phantompeace
phantompeace on Hacker News
Not sure if I agree. Remember the error rates stack up every turn of the conversation. If a 2bit model is 10% more wrong than the fp16 version, over 10 turns that adds up
Only real ones remember 110MB
History will look back on people like you quite unfavourably.
Nobody called anyone schizophrenic. They noted that the irrelevant link spam in a reply was reminiscent of schizophrenic posting. If you've spent any sort of time on places like 4chan, you'd understand. I don't think posting the HN guidelines is warranted here.
Care to explain how the terroristic Epstein Regime triple tapped a school in Iran recently? I'm waiting for your admonishment
I'm not surprised given the incessant bot activity. Top contributors are most likely bots. Other than that, I remember Ghislaine Maxwell as being a supermod across multiple high profile subreddits, so banning them seems like a good move.
Why did you consciously click into this article and comments section if you don't want to hear about it? You're free to scroll on at any time.
Which recent jailings were made for social media posts? Are you talking about the ones where people invited violence against migrants in hotels by calling for things like setting fire to their dwellings etc? Because that's a crime regardless of posting it on social media.
Commonly referred to as the Epstein Regime
Considering the very small difference between just SFT on the student model as compared to SFT + DPO on a proxy, doesn't it make sense to concentrate on ensuring the SFT dataset is perfect rather than sorry about DPO etc? And just train directly on the student model?
Why are we believing in marketing lies? For instance, crypto is useful (in a narrow context) even if it's 99.99% scams, and has positively impacted at least some people (the unbanked), even if it fell short of replacing the world's banking system. It's good tech.
Similarly with LLMs, it's good tech, but I do not believe for a single second that these things will ever replace anyone who is remotely skilled. People will be augmented by them. They won't be fully replaced. Well, I hope not anyway.
Do you really think an LLM will replace a surgeon?
How do you know it's being served at wildly below cost? AFAIK API pricing across the board is actually higher than what inference would cost? Yes there are subscription plans that subsidise tokens but they exist primarily to hook people in, so as to encourage their workplace to adopt it, whereby they then have to pay for an enterprise plan with zero subsidies like the subscription plans have. Training is very expensive, but you only do that once.
I predict costs will fall.
You're gonna smash it. Regardless of what your GCSE results are, stuff like this is what will take you far in life. All the best for your exams, but don't lose any sleep over them either.
You posted real discussions and opinions, the fact that you flagged them doesn't change the fact that you took some sort of value to the tune of spending effort posting replies beforehand. What I'm struggling to understand is why the call to shut down discussion on this topic when Israel is quite blatantly committing war crimes. Don't pretend you don't see it.
Read their submission history. They talk about "radical extremist Islamist regime in Iran" all the time. Seems rather odd to be this cagey when people are getting sick and tired of Israel's constant war crimes.
Less than 28 days ago you were commenting multiple times on Iran, Islamists and geopolitics involving war. Why are you suddenly so jumpy? I don't get it. Why is it suddenly unacceptable to talk about quadruple tappings on medical workers by Israel?
Wouldn't it be easier to put the unsophisticated blog behind cloudflare
Unfortunately I have zero experience with the Jetson family, and Parakeet itself is a pain to get running IMO - I took the easy option and used the ONNX version
What hardware are you running? Parakeet runs on nvidia and Mac and it’s way faster than Whisper. And I’ve had issues with training Qwen3 (and even Qwen2.5 but I think I was masking stop tokens wrong). I’ve had success with Gemma 3 though, and they have some really small models (270m and 1b). Maybe 270m for just transcript cleaning? I wonder if the 1b model can handle the transcript analysis…
Thank you for reading!
Cost is a big factor - i really want to make models that can run on average CPU only machines so most of the world can benefit, rather than needing expensive GPUs or an internet connection + subscriptions. Another big factor is privacy (you don't need to trust a 3rd party with your inputs).
As for the hyperparameters, pure bruteforce trial and error. It feels more like a dark art than a science. You roll the dice and then start tweaking things until the loss looks like it's dropping nicely and consistently, and the checkpoints are starting to output things resembling what we want. I sometimes do inference using checkpoints just to get a feel of if the model is learning (regardless of loss)
Posted by
phantompeace on Monday September 01, 2025 @02:34PM
from the show-hn dept.
https://bilawal.net/post/finetuning-llama32-3b-for-transcripts/
I wrote a small local tool to transcribe audio notes (Whisper/Parakeet). Code: https://github.com/bilawalriaz/lazy-notesI wanted to process raw transcripts locally without OpenRouter. Llama 3.2 3B with a prompt was decent but incomplete, so I tried SFT.
I fine-tuned Llama 3.2 3B to clean/analyze dictation and emit structured JSON (title, tags, entities, dates, actions).Data: 13 real memos →...
The UN and the ICJ are somehow less trustworthy than the Netanyahu government, which is being protested against by many Israelis? Perhaps one should stop and consider why a single country has amassed so many resolutions against itself compared to the rest of the world - no, the rest of the world must be wrong. Netanyahu must be right. /s
So now the people who can buy upvotes get to write history?
What is the Essence? In my head I picture movement or electrical fields propagating from some sort of proto cell like structure, but the more I try to “zoom” in, the more I try to find the finger that caused the ripple on the pond surface, so to speak