An Early Black Friday
Bitcoinโs rally to $126k reversed amid macro stress and a $19B futures wipeout. ETF inflows slowing and volatility spiking, the market enters a reset phase marked by a historic leverage flush.
Read the full Week On-Chain below๐
https://t.co/Osm96VjuJg
Last week, we've had a massive collapse on the markets.
Security & custody is a very important topic for #Altcoins & #Bitcoin.
That's why, in this episode, we discuss how $11 billion in $BTC is secured with this system, with Joe Kelly of @unchained:
https://t.co/jMN19nKghO
Ranknetix is such an exciting projectโusing AI and blockchain to redefine SEO for Web3 is a game-changer. The $RNX token utility and performance incentives make it even more innovative!
๐ ๐๐๐๐ฉ ๐๐ข๐ฏ๐ ๐ข๐ง๐ญ๐จ ๐ญ๐ก๐ ๐ฉ๐จ๐ฐ๐๐ซ ๐จ๐ ๐จ๐๐-๐๐ก๐๐ข๐ง ๐๐จ๐ฆ๐ฉ๐ฎ๐ญ๐ข๐ง๐ ๐ฐ๐ข๐ญ๐ก ๐๐๐ง๐ค๐ง๐๐ญ๐ข๐ฑ! ๐
Ranknetix seamlessly integrates off-chain AI processing with on-chain validation, making SEO optimization smarter and more efficient. Here's how:
โจ ๐๐-๐๐ซ๐ข๐ฏ๐๐ง ๐๐๐ซ๐ค๐๐ญ ๐๐ง๐ฌ๐ข๐ ๐ก๐ญ๐ฌ
Using advanced NLP and ML models, Ranknetix analyzes real-time market trends, token reputations, and search behaviors off-chain. The results? Actionable insights that are validated on-chain for transparency.
๐ก ๐๐๐จ๐ง๐จ๐ฆ๐ข๐ ๐๐จ๐ฅ๐ ๐จ๐ $๐๐๐
$RNX isn't just a tokenโit fuels the ecosystem. Users earn rewards for contributing off-chain SEO data or improving rankings through Stake-to-Earn models. All validated and tracked on-chain!
๐ฏ ๐๐ ๐-๐๐จ๐ฐ๐๐ซ๐๐ ๐๐๐ฉ๐ฎ๐ญ๐๐ญ๐ข๐จ๐ง
Ranknetix ties dynamic SEO performance scores to NFTs using off-chain calculations. These scores evolve in real-time and represent a user's SEO prowess, with direct utility in governance and ad prioritization.
๐ ๐๐ซ๐๐๐ข๐๐ญ๐ข๐ฏ๐ ๐๐๐
Off-chain AI crunches massive data sets to predict keyword trends and optimize ad performance, ensuring users maximize ROI while minimizing costs.
Ranknetix proves that off-chain computing is the backbone of scalable, intelligent Web3 applications. With $RNX driving incentives, it bridges the gap between efficiency and transparency.
๐ #BlockchainSEO #RNX
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.
The signals lined up before โ theyโre lining up again.
Hereโs what we forecasted, and what happened next.
Get โWhat we said and how it played outโ, Vol. II here. ๐ https://t.co/ZYFQ3xCla4
This is a good decision and I respect it.
Not every business should be trying to appeal to many customers as possible in the name of "not being maximalist". We need the stubborn ones who believe in their cause and their tribe and see their work as a labor of love to it.
Ever wonder what the hell happened when @sama was removed from @OpenAI . Or how ChatGPT rose to the prominent role itโs currently in?
@sebbunney and I discuss that and much more in this weekโs episode.