News for hackers. Stuff that matters.

A comment on Emerging Architectures for LLM Applications. Parent comment

Re:Emerging Architectures for LLM Applications by satvikpendem on Tuesday June 20, 2023 @09:05PM (#36413031)
Most enterprises likely won't be able to use OpenAI since they might have proprietary information, so setting up a good on-prem open source LLM can be necessary. In fact, I am doing so right now with my company.
[ Parent | on HN ]
Re:Emerging Architectures for LLM Applications by nonfamous on Tuesday June 20, 2023 @10:54PM (#36413816)
You can create your own instances of the OpenAI models in Azure to keep proprietary data with the bounds of your tenant there.
[ Parent | on HN ]
Re:Emerging Architectures for LLM Applications by abrichr on Tuesday June 20, 2023 @11:45PM (#36414161)
Interesting, can you please provide a link to the relevant documentation? The closest I found is https://azure.microsoft.com/en-us/blog/chatgpt-is-now-availa... and https://learn.microsoft.com/en-us/azure/cognitive-services/o... but wasn't able to find anything about the data provisioning terms you mention.
[ Parent | on HN ]
Re:Emerging Architectures for LLM Applications by rpeden on Wednesday June 21, 2023 @12:11AM (#36414303)
This might be what you're looking for: https://learn.microsoft.com/en-us/legal/cognitive-services/o...
[ Parent | on HN ]
Re:Emerging Architectures for LLM Applications by abrichr on Thursday June 22, 2023 @10:33AM (#36432753)
I believe you may be referring to this:

> Fine-tuned OpenAI models. The Fine-tunes API allows customers to create their own fine-tuned version of the OpenAI models based on the training data that they have uploaded to the service via the Files APIs. The trained fine-tuned models are stored in Azure Storage in the same region, encrypted at rest and logically isolated with their Azure subscription and API credentials. Fine-tuned models can be deleted by the user by calling the DELETE API operation.

My follow-up question is: how is this different from OpenAI's finetuning?

[ Parent | on HN ]
Re:Emerging Architectures for LLM Applications by arthurcolle on Tuesday June 20, 2023 @11:43PM (#36414140)
You don't have to necessary load "proprietary info" into the model to generate valuable data
[ Parent | on HN ]