abhikul0

While personally I’m not using them for work, just tinkering, 128GB can run qwen 3.8 flash next at q4(~111GB gguf). That model seems to be quite capable from what I’ve read.

It’s quite evident that the api models can and have been nerfed at the provider’s discretion, then there’s the abliteration aspect too(api models refusing what they consider “harmful”) which makes the case for local.

Dude you have a lot to learn from me I am a good guy I love to talk about things I love you so very so very so very so good luck and good night and I’ll talk soon and I’ll see if you have any more questions please call my phone number and I’ll talk about you

IMO getting a studio for LLMs is a wiser choice since it doesn't heat up and there are no batteries to degrade.

For an M5 max top bin with 128GB Ram:

- mac studio: $5,899.00 (with 2TB ssd bump to match the default macbook sku)

- macbook 16: $7,399.00

Portability gets a hit but I feel you can have a thin client and let the studio do the heavy lifting back at your home/studio(heh).

Thanks AI.

I wanted to upgrade my air(gen4) but with these prices it's just not happening. Oh and no support for the perfectly fine apple pencil I already got with it is just....

Tame & King Gizz collab! As always, GTA with a banger soundtrack.

It works quite well on iOS, I am satisfied. - ublock lite blocks most of the ads, the rest get filtered through an Adguard Profile(no app install needed).

- Not sure if possible with other browsers but you can install userscript extensions too.

- I have a shortcut that extracts the text of any tab(you can run a simple javascript, document.body.innerText to get) to send to my local llm, this javascript execution is limited to just safari.

On a mac though, I only use it to stream 4k,HDR on Netflix.

This was jarring when moving from windows where you can literally scroll a page in a circular motion smoothly. Reading text with the locked scroll made me wonder how have people not noticed this or am I in the minority of reading like this lol. I mostly use chrome so this isn't a huge annoyance anymore and hope it behaves the same on macos 27 and beyond too.

Aside for a peculiar trackpad issue:

Does anyone else have problem scrolling diagonally on a page that's a bit zoomed in(like on HN, you zoom in a little to read the threads)? It just doesn't move smoothly and instead does a staircase shape traversal.

This used to occur on all browsers before upgrading to Tahoe, now it's mostly fixed but still persists on Safari.

For reference: https://www.reddit.com/r/MacOS/comments/1hh5yjo/diagonal_tra...

Is this happening only when upgrading the os or does it force enable apple intelligence on clean re-installs too?

Local inference? I see no other way than to sign up for a vercel account, so pass.

Sounds a lot like the little voice, already inside our heads.

> Most people didn't have the same kind of experience, they got the LimeWire version which was the equivalent of wading through a dollar store that’s just been ransacked and shit’s all over the floor.

Heh, I often went a step down and recorded internet radio using RadioSure. This little utility split each track into its own file which was pretty neat and handy to a younger me. Shoutout to Ryan Seacrests' AT40 for the weekly charts on the weekends, it kept my "collection" fresh ;-) Although, it was mostly 128-256kbps mp3 but it didn't matter to me, it was fun.

I tried to lookup a Star Trek episode of this premise and found a Voyager episode: https://en.wikipedia.org/wiki/Friendship_One_(Star_Trek:_Voy...

This is a likely scenario of misinterpretation between species haha.

Interesting

https://science.nasa.gov/mission/swift/swift-boost-mission/

While not exactly the hub style you mentioned, maybe this is as close to Jamming you can get.

https://support.apple.com/en-us/guide/garageband-ipad/chsf2f...

Wow, didn't know about Spotify's basic-pitch or that any song to midi model existed. Gonna be a good time sink to play around with. The standalone mac app build was straightforward, thanks.

On-device video subtitles generation is exciting, should help with watching videos on mute. This seems like a low hanging fruit that should've already been grabbed by an app but I can't find any.

The blank page is what you'd get if you cleared your Watch History. I use this to watch a new-to-me topic/channel or just a video that I know I'd only watch once and don't want in my recommended history in Incognito mode(although you could just remove that video manually from history too).

This makes it easy to somewhat steer the algorithm, I just open and scrub through a few videos of the same "kind" that I'm interested in and this then makes the youtube homepage a bit wieldy. This still populates unrelated videos but at a lower rate.

Relevant Smarter Every Day video: https://www.youtube.com/watch?v=VPSm9gJkPxU

I'll try to use that, but llama-server has mmap on by default and the model still takes up the size of the model in RAM, not sure what's going on.

Of course the swap is there for fallback but I hate using it lol as I don't want to degrade SSD longevity.

For the 9B model, I can use the full context with Q8_0 KV. This uses around ~16GB, while still leaving a comfortable headroom.

Output after I exit the llama-server command:

  llama_memory_breakdown_print: | memory breakdown [MiB]  | total    free     self   model   context   compute    unaccounted |
  llama_memory_breakdown_print: |   - MTL0 (Apple M3 Pro) | 28753 = 14607 + (14145 =  6262 +    4553 +    3329) +           0 |
  llama_memory_breakdown_print: |   - Host                |                   2779 =   666 +       0 +    2112                |

Mac has unified memory, so 36GB is 36GB for everything- gpu,cpu.

A usable quant, Q5_KM imo, takes up ~26GB[0], which leaves around ~6-7GB for context and running other programs which is not much.

[0] https://huggingface.co/unsloth/Qwen3.5-35B-A3B-GGUF?show_fil...

I hope the other sizes are coming too(9B for me). Can't fit much context with this on a 36GB mac.

Have you ever tried going to the model registry and seeing that the model was recently updated? What updated? What changed? Should I re-download this 20GB file?

I guess if you're not frustrated with things like this then sure, no need to stop using it.

Real Science video on the slug: https://www.youtube.com/watch?v=IH_uv4h2xYM

I thought github was being wonky but yeah, getting 401 Unauthorized error.

Edit: Can't view discussions as well.