. Esto me recordó que mantener una web optimizada requiere un cuidado constante.

Duncan Ogilvie 🍍

@mrexodia

30 de set. de 2024

🤔Allocations in your C program too expensive? 💡Optimize them with this one simple trick!💅

DAN KOE

@thedankoe

11 de set. de 2024

Optimize for work you want to do, with people you care about, on problems that actually impact people's lives, even if that impact is small. If you don't, you'll end up with the opposite by default.

Maskara.ai

@MaskaraAi

24 h

This prompt reveals what you're actually optimizing for (hint: it's probably not what you think)

OxTøchi 🦇🔈

@OxTochi

24 de out. de

the problem: global compute is centralized the fix: cysic the solution is one word dawg, don't complicate things

Abhishek🌱

@Abhishekcur

27 de dez. de

Really great article for optimization really great article on optimization for software developers -multi-variable optimization is an incredibly useful tool for writing software -maths with cs stuffs

Abhishekcur's tweet image. Really great article for optimization
really great article on optimization for software developers
-multi-variable optimization is an incredibly useful tool for writing software
-maths with cs stuffs

BigTFromAZ

@BigTFromAZ

25 de out. de

I have seen this many times in my career. When optimizing, spend 90% of your time tuning the top 10% of anything in terms of volume. Spend the remaining 10% on the other 90%. This is true for database queries, program methods and algorithms.

Branko

@brankopetric00

24 de out. de

Developer kept complaining our RDS instance was too slow. Wanted to upgrade from db.t3.large to db.r5.8xlarge. Cost increase: $140/month to $3,400/month. I checked query performance first: - Query ran 47,000 times per hour - No indexes on the WHERE clause columns - Missing…

Ben Dicken

@BenjDicken

18 de nov. de

Js people: Ok your language is actually fast Python people: Come defend your language Go people: Needs more optimizations C people: Don't worry. If you enable the optimizer, this thing goes BRRRRRRRRRRRR.

javamuse🐧

@_axtone

22 de out. de

🧵 Optimizar antes de medir es la forma más común de romper un sistema estable. Veamos por qué la optimización prematura destruye más software del que mejora 👇