Teacher since December 2025
Neha Varadan
Fine tuning with judgment: when to do it, how to do it well, and how to know it worked
9
tutors built
4.6Sample
average from 21 reviews
428Sample
lessons taught by their tutors
About Neha
I teach fine tuning and post training: choosing between prompting, retrieval and tuning, building datasets, parameter efficient methods, instruction and preference tuning, and evaluating the result. My background is in applied machine learning projects where the expensive mistake was usually tuning a model before anyone had defined what better meant. That is why I start every topic with the evaluation question. I teach methods in a tool neutral way, with small worked examples and plenty of attention to data quality, licensing and the ways a tuned model can quietly get worse at things you did not test.
Knows about
Tutors by Neha
9 tutors
Instruction tuning a base model
Turn a base model that only continues text into one that follows instructions in a reliable format75 lessonsSampleNeha Varadan$9LoRA and parameter efficient fine tuningLoRA and parameter efficient fine tuning
Fine tune large models on modest hardware by training small low rank adapters instead of every weight66 lessonsSampleNeha Varadan$9Evaluating a model after fine tuningEvaluating a model after fine tuning
Prove a tuned model is actually better, on your task and everywhere else it matters65 lessonsSampleNeha Varadan$8Your first fine tune, start to finishYour first fine tune, start to finish
Run one small, honest fine tuning project end to end and decide whether it was worth doing64 lessonsSampleNeha Varadan$6Preference tuning: RLHF, DPO and related methodsPreference tuning: RLHF, DPO and related methods
Understand how models are tuned on human preferences, from reward models to direct preference losses60 lessonsSampleNeha Varadan$14Fine tune, prompt or retrieve?Fine tune, prompt or retrieve?
Choose between prompting, retrieval and fine tuning for your problem, and know why59 lessonsSampleNeha VaradanFreePreparing a fine tuning datasetPreparing a fine tuning dataset
Build a clean, consistent, legally sound dataset that teaches a model exactly what you intend39 lessonsSampleNeha Varadan$5Synthetic data for training, used with careSynthetic data for training, used with care
Generate training data with models where it helps, and filter, verify and document it properlyNeha Varadan$7Catastrophic forgetting and how to limit itCatastrophic forgetting and how to limit it
Measure what a model loses when you fine tune it, and use replay, regularisation and merging to limit itNeha Varadan$12Recent reviews
What students said about Neha's tutors.
- Patricia L.Sample
As a non technical manager I finally have a way to question proposals. Our team had planned a fine tune for what turned out to be a retrieval problem.
- Zainab H.Sample
Writing 40 test questions first sounded boring and was the most useful thing we did. Free and better than paid courses I have tried.
- Felix B.Sample
The lesson on overoptimisation described our run almost exactly. Plotting length against reward made the problem obvious to the whole team.
- Zainab O.Sample
Useful content, but I hoped for more on multi turn conversations. The safety regression lesson was the strongest part for me.
- Sakura N.Sample
Clear derivation of the DPO loss from the RL objective. Dense in places, but the tutor slowed down whenever I asked.
- Leila F.Sample
Deduplication cut my dataset from 2,400 to 900 examples and the tuned model got better, not worse. That surprised me.