Introducing Meta Llama 3: The most capable openly available LLM to date

Today, we’re introducing Meta Llama 3, the next generation of our state-of-the-art open source large language model.Llama 3 models will soon be available on AWS, Databricks, Google Cloud, Hugging Face, Kaggle, IBM WatsonX, Microsoft Azure, NVIDIA NIM, and Snowflake, and with support from hardware platforms offered by AMD, AWS, Dell, Intel, NVIDIA, and Qualcomm.We’re dedicated to developing Llama 3

[misshiki]

“新しい 8B および 70B パラメーターの Llama 3 モデル”をリリース。

[stealthinu]

Llama3来た！GPT-3.5は超えてて、Claude3 SonnetやGemini1.5 Proあたりよりちょっといいか並ぶ。Command R+との比較はないが、今後はそのあたりとの戦いになりそう。