OpenAI now allows us to fine-tune GPT-3.5 models. I have tested and fine-tuned the model with my own dataset but the problem is the fine-tuned model generates the answer randomly, not correct based on my custom dataset.
Is there any way to make the model only answer from my own fine-tuned dataset?
This is a completely wrong approach (as you've already figured out).
As stated in the official OpenAI documentation:
Some common use cases where fine-tuning can improve results:
- Setting the style, tone, format, or other qualitative aspects
- Improving reliability at producing a desired output
- Correcting failures to follow complex prompts Handling many edge cases in specific ways
- Performing a new skill or task that’s hard to articulate in a prompt
Fine-tuning is not about answering a specific question with a specific answer from the fine-tuning dataset.
What you need to implement is a semantic search based on embeddings, as stated in the official OpenAI documentation:
When should I use fine-tuning vs embeddings with retrieval?
Embeddings with retrieval is best suited for cases when you need to have a large database of documents with relevant context and information.
By default OpenAI’s models are trained to be helpful generalist assistants. Fine-tuning can be used to make a model which is narrowly focused, and exhibits specific ingrained behavior patterns. Retrieval strategies can be used to make new information available to a model by providing it with relevant context before generating its response. Retrieval strategies are not an alternative to fine-tuning and can in fact be complementary to it.
You have two options:
