LLM are all the hype these days in AI development. Having the ability to train and deploy a LLM on your own data without running into issues where the data is sent to a third party is a must have for many organizations.
Databricks introduced their own LLM called Dolly https://www.databricks.com/blog/2023/04/12/dolly-first-open-commercially-viable-instruction-tuned-ll...
This idea is for Dataiku to develop a LLM and make it available to customers as a model you can tune for your use-case.
Hi @M1-Nick , we are looking into introducing LLM fine-tuning.
Would you be able to share more insight into what you are looking to achieve in your use case? Have you explored existing approaches? Any challenges or observations to share?
@Krishnathere is any news regarding DSS being able to manage LLM fine tuning?
As business users we are interested in this feature, due our focus is business process, data discovery, extraction and preparation.
FYI, earlier this week there was an announcement from Dataiku or some forthcoming features related to LLMs.