πββοΈMy AskAI vs Fine-tuning
Your AI support agent isn't built using "fine-tuning", but "embeddings", another AI technique.
Is this fine-tuning? Whatβs the difference?
Not quite, fine-tuning is generally used (and most effective) when you have a load of data in a prompt (input) and completion (output) format already and you want to use that data to create more similar output information based on new input information. It requires very structured and curated data to get the best results.
AskAI uses something called βEmbeddingsβ and the benefit of it is that you donβt need to pre-format your data most of the time. You can just upload your content and then start querying.
Last updated