Gonzalo (@goncheck) 's Twitter Profile
Gonzalo

@goncheck

Tech insider | AI & Cloud | Insights que importan 💡

ID: 493805189

calendar_today16-02-2012 06:31:40

15,15K Tweet

268 Followers

700 Following

Jia-Bin Huang (@jbhuang0604) 's Twitter Profile Photo

New video! How do LLMs grow outrageously large yet blazingly fast? The secret: Mixture of Experts (MoE) In this video, we cover the role of FFNs, how to scale them up without slowing down, and how to maintain load balance and stability. Full video here: youtu.be/0QQlYR1r6pQ

New video! How do LLMs grow outrageously large yet blazingly fast?

The secret: Mixture of Experts (MoE)

In this video, we cover the role of FFNs, how to scale them up without slowing down, and how to maintain load balance and stability.

Full video here: youtu.be/0QQlYR1r6pQ
IvĂĄn (@ivaanblanco26) 's Twitter Profile Photo

Qué sangre fría para tirar a la escuadra el penalti que puede clasificar a tu equipo a los Octavos de Final de la Champions.

Juan FernĂĄndez (@7uanf) 's Twitter Profile Photo

Se hicieron cĂĄlculos que concluyeron que el plan de $200 de Claude estĂĄ tan subsidiado que podrĂ­a equivaler a $5000 en poder de cĂłmputo Hay que trabajar en soluciones propias, locales y en la medida de lo posible: gratis, para estar preparados cuando se termine el subsidio

Don Uhr (@donuhr24) 's Twitter Profile Photo

El Omega Constellation es de esos iconos raros. Todo el mundo lo reconoce pero casi nadie lo lleva. No es el reloj de siempre, ni de un cierto perfil de gente. Una buena oportunidad para aquellos que se quieren distinguir.

El Omega Constellation es de esos iconos raros.

Todo el mundo lo reconoce pero casi nadie lo lleva.

No es el reloj de siempre, ni de un cierto perfil de gente. Una buena oportunidad para aquellos que se quieren distinguir.
Gonzalo (@goncheck) 's Twitter Profile Photo

En vista de lo difĂ­cil que es estar al dĂ­a en el mundo de la Inteligencia Artificial, he creado un newsletter diario con las noticias mĂĄs relevantes. SuscrĂ­bete. đŸ‡Ș🇾 theaipulse.me/es 🇬🇧 theaipulse.me/en

Anish Moonka (@anisha_moonka) 's Twitter Profile Photo

Every time you message an AI chatbot, the model stores your entire conversation in temporary memory called a KV cache (a cheat sheet so it doesn’t re-read everything from scratch). On a large model like Llama 70B running a long conversation, that cache alone eats 40GB of GPU

Teja Karlapudi (@teja2495) 's Twitter Profile Photo

Google could have kept this algorithm private instead of open sourcing it. But they allowed everyone, including competitors, to benefit from their research. That is why I respect Google. If you did not know, this entire AI era became possible because Google openly published

Clint (@flinttrades) 's Twitter Profile Photo

I have a secret to share After your first $2–$3 million, a paid off home and a good car, there is no difference in quality of life between you and Jeff Bezos. Both of you have limited amount of time on earth; you have twice if not more than Jeff, so you are richer than him. A

ZZPotxolin (@zpotxolin) 's Twitter Profile Photo

EstĂĄs aburrido porque no haces misiones secundarias, tĂ­o. La vida es algo mĂĄs que solo trabajar y tirarte en la cama sin hacer nada. AquĂ­ tienes 50 misiones secundarias para completar:

Andrej Karpathy (@karpathy) 's Twitter Profile Photo

LLM Knowledge Bases Something I'm finding very useful recently: using LLMs to build personal knowledge bases for various topics of research interest. In this way, a large fraction of my recent token throughput is going less into manipulating code, and more into manipulating