Zaoral @zaoral@fedibird.com

https://x.com/2022_technology/status/1819370945418862687

https://huggingface.co/grapevine-AI/gemma-2-27b-it-gguf

gemma-2-27b-it-gguf

https://aistudio.google.com/app/prompts/new_chat?model=gemini-1.5-pro-exp-0801
gemini-1.5-pro-exp-0801

https://x.com/shanegJP/status/1819145571040743639

https://ollama.com/library
モデル

https://x.com/7shi/status/1819013108264050843
ローカルLLM

https://x.com/umiyuki_ai/status/1818583975356743772
> Llama3.1-405B（4bit量子化

>Llama-3.1-8B-EZO-1.1-itのggu
https://huggingface.co/mmnga/Llama-3.1-8B-EZO-1.1-it-gguf
https://x.com/WMjjRpISUEt2QZZ/status/1818609469447979202

https://prtimes.jp/main/html/rd/p/000000004.000129878.html

https://ai-scholar.tech/articles/survey/AI-Knowledge-Graphs

https://findy-tools.io/articles/ml/17

日本語事前学習モデル『Sarashina』
https://www.sbintuitions.co.jp/blog/entry/2024/07/26/150005

https://x.com/sbintuitions/status/1817802246530400390

https://x.com/hAru_mAki_ch/status/1817413835948797962
> LLaMa 3.1ユーザー必見！405B→8B変換で劇的コストダウンを実現
これは良さそう！！

どのモデルをどれくらい量子化するといいかがローカルLLMの醍醐味かも

https://x.com/npaka123/status/1817183372130832419
> Llama-3.1-70B-Japanese-Instruct-2407-gguf(Q8)をお試し中。
M3 128GB

https://x.com/CyberAgent_PR/status/1816668618811445422

https://aws.amazon.com/jp/blogs/machine-learning/llm-experimentation-at-scale-using-amazon-sagemaker-pipelines-and-mlflow/

https://gigazine.net/news/20240725-mint-1t-open-source-multimodal-dataset/?utm_source=x&utm_medium=sns&utm_campaign=x_post&utm_content=20240725-mint-1t-open-source-multimodal-dataset

https://x.com/AiXsatoshi/status/1815975930705432619

https://huggingface.co/aixsatoshi/Meta-Llama-3.1-8B-Instruct-plus-Swallow

Meta-Llama-3.1-8B-Instruct-plus-Swallow

https://x.com/Holy_fox_LLM/status/1815781865275727976

https://gigazine.net/news/20240724-google-research-neuralgcm/

https://x.com/bioshok3/status/1815767644253077530
https://llama.meta.com/

https://x.com/MLBear2/status/1815766851005382912