AndreasFischer1985

@[email protected]
Data Scientist & Bildungsforscher. Privat aktiv für die Bundesstelle für Open Data - und privat hier 👨🏻‍💻

[AndreasFischer1985 @AFischer1985]

So, hier ist sie nun - die Dokumentation zur API des umfangreichsten Berufe-Lexikons Deutschlands:
die BERUFENET-API 🎉

https://t.co/rsKWo1hZX1

[VAGO solutions @VAGOsolutions]

📢 What a week for open-source AI!

@AIatMeta Llama-3.1-8b-instruct impressed with its German skills.

Today, we're launching Llama-3.1-SauerkrautLM-8b-Instruct!

Built on our Sauerkraut Dataset V2

🔗 Details: https://t.co/sq3m1tgKES

[AI at Meta @AIatMeta]

Starting today, open source is leading the way. Introducing Llama 3.1: Our most capable models yet.

Today we’re releasing a collection of new Llama 3.1 models including our long awaited 405B. These models deliver improved reasoning capabilities, a larger 128K token context window and improved support for 8 languages among other improvements. Llama 3.1 405B rivals leading closed source models on state-of-the-art capabilities across a range of tasks in general knowledge, steerability, math, tool use and multilingual translation.

The models are available to download now directly from Meta or @huggingface. With today’s release the ecosystem is also ready to go with 25+ partners rolling out our latest models — including @awscloud, @nvidia, @databricks, @groqinc, @dell, @azure and @googlecloud ready on day one.

More details in the full announcement ➡️ https://t.co/hhJoLm5eLV
Download Llama 3.1 models ➡️ https://t.co/rRjvmxqCTC

With these releases we’re setting the stage for unprecedented new opportunities and we can’t wait to see the innovation our newest models will unlock across all levels of the AI community.

[André Bittermann @AndreBittermann]

✨ Now available ✨
Together with @AFischer1985, I guest-edited the Special Issue "#NLProc in #Psychology" in Zeitschrift für Psychologie.
Seven original articles demonstrate how text can be used as data in psychological research.
https://t.co/Pi15o5VMJj

[AndreasFischer1985 @AFischer1985]

❤️🤗

[clem 🤗 @ClementDelangue]

This isn’t a goal of ours because we have plenty of money in the bank but quite excited to see that @huggingface is profitable these days, with 220 team members and most of our platform being free (like model hosting) and open-source for the community!

Especially noteworthy at a time when most AI startups wouldn’t survive a year or two without VC money. Great job team!

[Arena.ai @arena]

Chatbot Arena Update!

1. Multilingual Arena -- four new languages (German, Spanish, Russian, Japanese).

GPT-4o is #1 in English, German, and Spanish. Gemini-1.5-Pro is #1 in Japanese, Chinese, and French. Claude-3 Opus is #1 in Russian. The competition is tight, and we need more votes 🗳️ to confidently rank them.

Let's challenge LLMs in any language!

2. Yi-1.5-34B-Chat shows impressive performance, matching larger models like Qwen-1.5-110B and GPT-4-0613. Congrats @01AI_Yi on this milestone!

3. Phi-3 Medium and Small are finally on the board! Medium (14B) ranks near GPT-3.5-Turbo-0613, Small (7B) ranks ~Llama-2-70B. We also see robust performance in Hard Prompts.

Congrats @Microsoft Phi team on these great models for the community!

Learn more
- Full leaderboard https://t.co/PBF1eCxRFy
- Chat & vote at https://t.co/IDFeIDIOtm

[AndreasFischer1985 @AFischer1985]

Wir haben im Laufe der letzten Jahre im Projekt #KIPerWeb eine lebhafte, kompetente und offene Austauschrunde zur Nutzung und Entwicklung von KI-gestützen Webanwendungen etabliert, die wir nun ehrenamtlich weiterführen werden. Interesse mitzumachen? 👉 https://t.co/iMPjb5P0jG

[AndreasFischer1985 @AFischer1985]

Die Folien unserer #KIPerWeb-Tagung vom 19.04.24 zu #KI-gestützter Personalisierung in der berufsbezogenen #Weiterbildung
finden sich neben weiterführenden Materialien zum Thema nun auch online:
https://t.co/ntW9ekaDNb

[Philipp Schmid @_philschmid]

Llama 3 released! 🚨🔔@AIatMeta just released their best open LLM! 👑🚀 Llama 3 is the next iteration of Llama with a ~10% relative improvement to its predecessor! 🤯 Llama 3 comes in 2 different sizes 8B and 70B with a new extended tokenizer and commercially permissive license! ✅

Blog: https://t.co/VrceFpjI1o
Models: https://t.co/SdkOeWURuM

New and improvements to v2✨:
🔠 Trained on 15T Tokens & fine-tuned on 10M human annotated samples
🧮 8B & 70B versions as Instruct and Base
🚀 Llama 3 70B best open LLM on MMLU (> 80 🤯)
🧑🏻‍💻 Instruct good at coding 8B with 62.2 and 70B 81.7 on Human Eval
✍🏻 Tiktoken-based tokenizer with a 128k vocabulary
🪟 8192 default context window (can be increased)
🧠 Used SFT, PPO & DPO for alignment.
💰Commercial use allowed ✅
🤗 Available on @Hugging Face
🤝 1-click deployments on Hugging Face, Amazon SageMaker, Google Cloud
🔜 more model sizes & enhanced performance

