Open-source artificial intelligence is the application of open-source practices to the development of artificial intelligence resources.
Many open-source artificial intelligence products are variations of other existing tools and technologies which have been shared as open-source software by large companies.[1]
Companies often develop closed products in an attempt to keep a competitive advantage in the marketplace.[2] A journalist for Wired explored the idea that open-source AI tools have a development advantage over closed products, and could overtake them in the marketplace.[2]
Popular open-source artificial intelligence project categories include large language models, machine translation tools, and chatbots.[3]
For software developers to produce open-source artificial intelligence resources, they must trust the various other open-source software components they use in its development.[4][5]
Main article: LLaMA |
LLaMA is a family of large language models released by Meta AI starting in February 2023.[6] Meta claims these models are open-source software, but the Open Source Initiative disputes this claim, arguing that "Meta’s license for the LLaMa models and code does not meet this standard; specifically, it puts restrictions on commercial use for some users (paragraph 2) and also restricts the use of the model and software for certain purposes (the Acceptable Use Policy)."[7]
Model | Developer | Parameter count | Context window | Licensing |
---|---|---|---|---|
LLaMA[6] | Meta AI | 7B, 13B, 33B, 65B | 2048 | —— |
LLaMA 2[8][9] | Meta AI | 7B, 13B, 70B | 4k | Custom Meta license |
Mistral 7B[10] | Mistral AI | 7 billion | 8k[11] | Apache 2.0 |
GPT-J[12] | EleutherAI | 6 billion | 2048 | Apache 2.0 |
Pythia[13] | EluetherAI | 70 million - 12 billion | —— | Apache 2.0 (Pythia-6.9B only)[14] |