HN Explorer

20 comments

[ 0.35 ms ] story [ 20.0 ms ] thread
mbuchel-hn
does this work similar to airllm? i am wondering how it would handle something like quantizing kimi k3 on a budget of 8 gbs, or is that something you are not attempting to solve yet?
rhgraysonii
Yes that is exactly what this does.
kennywinker
Could you explain what happens when you try to shoehorn a 2.4T parameter model into a 24gb m4 mac?
akshay_akula
Wondering the same thing but for 48gb M5 Max.
metalliqaz
extreme divergence would be my guess
jaylane
tried it out but based on the model sizing result i got i got an insufficient memory error when the server started running
rhgraysonii
If you could post an issue if you still have the error around that would be awesome.
jedbrooke
I gotta laugh at some of the models it suggests, for example:

> AnkitAI/Parable-Qwen3-4B-Claude-Fable-5-GGUF

you’re telling me you managed to fit Fable 5 into just 4B?

chompychop
I gotta laugh at your thought process: knowing Fable 5 is a large frontier model, you're telling me that the first thing that came to your mind on seeing that model name is that it's a quantized version of Fable? As opposed to a distillation/fine-tuning on Fable responses?
unrented7977
Don't make fun of people you think are ignorant, it's a pretty shitty look
chompychop
Well, then don't get all snarky and dismissive of things you might not be knowledgeable about ("you" here referring to OP).
metalliqaz
Well to be fair here... the title of this post doesn't mention fine tuning, it mentions quantization.
mannyv
Fyi that model name to me reads

Qwen3 4b params distilled/trained with fable 5

puttycat
This is really impressive. Can you say a bit about the underlying process? I'm guessing this is post-training qantization? Isn't PTQ also resource-intensive? (Ie might not work on any machine)