LocalLLaMA

1 readers

1 users here now

Community to discuss about Llama, the family of large language models created by Meta AI.

founded 10 months ago

MODERATORS

communick@poweruser.forum

Any good guides on building own datasets to then fine tune Mistral-based or similar models for completion? (alien.top)

submitted 10 months ago by LostGoatOnHill@alien.top to c/localllama@poweruser.forum

0 comments fedilink hide all child comments

Looking to move on in the next step of my LLM learning journey and:

a) generate a q&A dataset, say with GPT-4

b) use the dataset to instruction fine tune a 7B variant of mistral and evaluate

The Q might be to give me a sumamrised history of a company, with the dataset answer generated by GPT-4, to fine-tune the instruction fine tuned mistral 7B model.

If you know of any good guides for this, I’d highly appreciate, thank-you

EDIT: Reposted to fix title, god damn iPad auto complete!

no comments (yet)

sorted by: hot top controversial new old

there doesn't seem to be anything here