Massive kudos to Meta for continuing its commitment to open AI. Honored to partner with Joe and team! 🤗 The gap is melting. 🧊

[Devendra Chaplot @dchaplot]

We just released Mixtral-8x22B-v0.1 and Mixtral-8x22B-Instruct-v0.1:
- Free to use under Apache 2.0 license
- Outperforms all open models
- Native function calling
- Masters English, French, Italian, German and Spanish.
- Seq_len = 64K

https://t.co/SCG8s06Dbl

[WizardLM @WizardLM_AI]

🧙‍♀️ WizardLM-2 8x22B is our most advanced model, and just slightly falling behind GPT-4-1106-preview.
🧙 WizardLM-2 70B reaches top-tier capabilities in the same size.
🧙‍♀️ WizardLM-2 7B even achieves comparable performance with existing 10x larger opensource leading models.

The model weights of WizardLM-2 8x22B and WizardLM-2 7B are shared on Huggingface, and WizardLM-2 70B and the demo of all the models will be available in the coming days.

https://t.co/FWJs94FmB6

[Philipp Schmid @_philschmid]

We can do it! 🙌 First open LLM outperforms @OpenAI GPT-4 (March) on MT-Bench. WizardLM 2 is a fine-tuned and preferences-trained Mixtral 8x22B! 🤯

TL;DR;
🧮 Mixtral 8x22B based (141B-A40 MoE)
🔓 Apache 2.0 license
🤖 First > 9.00 on MT-Bench with an open LLM
🧬 Used multi-step synthetic data pipeline including Evol-instruct
🔄 data partitions and stage-by-stage training
👨‍🔬 Used SFT → DPO → PPO

Blog: https://t.co/gUuXlZPmyv
Model: https://t.co/s3bBxir0uR
Paper: coming soon

[clem 🤗 @ClementDelangue]

The new @MistralAI is now #1 on the openLLM leaderboard. Apache 2.0 license too! 🔥🔥🔥

[André Bittermann @AndreBittermann]

"Potenziale von #KI für die Personalisierung von (Weiter-)Bildung" - der Vortrag von @AFischer1985 ist nun im @ZPID Mediacenter frei verfügbar!

👉 https://t.co/KOQmWGY4ix

#TextAsData #Embeddings #LLM #NLP #Education

[André Bittermann @AndreBittermann]

Letzte Woche hatten wir @AFischer1985 vom @fbb_de zu Gast im @ZPID Kolloq. Seine Ideen mit #SentenceEmbeddings haben uns direkt zur Lösung eines Textklassifikationsproblems inspiriert😌Die großartigen Folien seines #KI-Vortrags sind nun auf PsychArchives!👉https://t.co/GNYJaneSq8

[Leibniz-Institut für Psychologie (ZPID) @ZPID]

📢 #Reminder
Potenziale von KI für die Personalisierung von (Weiter-)Bildung – von Content-Creation bis zum automatisierten Entscheidungsmanagement". Kolloquium am Fr, 15.3.2024 mit Dr.

@AFischer1985
@fbb_de
#bigdata #ki #bildung
Alle Infos:
https://t.co/eC0zd3pc1y

[Edward Snowden @Snowden]

OpenAI confessing 𝐨𝐧 𝐭𝐡𝐞𝐢𝐫 𝐨𝐰𝐧 𝐛𝐥𝐨𝐠 to a belief that "as we get closer to building AI, it will make sense to start being less open... but it's totally OK to not share the science..." is about as bad of a heel-turn as it gets.

[AndreasFischer1985 @AFischer1985]

Bei solchen Aushängen freut man sich doch gleich nochmal soviel auf den Vortrag am @ZPID 😘

https://t.co/IlEad0cuK3

[Leibniz-Institut für Psychologie (ZPID) @ZPID]

"Potenziale von KI für die Personalisierung von (Weiter-)Bildung – von Content-Creation bis zum automatisierten Entscheidungsmanagement".

Kolloquium am Fr, 15.3.2024 mit Dr. @AFischer1985
@fbb_de #bigdata #ki #bildung

Alle Infos: https://t.co/eC0zd3pc1y

[Alex Yanko 🇺🇦 @LeopolisDream]

Introducing Phind-70B – closing the code quality gap with GPT-4 Turbo while running 4x faster

https://t.co/sGjNGkPY1R

It will be released #opensource

[AndreasFischer1985 @AFischer1985]

Heute ist ein Beitrag von mir und Jens Dörpinghaus vom @BIBB_de in der Fachzeitschrift Knowledge erschienen! 🥳🎉🥂

Titel; „Web Mining of Online Resources for German Labor Market Research and Education: Finding the Ground Truth?“ 😎