fine_tune

7 ms·

1.
by fine_tune 1y ago
LLMs are only 3-5 years old (NLP is much older OFC), for all we know they'll be a dead end in research like LSTM are today - LLM/Multi Modal just look super hot. "Attention is all you need" was released in 2017, it took
2.
by fine_tune 1y ago
Its cloudflare trying to enshittify the internet with micro transactions[0] and take their N% cut (of course it will start at like 2% but ask any uber driver how thats going). The problem is the arguments they make for why this should happe
3.
by fine_tune 1y ago
I got rage baited by this so hard, cant comprehend thinking this way. Hung out with PhD's, economists, bankers, trust find kids, scientists, and artists - who maybe weren't top tier enough, but none thought this way. Literally the
4.
by fine_tune 1y ago
I was going to argue "LLM's need code samples to-do well on languages and if we are honest C# is a language mostly held in private repo's" but Github's 2024 report[0] says its the 5th most used language (I'm t
5.
by fine_tune 1y ago
I'm no ruby expert, so forgive my ignorance, but it looks like a small "NER model" packaged as a string convince wrapper named `filter` that tries to filter out "sensitive info" on input strings. I assume the NER mo
6.
by fine_tune 1y ago
Yes and no, for human search - its kinda neat, you might find some duplicates, or some nearby neighbour bugs that help you solve a whole class of issues. But the cool kids? They'd do something worse; They'd define some complicated
7.
by fine_tune 1y ago
RAG is taking a bunch of docs, chunking them it to text blocks of a certain length (how best todo this up for debate), creating a search API that takes query (like a google search) and compares it to the document chunks (very much how your
8.
by fine_tune 1y ago
You bought a house that had a murder X years ago and are wondering if your guilty for the murder, probably not - aslong as you don't do more murder in it. I suppose real life is more interesting though, the guy who picked up the domain
9.
by fine_tune 1y ago
Attempting to train this on a real workload I converted over the weekend after, "step" 8M~ so far and rarely scores above 5% and most are 0% but has scored 60% once 7M~ steps ago. Adding more than 1 GPU didn't improve speed b
10.
by fine_tune 1y ago
Love unsloth btw, use it for some other stuff at work, GRPO stuff was fun :) I know its coming but "mUlTi GpU PlZ" :pleading: <3
11.
by fine_tune 1y ago
> I'm interested to know if anyone is using fine-tuning to train a model on proprietary or in-house codebases and documentation. I've done it, 1/2 the team though it was great 20% of the time, 1/2 the team hated it fr