Andrej Karpathy
nanoGPT, llm.c, and Zero to Hero — language models built from first principles.
2 evidence links
← Discover
Fine-tuning
Hire for people who know when not to train
Adapters, data pipelines, and a baseline that prompting or RAG has to beat.
4Published profiles in this specialty
100%Of those rows attach evidence
8Capabilities in the taxonomy
Why hire through this index
Fine-tuning is expensive to do badly. Profiles that list it should be able to show a dataset, an adapter method, and a comparison against the untuned system.
Available fine-tuning
4 published profiles listing this capability.
Hamel Husain
Independent ML consultant; the field notes on LLM evaluation.
2 evidence links
Jeremy Howard
fast.ai co-founder; the fastai library and Deep Learning for Coders.
2 evidence links
Sebastian Raschka
Build a Large Language Model (From Scratch); monthly LLM research reviews.
2 evidence links
Frequently asked
Should we fine-tune or write a better prompt?
Start with prompting, retrieval, and evals. Fine-tune when those have a measured ceiling. This capability is as much judgment as training.