2 ms·
The two sentences prior are important: > We are also releasing a set of research models that are instruction fine-tuned. Initially, these fine-tuned models wil
by freedomben 4y ago
The two sentences prior are important:
> We are also releasing a set of research models that are instruction fine-tuned. Initially, these fine-tuned models will use a combination of five recent open-source datasets for conversational agents: Alpaca, GPT4All, Dolly, ShareGPT, and HH. These fine-tuned models are intended for research use only and are released under a noncommercial CC BY-NC-SA 4.0 license, in-line with Stanford’s Alpaca license.
The snippet you quoted is not talking about the main model in the announcement. It's talking about fine-tuned models based on other models. Stability has to respect the license of the originals. They cannot change it.
The main model is described higher up in the post and is permissible for commercial:
> Developers can freely inspect, use, and adapt our StableLM base models for commercial or research purposes, subject to the terms of the CC BY-SA-4.0 license
- btown 4y agoIt also appears that CC BY-SA-4.0 is GPL-compatible. Not a lawyer and this is not legal advice, but it certainly seems like one could operate their own StableLM server/service and allow proprietary code to use it over a network interface, much like one could use a GPL-licensed database system. https://creativecommons.org/faq/#can-i-apply-a-creative-commons-license-to-software https://creativecommons.org/faq/#can-i-apply-a-creative-comm...