Meta announces a large language model which can be run on one GPU
Meta announced a new AI powered large language model called LLaMA-13B on Friday. It claims that this model can outperform OpenAI’s GPT-3 model despite it being \”10x\” smaller. Smaller AI models may lead to ChatGPT style language assistants running locally on devices like PCs and Smartphones. This is part of a family of new language models known as \”Large Language Model Meta AI,\” also called LLAMA.
Guillaume Lample, a project member, tweeted: \”Unlike Chinchilla or PaLM or GPT-3 we only use publicly available datasets, making our work reproducible and compatible with open-sourcing. Most existing models rely upon data that is not publicly accessible or undocumented.\”
Source:
https://arstechnica.com/information-technology/2023/02/chatgpt-on-your-pc-meta-unveils-new-ai-model-that-can-run-on-a-single-gpu/?utm_brand=ars&utm_medium=social&utm_social-type=owned&utm_source=facebook
Leave a Reply