Show HN: FinetuneDB – AI fine-tuning platform to create custom LLMs

Re:Show HN: FinetuneDB – AI fine-tuning platform to create c... by thomashop on Wednesday October 09, 2024 @04:20PM (#41792196)

These days, I'd say the easiest and most effective approach is to put the whole book in the context of one of the longer context models.

[ Parent | on HN ]

Re:Show HN: FinetuneDB – AI fine-tuning platform to create c... by felix089 on Wednesday October 09, 2024 @04:26PM (#41792265)

Agreed, for this use case probably the easiest way to go.

[ Parent | on HN ]

Re:Show HN: FinetuneDB – AI fine-tuning platform to create c... by swyx on Wednesday October 09, 2024 @05:40PM (#41792963)

(and most expensive)

[ Parent | on HN ]

Re:Show HN: FinetuneDB – AI fine-tuning platform to create c... by felix089 on Wednesday October 09, 2024 @06:22PM (#41793378)

Agreed too

[ Parent | on HN ]

Re:Show HN: FinetuneDB – AI fine-tuning platform to create c... by thomashop on Monday October 14, 2024 @06:46PM (#41842986)

Prompt caching to the rescue?

[ Parent | on HN ]

Re:Show HN: FinetuneDB – AI fine-tuning platform to create c... by KaoruAoiShiho on Wednesday October 09, 2024 @04:31PM (#41792295)

Not really, for something like gemini the accuracy and performance is very poor.

[ Parent | on HN ]

Re:Show HN: FinetuneDB – AI fine-tuning platform to create c... by farouqaldori on Wednesday October 09, 2024 @05:15PM (#41792723)

The magic behind NotebookLM can't be replicated only with fine-tuning. It's all about the workflow, from the chunking strategy, to retrieval etc.

For a defined specific use-case it's certainly possible to beat their performance, but things get harder when you try to create a general solution.

To answer your question, the format of the data depends entirely on the use-case and how many examples you have. The more examples you have, the more flexible you can be.

[ Parent | on HN ]