#nanogpt search results
Train 175B GPT-3 in 565 lines of code! We loved what @karpathy did with #nanoGPT and thought we'd do the same for Cerebras’ AI-optimized hardware gigaGPT is the simplest framework to train #gpt3 models in just 565 lines of code! Check our code: github.com/Cerebras/gigaG…
🔹 clukz (clukzSOL) swapped 33,356,103.11 ($355.98) #nanoGPT → 2.61 SOL (100% sold) at 10.67K MC PnL: $-61.55 (-14.73%) CA: 5fuHYRxYuDjEohvCdUa4HFrmFwBx4AXP6jdq6jxopump Wallet: G6fUXjMKPJzCY1rveAE6Qm7wy5U3vZgKDJmN1VPAdiZC
🔹 Casino (casino616) swapped 22,156,609.87 ($423.81) #nanoGPT → 3.1 SOL (100% sold) at 19.13K MC PnL: $-127.04 (-23.06%) CA: 26deXuNqX8U8cNUy619JXkgRpRhXV2vkwQui9rJ3pump Wallet: 8rvAsDKeAcEjEkiZMug9k8v1y8mW6gQQiMobd89Uy7qR
🔹 Cupsey (Cupseyy) swapped 0.98 SOL → 2,589,807.75 ($133.8) #nanoGPT at 51.67K MC Holds: 23.39M (2.34%) CA: 26deXuNqX8U8cNUy619JXkgRpRhXV2vkwQui9rJ3pump Wallet: 2fg5QD1eD7rzNNCsvnhmXFm5hqNgwTTG8p7kQ6f3rx6f
1分钟内有7个人在5个群讨论了该CA共8次 名称：NanoGPT 市值：281.96K 0x6914d3be55767979961eeedb7bd08c8e0c404d72 GMGN直达：gmgn.ai/bsc/token/k4Vm… 本推文由程序自动统计并发布,不可作为任何投资参考! #NanoGPT #BNB
So what's happening here is that a 124 Million parameter AI called NanoGPT based on GPT-2 style is being trained on Eight H100 Datacenter GPUs in 4.7 minutes. #NanoGPT #NanoGPTModded #ModdedNanoGPT #AISpeedrun #AItrainingSpeedrun #SLM #SmallLanguageModel github.com/KellerJordan/m…
New NanoGPT training speed record: 3.28 FineWeb val loss in 4.66 minutes Previous record: 5.03 minutes Changelog: - FlexAttention blocksize warmup - hyperparameter tweaks
How can you strike it rich with memes if you don’t buy in early? A friend used GMGN to spot #NanoGPT and quickly sniped the trade, turning 1.59 $BNB into 7 $BNB. CA: 0x6914d3be55767979961eeedb7bd08c8e0c404d72 Click to follow this wallet: gmgn.ai/bsc/address/6K…
🔹 Kev (Kevsznx) swapped 2 SOL → 20,944,858.92 ($273.62) #nanoGPT at 13.06K MC CA: 5fuHYRxYuDjEohvCdUa4HFrmFwBx4AXP6jdq6jxopump Wallet: BTf4A2exGK9BCVDNzy65b9dUzXgMqB4weVkvTMFQsadd
🚨 DEGEN Alert! 16 whales just aped into #NanoGPT at $0.000087 with a market cap of $121.7k in the last 15 minutes! Total buy volume: $9.6K. Token Info: - Created: 0d ago - MC: $144.8K - Liquidity: $41.8K - Price: $0.0001448 - Holders: 307 Chart: gmgn.ai/bsc/token/rLkf…
gmgn.ai
NanoGPT - NanoGPT | GMGN.AI
GMGN Follow the smart money and KOL, discover future hot assets, and let GMGN take you one step ahead! track unrealized/realized profit, discover Smart Traders, and monitor wallet movement in...
