Who is investigating on-premise alternatives to ChatGPT / Cloud-based GenerativeAI currently and if so, based on which available (open source) platforms?

18.9k views60 Upvotes4 Comments

Sort by:

IT Manager in Constructiona year ago

I am working with a local setup via Text Generation WebUI of Mixtral
https://huggingface.co/model-hub/Mixtral-8x7B-Instruct-v0.1
but I am also exploring the uncensored models as Vicuna13B.

CEO2 years ago

On NLP Cloud (https://nlpcloud.com) you have a comprehensive on-premise offer. For example you can deploy NLP Cloud's Dolphin generative model (an alternative to ChatGPT) on-premise.

CTO in Software2 years ago

We are currently evaluating Vicuña-33b

Senior Data Scientist in Miscellaneous2 years ago

Under https://slashdot.org/software/large-language-models/on-premise/ there are 42 on-premise alternatives listed. Some of them (e.g. GPT-J and GPT-NeoX from EleutherAI) are already around for quite a while.

Content you might like

What are the biggest technical challenges preventing organizations from being able to use their existing data to enable digital transformation?

Finding data and putting it to good use13%

Controlling the security and privacy of data45%

Understanding how data is currently being used20%

All of the above19%

None of the above1%

View Results

Do you think that the tech industry is starting to temper its expectations of what GenAI can do?

Yes, we’re starting to see the limits of GenAI74%

No, the hype still hasn’t died down26%

How do you tell if AI tools are genuinely helping your engineers, instead of just adding extra steps or confusion?

Has anyone had any success with evaluating the impact of using Generative AI tools such as GitHub's Copilot on the productivity or performance impact on developers? I see a lot of qualitative discussions about how developers say they are more productive, but how are you measuring that impact?