Werner Bogula (@Bogula) header
Werner Bogula (@Bogula) profile picture

True digital native and analogue lute player / currently AI enabler @ https://teapawt.net/t.co/HmWbJswrPd

Hamburg
Joined January 2009
Photo by Werner Bogula (@Bogula)Photo by Werner Bogula (@Bogula)Photo by Werner Bogula (@Bogula)Photo by Werner Bogula (@Bogula)Photo by Werner Bogula (@Bogula)Photo by Werner Bogula (@Bogula)Photo by Werner Bogula (@Bogula)Photo by Werner Bogula (@Bogula)Photo by Werner Bogula (@Bogula)Photo by Werner Bogula (@Bogula)
Mission to Mars,an alternative inauguration hymn. The future if the world according to @elonmusk
30
Todays news left me speechless, only a good sond can save us now: udio.com/songs/rvHJBJEPiSMux…
1
1
22
Werner Bogula retweeted
This is insane! 😱 You can now train a 100-billion parameters LLM on Google Colab. Explanation and code below ⤵️⤵️⤵️ --- TL;DR --- Distributed training over the Internet has become operational with the release of the new version of the PETALS distributed training package. Join forces with people from around the world and receive immense computing power in return for your use! 🌐⚡️ --- Intro --- Distributed training over the internet has turned from a dream into a reality with the release of the new version of the Petals distributed training package. Upon connection, the package grants you access to immense computing power that allows you to do anything you can imagine. It doesn't even have to be your own computing power: Attaching below is a Colab notebook that trains a 65-billion-parameter model (and also supports 100 billion. Yes, models with a hundred billion parameters are accessible to you). 💻🚀 --- It surely takes years, doesn't it? --- No. For example, with a 65-billion-parameter inference, you'll get a speed of 5 tokens per second. ⏱️⚡️ --- How does it work? --- The concept is very simple. In LoRA-based training (My tweet about LoRA - teapawt.net/ItakGol/status/1…), the base model is the same for everyone and does not change until the end of training, so many people need it at the same time. With Petals, people load small parts of the model, and each executes their part in the computation of running the full model. All the headache and complex software engineering required to make this madness run smoothly and quickly happens inside the package. 🤯🔧 --- How complex? --- From your own perspective, simply run: AutoDistributedModel.from_pretrained --- Important! Privacy! --- Obviously, this infrastructure sends your data outside to the world, so it's crucial to take that into account. Your code is never sent outside. 🌐🔒 --- Links --- 1. Article: Petals: Collaborative Inference and Fine-tuning of Large Models - aclanthology.org/2023.acl-de… 2. See the speed for yourself: Distributed Chatbot - chat.petals.ml/ 3. Notebook for training a 65-billion-parameter model: colab.research.google.com/dr… 4. Code: github.com/bigscience-worksh… 5. Paper: arxiv.org/abs/2209.01188 6. Join the community: discord.com/invite/D9MwApKgW…
How to Fine-Tune a Large Language Model (LLM) on a Single GPU or Google Colab? 😊 LoRA (Low-Rank Adaptation)! 🚀 Recently, LoRA has been a hot topic in the context of fine-tuning LLMs. But what does LoRA really mean? Let's dive in 📚 1 / 18
22
314
11
1,427
372,432
Bookmarklist of a computer linguist 2022 vs 2023 #NLP #ArtificialIntelligence #ChatGPT
2
78
#AI in #Hamburg für alle. Warum nicht den Sonntagsspaziergang bei einem kostenlosen Kaffee mit Kuchen im Dockland abrunden? Am Sonntag den 19.6 im @aric_hamburg Showroom für Künstliche Intelligenz #ai #sonntag #freecoffee
Ihr habt Lust, das ARIC kennenzulernen und ins Thema KI einzusteigen? Dann kommt zum Open Day am 19.6.! Es gibt: 🤖Showroom-Führungen 🤖KI-Einsteiger-Workshop für junge Menschen ab 16 J. 🤖Kuchen Hier informieren und kostenlos auf Eventbrite anmelden: aric-hamburg.de/workshops/ka…
1
1
3
Zwei Jahre nach Stapellauf in Hamburg festgemacht. Die Pionier One von Garbor Steingarts Crew lud heute zur Hafenrundfahrt.
