Rhys Campbell
184 posts

Rhys Campbell retweetledi

@shadcn Everyone’s busy praising them, but they’re clearly squeezing margins from everywhere—and could end up killing tools like shadcn and Figma
English


@MankyDankyBanky @hnshah Part time model builder, full time looksmaxxer
English

I trained a 12M parameter LLM on my own ML framework using a Rust backend and CUDA kernels for flash attention, AdamW, and more.
Wrote the full transformer architecture, and BPE tokenizer from scratch.
The framework features:
- Custom CUDA kernels (Flash Attention, fused LayerNorm, fused GELU) for 3x increased throughput
- Automatic WebGPU fallback for non-NVIDIA devices
- TypeScript API with Rust compute backend
- One npm install to get started, prebuilt binaries for every platform
Try out the model for yourself: mni-ml.github.io/demos/transfor…
Built with @_reesechong. Check out the repos and blog if you want to learn more.
Shoutout to @modal for the compute credits allowing me to train on 2 A100 GPUs without going broke
cc @sundeep @GavinSherry
English

@theo “And you can actually save.. oh you can only copy it because they’re app sucks” hahaha 😂
English
Rhys Campbell retweetledi
Rhys Campbell retweetledi

@auboomers Are you actually retarded?
Houses back then cost like 70k for fuck sake, the average home now is around 900k to 1m, our wage has maybe doubled, while the prices of homes has almost 9x’d . And that’s not including how interest comes into the equation.
English

He wants one logical answer. Here it is. Stop renting and buy. We did it on $41,500 combined in 1983. The median Australian earns double that now. Twice the income and somehow worse off? I'd be looking at the lifestyle before I blamed the system 🏡
BroBro🇦🇺🏇🏻@realRick_AUS
The median Australian earns between 80-90k a year, Someone tell me how they’re suppose to afford a home, an electric car and installation equipment required? Not including all the other bills they have to pay. Just one logical answer would be nice.
English

@Matt_Pinner That man has no peace and quiet, and dinner isn’t even ready
English

@iruletheworldmo Just 10x that shit and make it 100t models and manipulate time
English

first class of 10t models trained on blackwells.
spud will follow. gemini next month. grok a few months after.
you need to prepare yourself psychologically for one obvious fact.
this does not stop.
more compute. more efficiency. more breakthroughs. more intelligence.
now we’ve entered a point, the new models will accelerate all of these accelerants.
the government won’t act in time so we are in the hands of demis, sam, dario, elon, and mark.
may they be kind.
martin_casado@martin_casado
Mythos appears to be the first class of models trained at scale on Blackwells. Then will be Vera Rubins. Pre-training isn't saturated. RL works. And there is *so much* computing coming online soon. Buckle your chin strips. It's going to be fucking wild.
English



















