Problem Solver for engineers who want to get stuff done | Sh%#-posting
127.0.0.1
Joined May 2011
- Tweets 5,153
- Following 216
- Followers 220
- Likes 2,154
KAY 🍃 retweeted
Spring Boot 4's null-safety in action! 🎯
The real problem? In Java, nullness has always been implicit. When you see User findUserByEmail(String email), can it return null? Who knows? 🤷
Spring Boot 4 changes the game with NullMarked (powered by JSpecify + NullAway). One package-level annotation makes nullness explicit in your code. The IDE instantly flags potential NPEs at compile-time, no more guessing, no more surprises in production.
The tooling guides you to handle nulls properly, exactly where they matter. Your type signatures now tell the truth. This is the developer experience we've been waiting for!
KAY 🍃 retweeted
we don't have to live this way. senior engineers have the duty of building guardrails and teaching the next generation of programmers the practices that lead to high quality software.
AIs may write most of the code, but you still decide what ships. we should never lose our taste for quality software.
i made this analogy a while back: everyone can cook at home but yet restaurants still exist. the floor is rising, but so is the ceiling. the best software will always be made by those who care.
KAY 🍃 retweeted
I was fairly certain Anthropic banned me.
But no:
Claude is down so hard that even logging in doesn't work... and they don't tell you on login, but instruct you to contact support. Really should do better communicating an outage...
The security vulnerability we found in Perplexity’s Comet browser this summer is not an isolated issue.
Indirect prompt injections are a systemic problem facing Comet and other AI-powered browsers.
Today we’re publishing details on more security vulnerabilities we uncovered.
VPS people love to flex when AWS goes down, but they always overlook the bigger picture
When AWS is down, everything is, so nobody is angry with you. The entire internet blames Amazon and you're off the hook
When your VPS is down, nothing else is, so all customer focus is on your dumb infrastructure and why haven't you fixed it yet? What the fuck is a "hetzner"? I want a refund.
Source: 15 years of running infrastructure that isn't on AWS.
KAY 🍃 retweeted
Excited to release new repo: nanochat!
(it's among the most unhinged I've written).
Unlike my earlier similar repo nanoGPT which only covered pretraining, nanochat is a minimal, from scratch, full-stack training/inference pipeline of a simple ChatGPT clone in a single, dependency-minimal codebase. You boot up a cloud GPU box, run a single script and in as little as 4 hours later you can talk to your own LLM in a ChatGPT-like web UI.
It weighs ~8,000 lines of imo quite clean code to:
- Train the tokenizer using a new Rust implementation
- Pretrain a Transformer LLM on FineWeb, evaluate CORE score across a number of metrics
- Midtrain on user-assistant conversations from SmolTalk, multiple choice questions, tool use.
- SFT, evaluate the chat model on world knowledge multiple choice (ARC-E/C, MMLU), math (GSM8K), code (HumanEval)
- RL the model optionally on GSM8K with "GRPO"
- Efficient inference the model in an Engine with KV cache, simple prefill/decode, tool use (Python interpreter in a lightweight sandbox), talk to it over CLI or ChatGPT-like WebUI.
- Write a single markdown report card, summarizing and gamifying the whole thing.
Even for as low as ~$100 in cost (~4 hours on an 8XH100 node), you can train a little ChatGPT clone that you can kind of talk to, and which can write stories/poems, answer simple questions. About ~12 hours surpasses GPT-2 CORE metric. As you further scale up towards ~$1000 (~41.6 hours of training), it quickly becomes a lot more coherent and can solve simple math/code problems and take multiple choice tests. E.g. a depth 30 model trained for 24 hours (this is about equal to FLOPs of GPT-3 Small 125M and 1/1000th of GPT-3) gets into 40s on MMLU and 70s on ARC-Easy, 20s on GSM8K, etc.
My goal is to get the full "strong baseline" stack into one cohesive, minimal, readable, hackable, maximally forkable repo. nanochat will be the capstone project of LLM101n (which is still being developed). I think it also has potential to grow into a research harness, or a benchmark, similar to nanoGPT before it. It is by no means finished, tuned or optimized (actually I think there's likely quite a bit of low-hanging fruit), but I think it's at a place where the overall skeleton is ok enough that it can go up on GitHub where all the parts of it can be improved.
Link to repo and a detailed walkthrough of the nanochat speedrun is in the reply.