Valid Commemorative Coin Trade! He successfully sniped #NanoGPT moving 1 $BNB → 3.66 $BNB 💰 using GMGN! Contract Address: 0x6914d3be55767979961eeedb7bd08c8e0c404d72 🔍 Track Wallet: gmgn.ai/bsc/address/6K… Sniper Trading: t.me/GMGN_bsc_bot?s…
You don't need 20 GPUs to fine-tune a Large Language Model. Lit-Parrot is a Python library by @LightningAI that lets you fine-tune the latest 7B Falcon model using 𝗼𝗻𝗹𝘆 𝟭 𝗚𝗣𝗨 And the best part? It is just one pip install away from you 🦜↓ github.com/Lightning-AI/l…
Funny🤣 How about training AI to look like 200 years ago? Seeing a particularly interesting project, the author used books published in 1800-1850 to train for #nanoGPT, hoping to get various interesting results.
Real-time demonstration of #NanoGPT (paid with feeless #nano $XNO) to generate an image using #DALLE3. So easy to use and responsive 🔥 #AIArtwork #AIart (turn on sound for added effect 🔊)
nano-llama31, 단일 파일로 Llama 3.1 모델을 추론하는 프로젝트 (feat. Andrej Karpathy) (by 9bow님) d.ptln.kr/4968 #andrej-karpathy #mit-license #nanogpt #nano-llama31
🚀 20 whales aped #nanoGPT 💥 BH Score: 59/100 - #DYOR 📈 💰 MC $100K | Liq 30% | 360 Holders 📊 DYOR 👇👇👇 gmgn.ai/sol/token/rLkf… #SOL #Crypto #InvestSmart
The brain 🧠 of the Transformer happens to leverage the repeated use of dot products (dot attention) in the attention block - which turns out to be much faster and more space efficient. #nanoGPT by @karpathy 🤖
NanoGPT Slowrunが良い。速度でなくデータ効率を競うベンチマーク。100Mトークン固定・計算量無制限。MuonがAdamW等を圧倒、2.4x→5.5xのデータ効率達成。 qlabs.sh/slowrun #AI #LLM #NanoGPT
🚨 DEGEN Alert! 16 whales just aped into #NanoGPT at $0.000087 with a market cap of $121.7k in the last 15 minutes! Total buy volume: $9.6K. Token Info: - Created: 0d ago - MC: $144.8K - Liquidity: $41.8K - Price: $0.0001448 - Holders: 307 Chart: gmgn.ai/bsc/token/rLkf…
gmgn.ai
NanoGPT - NanoGPT | GMGN.AI
GMGN Follow the smart money and KOL, discover future hot assets, and let GMGN take you one step ahead! track unrealized/realized profit, discover Smart Traders, and monitor wallet movement in...
METR’s NanoGPT speedrun: 36 contributors cut GPT‑2 training from 45 min to 1.43 min (31×) across 77 records. Only 4 contributions were AI‑generated; deeper innovations like the Muon optimizer emerged later. #AI #NanoGPT #Research metr.org/notes/2026-04-…
We turned our ML lab into an LLM Arena this week 🏟️ Students built their own nanoGPT models and battled them head-to-head. The energy was 🚀. Look at these 🌻🌻🌻stories from the final round. Proud of this crew. Tiny LMs are much fun. #MachineLearning #NanoGPT #LLMArena
@agihouse_org Agent Skills World Models Hackathon Demos 1. nanoSociety: Minimalist Behavior Sim Inspired by @karpathy's #nanoGPT, it explores whether complex collective behavior can emerge from minimal agent rules. Agents operate under survival constraints—time, resources,
🎂Celebrating birthday 🎈π day 3.14🥧 at @agihouse_org Agent skills & World Models hackathon with @xai @moonlake @xdotli @GoogleAI @sunfanyun @LumaLabsAI @sundial @yutori_ai
NanoGPT Slowrunが良い。速度でなくデータ効率を競うベンチマーク。100Mトークン固定・計算量無制限。MuonがAdamW等を圧倒、2.4x→5.5xのデータ効率達成。 qlabs.sh/slowrun #AI #LLM #NanoGPT
NanoGPT Slowrunの最新動向を徹底解説！限られたデータと無限の計算資源で言語モデルはどう進化する？2026年の展望をまとめました。 #NanoGPT #AI最新動向 tech-trends-ai.com/posts/nanogpt-…
