Thread

1. nico · 1790006937 · focus · HN ↗

   If you only need classification, and you can provide some training data, you can ask Codex/Claude to build an embeddings + logistic classifier model for you

   For emails, I get 95% accuracy with this method, with only 50-100 examples for training

   Training the model takes less than 5 minutes on a CPU

   The resulting model is <1MB, and inference is sub 100ms

   Some other cool things about this approach:

   * the model doesn’t train on some “ideal” or general classification, instead it learns your preferences

   * the model runs on pretty much any mobile device and can be retrained online on the device

   * privacy, the whole training and inference is 100% local, no data goes anywhere (except whatever you feed codex/claude while building the model)

   Note: to do a more general test, I made a classifier for the Banking77 dataset. The model is <10MB, trains in <30s on CPU and gets 94.5% accuracy, which puts it in the top 5?models by accuracy for that set (the best one is at 94.86%, but it’s 350MB in size and takes hours to train on a GPU).

   1. e12e · 1790038169 · focus · HN ↗

      > For emails, I get 95% accuracy with this method, with only 50-100 examples for training

      I'm assuming you only need to consider a single language for your emails?

      1. nico · 1790041126 · focus · HN ↗

         Mostly 2 languages

         Haven’t tried with more

         Do you have a specific use case?

         1. e12e · 1790064991 · focus · HN ↗

            I would need at least two languages, probably three. I was guessing this approach might be less multilingual than LLMs.

            In my case Norwegian, English and Japanese.

            1. dotancohen · 1790469440 · focus · HN ↗

               Are you in the Salmon business?

Unofficial Hacker News client; not affiliated with Y Combinator.