3 ms·
Almost. If your dataset contains questions and answers about your own projects documentation, then yes. The UX around how to prompt a fine-tuned model depends o
by aqader 4y ago
Almost. If your dataset contains questions and answers about your own projects documentation, then yes. The UX around how to prompt a fine-tuned model depends on the format of the dataset it's trained on.
One way you can do this is pass your documentation to a larger model (like a GPT3.5 / OSS equivalent) and have it generate the questions/answers. You can then use that dataset to fine-tune something like Llama to get conversation / relevant answers.
- underlines 4y agoto my understanding, fine tuning is slow and would be quite bad to update. embeddings seems to be the way to go. i don't understand it well enough, but it seems with the langchain framework you can create an embedding of your own data and submit it to the GPT API and i believe emeddings should be a similar principle in llama. at least i did it with diffusers in stablediffusion.