Skip to content

Notes on OpenAI Q&A Finetuning GPT-3 Vs Semantic Search - Which to Use, When, and Why

A great video about finetuning vs semantic search. Finetuning teaches a model to write new patterns, not to have a theory of mind.

Bram Adams
Bram Adams
1 min read

Table of Contents

Overall Thoughts

This is really great video; it was very thorough and had great analogies. I didn't know about the unfreezing of the partial model, that's a neat fact!

I've had many office hours with people coming with finetune related questions that I believe would be better off 99% of the time with semantic search and the Hypothetical Document Embeddings (HyDE) in the rarest of cases perhaps.

Notated Transcript


a history of transfer learning


long live NLU, rip NLP


fine tuning is tweaking a task


only similarity b/w finetuning is q/a search is that they both use embeddings at some point


fine tuning unfreezes part of a model -- does not stop confabulation (hallucination)


unfreezing an entire model is expensive af


models barf out patterns, they do not have a theory of mind or knowledge

bigger models are more convincing, but a largest model will never know itself (as an information store)


finetuning is way more difficult than prompt engineering (10,000x harder)


finetuning at scale is very hard -- how much do we share in alignment


cost of fine tuning goes up with more data -- needs constant retraining


instruct -> question + body of info -> is answer in here?


finetuning teaches model to write a new pattern


formulate -> research -> criticize -> answer


dewey decimal is indexing on a smaller set of data, compile all the relevant research and scale it


Bram Adams

writer, programmer


Related Posts

Members Public

An Embedding Technique I've Been Enjoying

boats 'n' rows, boats 'n' rows, I gotta have me more boats 'n' rows

An Embedding Technique I've Been Enjoying
Members Public

September 2 2023

A lot of calories!

September 2 2023
Members Public

September 1 2023

Do you remember?

September 1 2023