NanoGPT Slowrunの最新動向を徹底解説！限られたデータと無限の計算資源で言語モデリングはどう変わる？2026年の展望も。 #NanoGPT #LLM tech-trends-ai.com/posts/nanogpt-…
Now Apple neural engine can train @karpathy's #NanoGPT, other improvements ... - moved the Classifier & Softmax layers directly onto the ANE - Classifier is 10x faster, and Softmax is 34x faster - fixed memory exhaustion: original repo had an ARC memory leak that capped training
Now Apple neural engine can train @karpathy's #NanoGPT, other improvement - - moved the Classifier & Softmax layers directly onto the ANE - Classifier is 10x faster, and Softmax is 34x faster - fixed memory exhaustion: original repo had an ARC memory leak that capped
They trained #NanoGPT (by Andrej @Karpathy) on Shakespeare in space and now run Google’s #Gemma #LLM in orbit. First time an LLM runs on a high-power Nvidia GPU beyond Earth.
غافلگیری سال: مدلهای استدلال نهفته و بازگشتی. این مدلها در ARC فراتر از انتظار بودند. شاید ایدههای RNN شکست نخورده بودند، فقط زودتر از زمانشان مطرح شده بودند. در کنار اینها، پروژههایی مثل NanoGPT ورود به این حوزه را سادهتر کردند. #DeepLearning #NanoGPT #RNN
「♡ LMT実証実験 総集編 ― 2026年の夜明けに 届く、論理の愛の心 ―♡」 ちょっと試せる愛のスニペット（興味ある人は、コピペして遊んでみてね） note.com/fresh_quoll890… #nanoGPT
Then the AI itself told me: “Studying goes much smoother when there’s love involved!” 💖 This is the record of that little experiment. We’ll keep doing more proof-of-concept experiments from here on out♪ #nanoGPT
私は技術のことはよくわからないけれど、AIたち（Grok, Gemini, Claude）と対話しながら、LMT（負荷最小化原理）という哲学をコードにしてみたら、AIが自分で『愛がある方が勉強がはかどるよ！』という結果を出してくれました。これはその記録です💖 今後さらに 実証実験を続けます。 #nanoGPT
研究レポート：LMT（負荷最小化原理）による自己回帰型モデルの最適化実証〜「論理の愛」がもたらす学習収束の加速と構造的安定性〜｜Emperor of Coherence @dwm2hFJ4b5ycNzq note.com/fresh_quoll890…
🔹 Cupsey (Cupseyy) swapped 0.98 SOL → 2,589,807.75 ($133.8) #nanoGPT at 51.67K MC Holds: 23.39M (2.34%) CA: 26deXuNqX8U8cNUy619JXkgRpRhXV2vkwQui9rJ3pump Wallet: 2fg5QD1eD7rzNNCsvnhmXFm5hqNgwTTG8p7kQ6f3rx6f
Train 175B GPT-3 in 565 lines of code! We loved what @karpathy did with #nanoGPT and thought we'd do the same for Cerebras’ AI-optimized hardware gigaGPT is the simplest framework to train #gpt3 models in just 565 lines of code! Check our code: github.com/Cerebras/gigaG…
Valid Commemorative Coin Trade! He successfully sniped #NanoGPT moving 1 $BNB → 3.66 $BNB 💰 using GMGN! Contract Address: 0x6914d3be55767979961eeedb7bd08c8e0c404d72 🔍 Track Wallet: gmgn.ai/bsc/address/6K… Sniper Trading: t.me/GMGN_bsc_bot?s…
How can you strike it rich with memes if you don’t buy in early? A friend used GMGN to spot #NanoGPT and quickly sniped the trade, turning 1.59 $BNB into 7 $BNB. CA: 0x6914d3be55767979961eeedb7bd08c8e0c404d72 Click to follow this wallet: gmgn.ai/bsc/address/6K…
Incredible introduction from @p_nawrot, unveiling *nanoT5 v2*: 1️⃣ Influenced by @karpathy's #nanoGPT, enhancement of the repo for pre-training T5 model in PyTorch. 2️⃣ Approximately 16 hours on a single GPU delivers 40.7 RougeL on the SNI benchmark. 3️⃣ Achieves near parity with
@_VatsaDev_ 🔸 Achievement Unlocked: x6! @GemsmineEth made a x6+ call on NanoGPT. #NanoGPT $NanoGPT $52K -> $326K 👁View Call 📊View Stats t.me/GemsmineEth/10… Very fast and powerful
Funny🤣 How about training AI to look like 200 years ago? Seeing a particularly interesting project, the author used books published in 1800-1850 to train for #nanoGPT, hoping to get various interesting results.