Using @Aleph__Alpha's multimodal language model MAGMA to find creative titles for thousands of pieces of #cryptoart generated by @peezmo_ . TLDR; it works!
Kann man abstrakte Kunstwerke von @peezmo_ mit KI sinnvoll betiteln? @Bogula hat es ausprobiert und dabei das Magma-Tool unseres neuen Mitglieds @Aleph__Alpha benutzt. Fazit: Es funktioniert!👍🤖 Wie er dabei vorgegangen ist, erklärt er in seinem Paper: tinyurl.com/yvse757t
3
Werner Bogula retweeted
Das ARIC im Max Magazin: Wenn eine Künstliche Intelligenz eine Weihnachtsgeschichte verfasst, endet das auch schon mal mit einem Axtmörder. @Bogula erklärt im Interview mit @OsingTim, wie es dazu kommt und, wie #NLP funktioniert. Jetzt am Kiosk! #artificialintelligence #AI
1
4
Danke für die Referenz mit dem "Mist". Es steckt außerdem auch eine Menge Arbeit von Grafikern, DidaktikerInnen und Fachleuten drin, für die ich mich noch einmal bedanken will.
This Post is from an account that no longer exists. Learn more
1
Werner Bogula retweeted
Here’s your chance to win a free NFT, @PEEZMO_Art, because we are on 🔥 and we marked 200 followers. - follow, like & RT - tag 2 friends in your comment - leave or dm your ETH -address Winner announced on Sat, Aug 21th at 8PM CEST #NFTGiveaway 🤟 opensea.io/assets/matic/0x92…
10
12
1
20
As someone who knew Frank Schirrmacher - and his views on Silicon Valley - rather well, I'm quite appalled that a prize in his honor has been given to Peter Thiel. What cosmic injustice is that?
3
16
3
137
Celebrating the successful launch of the #AI art collection of @PEEZMO_Art with creator DarioGi and #crypto gallerists @magall_art. #AI will help to generate >4000 pieces of art.
Introducing "Dario Gi" Inventor and Programmer of PEEZMO AI Engine with @magall_art, posing with a physical @PEEZMO_Art at fontana di Rathaus Hamburg! #AI #AIArt #nft #opensea @opensea #polygonmatic #peezmo #nftcollector #NFTcollection
1
2
A new carnivorous plant in my garden. Bought my first piece of Crypto Art from @magall_art. It's generated by a sophisticated AI process designed by @PEEZMO_ Checkout the whole collection at opensea.io/assets/peezmo-ai-…
1
2
6
Share of electricity that comes from fossil fuels: 🇿🇦 South Africa: 89% 🇦🇺 Australia: 75% 🇮🇳 India: 74% 🇯🇵 Japan: 69% 🇨🇳 China: 66% 🌍 World: 61% 🇺🇸 USA: 60% 🇩🇪 Germany: 44% 🇬🇧 UK: 41% 🇧🇷 Brazil: 14% 🇫🇷 France: 10% 🇸🇪 Sweden: 2% Explore any country here: ourworldindata.org/grapher/p…
98
829
222
2,252
Das (kulturelle) Leben kehrt zurück. Gestern Abend Staatsoper Hamburg Händels Agrippina. ndr.de/kultur/sendungen/oper…
Werner Bogula retweeted
"Now the script has flipped and a new solar plant is almost three times cheaper than a new coal one. The price of electricity from solar declined by 89% between 2009 and 2019." ourworldindata.org/battery-p…
45
309
46
1,596
Werner Bogula retweeted
MAGALL - Club - Berlin @cryptovoxels Opening Party with @EclecticMethod Supported by @nftsmart Join here ➡️https://teapawt.net/t.co/9SRxgCggMO⬅️ Sun, May 9 2021 BERLIN, 10:00p / NEW YORK 4:00p / LOS ANGELES, 1:00p NFTs Backstage Area: @SkullsOnETH @mamadousow221 @PEEZMO_Art
5
15
1
26
Heute um 12.00 gebe ich in der beliebten @ARIC Brownbag Session einen schnellen Überblick über die KI Trends 2020/21 mit den 5 mir wichtigen Trends: 1 NLP, 2 Meta Science, 3 Hardware, 4 Ethical AI und 5 AI Powers #ai #ArtificialIntelligence meetup.com/de-DE/ARIC-Brown-…
Morgen (18.03.) die KI Trends 20/21 mit Werner Bogula
2