Hacker Newsnew | past | comments | ask | show | jobs | submit | marinhero's commentslogin

Excellent. This plus OPDS will make for easier transfer of files locally.

Yes and with kavita there's now even progress sync with koreader! I use it on my kindles too.

Or even not locally!

What a great thing you built!


Serious question but if it hallucinates about almost everything, what's the use case for it?


Fine-tuning for specific tasks. I'm hoping to see some good examples of that soon - the blog entry mentions things like structured text extraction, so maybe something like "turn this text about an event into an iCal document" might work?


Google helpfully made some docs on how to fine-tune this model [0]. I'm looking forward to giving it a try!

  [0]: https://ai.google.dev/gemma/docs/core/huggingface_text_full_finetune


Fine tuning messes with instruction following and RL'd behavior. I think this is mostly going to be useful for high volume pipelines doing some sort of mundane extraction or transformation.


This is exactly the fine-tuning I am hoping for, or I would do if I had the skills. I tried it with gemma3 270M and vanilla it fails spectacularly.

Basically it would be the quickadd[1] event from google calendar, but calendar agnostic.

[1] https://developers.google.com/workspace/calendar/api/v3/refe...


It's intended for finetuning on your actual usecase, as the article shows.


I feel like the blog post, and GP comment, does a good job of explaining how it's built to be a small model easily fine tuned for narrow tasks, rather than used for general tasks out of the box. The latter is guaranteed to hallucinate heavily at this size, that doesn't mean every specific task it's fine tuned to would be. Some examples given were fine tuning it to efficiently and quickly route a query to the right place to actually be handled or tuning it to do sentiment analysis of content.

An easily fine tunable tiny model might actually be one of the better uses of local LLMs I've seen yet. Rather than try to be a small model that's great at everything it's a tiny model you can quickly tune to do one specific thing decently, extremely fast, and locally on pretty much anything.


It's funny. Which is subjective, but if it fits for you, it's arguably more useful than Claude.


Because that's not the job it was designed to do, and you would know by reading the article.


The same as having a goldfish. You can train it to do a trick I guess.


Games where you need NPCs to talk random jiberrish.


Nothing, just like pretty much all models you can run on consumer hardware.


This message brought to you by OpenAI: we're useless, but atleast theres a pay gate indicating quality!


robotic parrots?


An army of troll bots to shift the Overton Window?


oh no now we'll never hear the end of how LLMs are just statistical word generators


This sounds like bro science. Having boring sessions is not the point, Z2 training is designed to build endurance, improve cardiovascular efficiency, and increase your body's ability to burn fat as a fuel source at moderate intensities. It’s not about enduring boredom or embracing “mental pain” but rather about consistently training at a level that is sustainable for extended periods.


If the truly cared about that they would let us run a desktop OS on their iPhones so we don’t carry laptops all around.


How many android users actually use their phone for this?

The utility of a MacBook (or any laptop) is the display, keyboard, and trackpad, not the compute


Have you tried using an nvim distribution? They take care of all the plugin loading, config and testing. They stay reasonably updated too.


I tried using LazyVim, but I didn't put much effort customizing it so I missed my custom config and I found it somewhat janky


I think they are a good middle ground, but you're still left with some of the busy work. Further, you're a bit more at the mercy of the maintainer. Likely, Lazyvim isn't going anywhere, but it isn't out of the realm of possibilities either.


What distribution would you recommend?


I recently started with Kickstart paired with this video, it's really easy and beginner friendly.

https://github.com/nvim-lua/kickstart.nvim

https://www.youtube.com/watch?v=m8C0Cq9Uv9o


I think lazyvim is one of the more successful/dependable ones


I get frustrated seeing this go into the iPad and knowing that we can't get a shell, and run our own binaries there. Not even as a VM like [UserLAnd](https://userland.tech). I could effectively travel with one device less in my backpack but instead I have to carry two M chips, two displays, batteries, and so on...

It's great to see this tech moving forward but it's frustrating to not see it translate into a more significant impact in the ways we work, travel and develop software.


> instead I have to carry two M chips

What's the incentive for Apple to unify them, since you've already given them the money twice?


> Not even as a VM

WWDC is next month. There's still a chance of iPadOS 18 including a Hypervisor API for macOS/Linux VMs on M4 iPads.


I hope for this every single year. I just don’t see it happening. But I hope I am wrong.


2022, https://appleinsider.com/articles/22/10/20/apple-rumored-to-...

