Zaoral @zaoral@fedibird.com

新マルチアスペクト埋め込みモデル
https://note.com/npaka/n/n531d327a6c66?sub_rt=share_h
>「マルチアスペクトデータ」は、複数の概念と関係を含むデータ..

https://twitter.com/alfredplpl/status/1780042739318767892

https://huggingface.co/microsoft/WizardLM-2-8x22B
なぜかない

wizardLM

https://www.sbbit.jp/article/cont1/137771

https://twitter.com/yoko_materialDX/status/1779827015446610328
> LLMの化学知識

https://huggingface.co/TFMC/Japanese-Starling-ChatV-7B-GGUF

https://twitter.com/mutaguchi/status/1779810948796272641

https://twitter.com/OpenAI/status/1779754674242892014

https://japan.cnet.com/article/35217781/

https://twitter.com/jaguring1/status/1779758770324988230
>日本語特化GPT-4

https://x.ai/blog/grok-1.5v

https://twitter.com/jaguring1/status/1778974703010648065

https://twitter.com/jaguring1/status/1778771756586840204
> Drop 86%
> GPT-4 turbo-2024-04-09

https://twitter.com/kajikent/status/1778587614754582635
gpt4 turbo

https://ascii.jp/elem/000/004/192/4192907/

"Leave No Context Behind: Efficient Infinite Context Transformers with Infini-attention"

https://twitter.com/s_tat1204/status/1778297362131374363

> 既存の Transformer ベースのモデルに "メモリ" 機構を導入する Infini-attention の提案。

https://twitter.com/kyo_takano/status/1777695664035803534

https://twitter.com/ZeyuanAllenZhu/status/1777513016592040248

https://arxiv.org/pdf/2404.05405.pdf

Physics of Language Models: Part 3.3, Knowledge Capacity Scaling Laws

https://twitter.com/npaka123/status/1777810734447276249
>Gemini 1.5 Proの新機能

Gemma
https://twitter.com/_philschmid/status/1777673558874829090

>AI ソフトウェアエンジニアに関するプロダクト紹介
https://tech.algomatic.jp/entry/2024/04/04/180922

https://note.com/masayuki_abe/n/nfa4a2367de86?sub_rt=share_pb

https://huggingface.co/datasets/LanguageBind/Open-Sora-Plan-v1.0.0
Open-Soraデータ