omneity

12 ms·

181.
by omneity 1y ago
This looks great, congrats on the release! So what are the potential risks/downsides? As in what’s the tradeoff that is being made when relying on pipask?
182.
by omneity 1y ago
Congrats for shipping! Have you seen LiteLLM Gateway before? As a bonus it supports a lot of different providers. https://docs.litellm.ai/docs/simple_proxy
183.
by omneity 1y ago
This rings such a vague and distant bell... I'm several videos in and totally hooked, thank you for sharing. This would be an amazing interactive music app in VR, both to perform and to record trippy music videos.
184.
Meta Synthetic Data Kit(github.com)
1 points by omneity 1y ago|0 comments
185.
by omneity 1y ago
You are certainly right, managed platforms make finetuning much easier. But managed/closed model finetuning is pretty limited and in fact should be named “distribution modeling” or something. Results with this method are significantly
186.
by omneity 1y ago
I read your comment first and was like it can’t be that bad. I then opened the page, scrolled frantically down and boy oh boy were you right.
187.
by omneity 1y ago
This assumes the team deploying the RAG-based solution has equal ability to either engineer a RAG-based system or to finetune an LLM. Those are different skillsets and even selecting which LLM should be finetuned is a complex question, let
188.
by omneity 1y ago
It's okay for extracting simple things like addresses, or for formatting text with some input data, like a more advanced form of mail merge. I haven't evaled these tasks so YMMV. I'm exploring other possibilities as well. I s
189.
by omneity 1y ago
RAG is infinitely more accessible and cheaper than finetuning. But it is true that finetuning is getting severely overlooked in situations where it would outperform alternatives like RAG.
190.
by omneity 1y ago
128GB but it's not using much. I'm running Q4 and it's taking 17.94 GB VRAM with 4k context window, 20GB with 32k tokens.
191.
by omneity 1y ago
I'm using Qwen3-30B-A3B locally and it's very impressive. Feels like the GPT-4 killer we were waiting for for two years. I'm getting 70 tok/s on an M3 Max, which is pushing it into the "very usable" quadrant. W
192.
by omneity 1y ago
Looking at the Persian IPA table[0] for the letters you wrote, we get `/ʒ/` for `ژ` and `/tʃʰ/` for `چ` In Arabic[1], there are two close phonemes: `/dʒ/` for `ج` and `/ʃ/` for `ش` The difference in b
193.
by omneity 1y ago
Arabic is not Aramaic. Please correct your sources. I’m also quite curious about the sounds of “ch” and “zh” which exist in Arabic as ش and ج, or did you mean something else?
194.
by omneity 1y ago
It doesn’t have to be contiguous AI-written chunks. This could also mean X% accepted AI suggestions on Y% codebase where X*Y < 30%. AI suggestions can also be as simple as autocomplete and still be counted for the sake of engagement metr
195.
by omneity 1y ago
You are right, it doesn’t make sense. If that was a real reason then why not augment their current staff with AI tools to “scale content to more learners” and still have a sense of quality and human touch. Encrappification at play.
196.
by omneity 1y ago
I already found Duolingo quite bad, struggling to make it past the repetitive and often not-so-sensical question/answers, but this finally proves it’s not for me.
197.
by omneity 1y ago
I would like to challenge the fundamental premise of the article. Just because you can generate 50 PRs doesn’t mean you should. In fact the same bottleneck they’re describing is present if you have 50 coders in your team. The problem theref
198.
by omneity 1y ago
Excellent release by the Qwen team as always. Pretty much the best open-weights model line so far. In my early tests however, several of the advertised languages are not really well supported and the model is outputting something that only
199.
by omneity 1y ago
For the uninitiated, what this comment means is that LEO-orbiting payloads have a naturally limited lifespan as the orbit decays over time on its own causing the LEO satellites to get closer to earth and burn down to smithereens due to fric
200.
by omneity 1y ago
This somehow reminds me of Agent-3 from [0]. 0: https://ai-2027.com
201.
by omneity 1y ago
You probably know this already but I’ll leave it for posterity. https://mermaid.js.org/syntax/entityRelationshipDiagram.html
202.
by omneity 1y ago
Yeah those are pretty cool and on my radar! And thanks for sharing the tip :) Just checked your work on covary and it's pretty rad! What's your backend like?
203.
by omneity 1y ago
> So, our question was: what happens to brain fingerprints when we lose consciousness, such as during the artificial sleep induced by general anesthesia? The most important question here, and yet the article keeps talking about "fin
204.
by omneity 1y ago
Maybe by doing it enough times o3-mini will end up reimplementing itself?
205.
by omneity 1y ago
A long time ago, I believed I wanted to implement a custom interpreter in Ruby. And then I discovered method_missing, and later on the rabbit hole that is Metaprogramming in Ruby, and that worked wonders to get the interpreter thought of my
206.
by omneity 1y ago
If I was the paper author I would have done it and include the results as an appendix or a repo.
207.
by omneity 1y ago
Or your model wasn't trained so well (weights are too spiky)
208.
by omneity 1y ago
This is similar to the halting problem. You can only write an effective policy if you can predict all the side effects and their ramifications. Of course you could do like deno and other such systems and just deny internet or filesystem acc
209.
by omneity 1y ago
I guess the "lucky 10000" effect is extremely strong today! This is like an OS developer who has never heard of Linux.
210.
by omneity 1y ago
Block-based editors maybe?

More ›