> A leaker has claimed that Apple is working on a version of macOS exclusive for the M2 iPad Pro ... the exclusivity to M2 iPad Pro could be a marketing push. If the feature is only available on that iPad, more people would buy it.

Based on the M4 announcement, vMacOS could be exclusive to the 1TB/2TB iPad Pro with 16GB RAM that would be helpful for VMs.


At this point, you would have a better chance of running your own apps by relocating to the EU ;)


yup - im honestly tired of the Apple ~~~jail~~~ ecosystem.

I love the lower power usage/high efficiency of ARM chips but the locked down ecosystem is a drag.

Just the other day, I was trying to get gpu acceleration to work within a vm on my m1 mac. I think it’s working? But compared to native it’s slow.

I think it’s just a misconfig, somewhere (ie, hypervisor or qemu or UTM or maybe the emulation service in vm).

On other systems (intel/amd + nvidia/radeon) this is more or less a “pass through” but on mac it’s a different beast.


gpu passthrough for VMs is not supported on apple silicon period afaik. there may be some "native" renderer built on top of metal but apple doesn't support SR-IOV or "headless passthrough".

https://chariotsolutions.com/blog/post/apple-silicon-gpus-do...

otoh no, it is not "more or less [automatic]" in other hardware either, SR-IOV has been on the enthusiast wishlist for a ridiculously long time now because basically nobody implements it (or, they restrict it to the most datacenter-y of products).

intel iGPUs from the HD/UHD Intel Graphics Technology era have a concept called GVT-g which isn't quite SR-IOV but generally does the thing. Newer Xe-based iGPUs do not support this, nor do the discrete graphics cards.

AMD's iGPUs do not have anything at all afaik. Their dGPUs don't even implement reset properly, which is becoming a big problem with people trying to set up GPU clouds for AI stuff - a lot of times the AMD machines will need a hard power reset to come back.

NVIDIA GPUs do work properly, and do implement SR-IOV properly... but they only started letting you do passthrough recently, and only 1 VM instance per card (so, 1 real + 1 virtual).

Curious what you're using (I'm guessing intel iGPU or nvidia dGPU) but generally this is still something that gets Wendell Level1techs hot and bothered about the mere possibility of this feature being in something without a five-figure subscription attached.

https://www.youtube.com/watch?v=tLK_i-TQ3kQ

It does suck that Apple refuses to implement vulkan support (or sign graphics drivers), I think that's de-facto how people interact with most "hardware accelerated graphics" solutions in vmware or virtualbox, but SR-IOV is actually quite a rare feature, and "passthrough" is not sufficient here since the outer machine still needs to use the GPU as well. The feature point is SR-IOV not just passthrough.


UTM can be built for iOS.


Hypervisor.framework is not exposed without a jailbreak which makes this quite limited in terms of usability and functionality.


best you can hope for is cpu pass through. Gl with using the rest of the chip


Think the play is "consumer AI". Would you really write code on an iPad? And if you do, do you use an external keyboard?


Tablets are the perfect form factor for coding because you can easily mount them in an ergonomic position like this: https://mgsloan.com/posts/comfortable-airplane-computing/

Most laptops have terrible keyboards so I'd be using an external one either way.


Those keyboards are absolutely ridiculous, sorry.


Yes. If I’m plugging it to a thunderbolt dock I’d expect it to work like a MacBook Air


I loved Children of Time. Made me appreciate multi-generational stories for the first time. Also, the ants!


If you haven't read it, the Octavia Butler trilogy called Lilith's Brood is a great multi-generational story as well.


You can download it and run it with [this](https://github.com/oobabooga/text-generation-webui). There's an API mode that you could leverage from your VS Code extension.


How well do LLMS like this work with a non-English language? Or are these open source models limited to English?


Quite a few of the top ranked models on this leaderboard are multilingual: https://huggingface.co/spaces/mteb/leaderboard

https://huggingface.co/BAAI/bge-large-en-v1.5 FlagEmbedding for example describes itself as covering Chinese and English.


Stability has a Japanese port which is getting lots of work https://twitter.com/StabilityAI_JP/status/171699857824440759...


This is not an embedding model though. Yes you can always extract some embeddings from somewhere, but for most LLMs those won't perform well for retrieval (which makes sense as it's not what the models are optimizing for)


This isn't an embedding model, but it is a group of people working in this general area in a language other than English. Maybe they'll get to an embedding model next?


That depends on whether the training data contained languages other than English.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: