Dosa Maximizer

Something about everything. Everything about something. My reward function is shaped to maximize dosas!
AI/ML, Data and Dosas 🥘 RT != Views

[ADHD @ADHDGoat]

this girl lowkey has wiped Sydney Sweeney from all social media in the last two weeks

[Dosa Maximizer @Brombadil]

@kingofknowwhere @snr_boost DM?

[Dosa Maximizer @Brombadil]

@kingofknowwhere @snr_boost You interested in new opportunities?

[Dosa Maximizer @Brombadil]

@Laz4rz @lilianweng Dropped you a msg

[Dosa Maximizer @Brombadil]

@Laz4rz @lilianweng Cool. DM?

[Dosa Maximizer @Brombadil]

@NielsRogge @stetsblake @levelsio @FAL @replicate What’s the requirement for inference?

[Dosa Maximizer @Brombadil]

Stick it to him Yann!

[Yann LeCun @ylecun]

Join xAI if you can stand a boss who:
- claims that what you are working on will be solved next year (no pressure).
- claims that what you are working on will kill everyone and must be stopped or paused (yay, vacation for 6 months!).
- claims to want a "maximally rigorous pursuit of the truth" but spews crazy-ass conspiracy theories on his own social platform.

[Dosa Maximizer @Brombadil]

@svpino What's a good document layout detection+ parsing model? That's open-source preferably

[Dosa Maximizer @Brombadil]

@nickfloats @character_ai Wait there's no API at all??

[Sasha Rush @srush_nlp]

If you know Torch, I think you can code for GPU now with OpenAI's Triton language.

We made some puzzles to help you rewire your brain. Starts easy, but gets quickly to fun modern models like FlashAttention and GPT-Q.

Good luck! https://t.co/psu31KpdR6

[Dosa Maximizer @Brombadil]

@victorialslocum Does it work with local LLMs?

[Dosa Maximizer @Brombadil]

@dofelite @Suhail You claiming 40gigs of SRAM and you're saying you can't think of a single way to demonstrate that capability???

[Dosa Maximizer @Brombadil]

We really did go from 230MB SRAM to 44GB SRAM in one month lol

[Dosa Maximizer @Brombadil]

1) what

[Cerebras @cerebras]

📣ANNOUNCING THE FASTEST AI CHIP ON EARTH📣

Cerebras proudly announces CS-3: the fastest AI accelerator in the world.

The CS-3 can train up to 24 trillion parameter models on a single device. The world has never seen AI at this scale.

CS-3 specs:
⚙ 46,225 mm2 silicon | 4 trillion transistors | 5nm
⚙ 900,000 cores optimized for sparse linear algebra
⚙ 125 petaflops of AI compute
⚙ 44 gigabytes of on-chip memory
⚙ 1,200 gigabytes of external memory
⚙ 21 PByte/s memory bandwidth
⚙ 214 Pbit/s fabric bandwidth

📰 Press Release: https://t.co/xxKBZ2nKYA

👨‍🎓 Learn More: https://t.co/3A7iq9q5n3

📢 Contact us: https://t.co/IUMEORD4Jb

#AIChip #GenerativeAI #ML #AI #Supercomputers #LLMs #AICompute

[Dosa Maximizer @Brombadil]

@levelsio Post-AGI post-Fusion would most certainly be post-planetary shortly afterwards

[Dosa Maximizer @Brombadil]

@drummatick Similarly we found that a tuned MLP works better than a Transformer model for our usecase (neural operator learning) 🥲

[Dosa Maximizer @Brombadil]

@beffjezos Humility doesn't play nice with valuation

[Dosa Maximizer @Brombadil]

@jrosell @andromeda74356 @jeremyphoward Unless temp=0, how would the model return the exact completion both times?