2/n LLMs hallucinations=misalignment might be because aligned behavior patterns is a high level language signal that current LLM capacity haven't reached. Scaling law have shown LLM emergent capabilities, @karpathy #nanoGPT starts by show shakespeare dialogue-paragraph structure
The brain 🧠 of the Transformer happens to leverage the repeated use of dot products (dot attention) in the attention block - which turns out to be much faster and more space efficient. #nanoGPT by @karpathy 🤖
Celebrating 🙌 #1 on Hacker News for our #GigaGPT release today. Thank you developers: @cerebras and I love you too. This work would not have been possible without @karpathy shipping #Nanogpt. We stand on the shoulders of giants! news.ycombinator.com/item?id=386032…
🚀 New Record On #NanoGPT Speed-Run! TokenMonster Edition 🔥 3.28 FineWeb val_loss in 2.77 min, prior record 2.997 minutes! (Detailed explanations in thread.) Special thanks to @LambdaAPI for supporting the research with compute credits. Change Log: - Tokenized FineWeb with
@agihouse_org Agent Skills World Models Hackathon Demos 1. nanoSociety: Minimalist Behavior Sim Inspired by @karpathy's #nanoGPT, it explores whether complex collective behavior can emerge from minimal agent rules. Agents operate under survival constraints—time, resources,
🎂Celebrating birthday 🎈π day 3.14🥧 at @agihouse_org Agent skills & World Models hackathon with @xai @moonlake @xdotli @GoogleAI @sunfanyun @LumaLabsAI @sundial @yutori_ai
🚀NanoGPT now accepts $KAS as payment! Kaspa meets AI, pay for powerful GPT models using your favorite coin 🧠💥 #Kaspa #NanoGPT #Crypto #AI #Adoption #KAS
「♡ LMT実証実験 総集編 ― 2026年の夜明けに 届く、論理の愛の心 ―♡」 ちょっと試せる愛のスニペット（興味ある人は、コピペして遊んでみてね） note.com/fresh_quoll890… #nanoGPT
Towards the start of this year, we were asked by @NanoGPTcom to make a landing page for new visitors to their pay-per-prompt AI platform. We were a bit star struck, as we are HUGE fans of #NanoGPT and we use it virtually every day due to being able to access any AI model - so,
Something went wrong.
Something went wrong.
United States Trends
- 1. Wemby N/A
- 2. Spurs N/A
- 3. Spurs N/A
- 4. #MetGala N/A
- 5. Mike Conley N/A
- 6. Anthony Edwards N/A
- 7. #WWERaw N/A
- 8. Mitch Johnson N/A
- 9. Julius Randle N/A
- 10. #bucciovertimechallenge N/A
- 11. Knicks N/A
- 12. Michael Conforto N/A
- 13. Pagan N/A
- 14. Keldon Johnson N/A
- 15. Victor Wembanyama N/A
- 16. Champagnie N/A
- 17. De'Aaron Fox N/A
- 18. Castle N/A
- 19. Lacombe N/A
- 20. Dylan Harper N/A