filterfiber

9 ms·

61.
by filterfiber 3y ago
Realistically I think you'll often want both depending on what you're doing. Especially for things like blur detection, what's your acceptable specificity, acceptable/scale of performance, where you're running the a
62.
by filterfiber 3y ago
Yeah the client is unfortunately closed source. The only reason I'm okay with that is that it's a client side application that works with markdown, and their schema for "canvas" is opensource. > It's also not sel
63.
by filterfiber 3y ago
> My bias is that 50% of ADHD'ers have anxiety disorders due to noradrenaline running high. I think this is partially the case. I think the glutamatergic system is messed up which causes the release of too much norephenephrine and
64.
by filterfiber 3y ago
I'm curious too. I know cocoa contains a lot of Theobromine which is very similar to caffeine (caffeine is Methyl-theobromine) although it does behave a little different. Cocoa also has a little caffeine as well. I would think these tw
65.
by filterfiber 3y ago
Obsidian with plugins!!!! You can create some very powerful workflows with dataviewer, properties and templates. Whatever note taking style you want really, automatically moving them to folders, etc. I highly recommend using unique notes wi
66.
by filterfiber 3y ago
> Has anyone come across a solution where model can iterate (eg, with prompts like "move the bicycle to the left side of the photo")? It feels like we're close. Emu can do that. The bluejay/toronto thing may be addres
67.
by filterfiber 3y ago
fwiw this specific subreddit has very "hacker news" like conversations, and is very technically focused. So this is a link to a actual conversation about it.
68.
by filterfiber 3y ago
Especially since it only has like 8 MB of memory I think?
69.
by filterfiber 3y ago
The light spectrum from LEDs is undesirable. But otherwise yes they would be less hot.
70.
by filterfiber 3y ago
I don't think the article mentions it, how well does the rpi 4 and 5 do for inference with whisper especially v3?
71.
by filterfiber 3y ago
I've always struggled with where people draw the line. Video scripts have been near algorythmic for humans already. Does using chatgpt to make your script count as AI made? What if you gave it an outline, walked through it, and guided
72.
by filterfiber 3y ago
> Might it be that people are just used to Google and not that it really produces better results? Maybe, but for me it struggles to recommend relevant reddit articles, copy+pasted error messages were hit/miss. It just feels like it
73.
by filterfiber 3y ago
Most people who utilize this hardware aren't programming kernels directly for the GPU, they're using abstraction layers like pytorch, tensorflow, etc. For the developers of those type of frameworks, cuda itself offers a lot of lib
74.
by filterfiber 3y ago
Several of the GQ counters have serial access, does anyone know if you can get the individual events from those?
75.
by filterfiber 3y ago
There are esp-lora boards including lipo power management you can start with. The ESP has a few deep sleep modes, and there's a lot you can do to optimize them. I highly recommend Andreas Spiess on youtube. EDIT: Heads up that moisture
76.
by filterfiber 3y ago
> I don’t think it’s a power issue as it’s getting 5V1A from a power outlet directly to USB-C into the device. It's not the total voltage/wattage the PSU can provide, but the voltage at the processor. The ESP's varying cur
77.
by filterfiber 3y ago
This is an excellent write up! > Seeed Studio XIAO ESP32S3/C3, WaveShare ESP32S3 Zero, Unbranded ESP32-WROOM with OLED, Orange Pi Zero W (untouched), Raspberry Pi Zero W (L->R, T->D) After testing all of these, the only one re
78.
by filterfiber 3y ago
tl;dr - google has big computer (a lot of TPUs) they just showed they could indeed make 50k TPUs do some flops. With no paper this is just a marketing press release - the only takeaway is that existing tech stacks can utilize it probably.
79.
by filterfiber 3y ago
IIRC apple is actually going this same route as proton - wine + dx12/11 to metal
80.
by filterfiber 3y ago
A bit tongue in cheek - I said "instantly destroy" because if the main selling point is an ai voice assistant, then people would just use what's already built into their phone/watch/airpods instead of paying 600$ if
81.
by filterfiber 3y ago
Doesn't the projection require one hand to be up and in a flat position, and the other hand interacts with it? Meaning it requires two hands? Apple's implementations are for 1 hand operation. You can operate the watch's touch
82.
by filterfiber 3y ago
The apple watch can already Pay, music, communication without a phone. Siri itself is lacking but I expect that to change with an LLM soon. You can already lock down your phone to prevent distracting apps. I've skeptical that people wo
83.
by filterfiber 3y ago
Genuinely asking, problem does this solve? And won't apple instantly destroy them the moment they use an LLM for siri? Why wouldn't I use my existing watch/phone/earbuds/pods instead of paying 600$+subscription for
84.
by filterfiber 3y ago
Is this fully local? I glanced around and didn't see any mention of chatgpt/gpt/codex so I'm thinking it is?
85.
by filterfiber 3y ago
The CI shows the entire team that the tests are passing, and doesn't require other devs to pull it and run the tests themselves. It prevents devs from accidentally forgetting to run tests, or thinking "This is just a typo fix, it
86.
by filterfiber 3y ago
Woah that's massive progress this year, dang.
87.
by filterfiber 3y ago
Maybe I'm remember incorrectly but I was in middleschool when the RPI 1 came out and my dad got me one. There was tons of guides, tutorials, etc. when I got it, which I think was close to launch? The os image was already made, tutorial
88.
by filterfiber 3y ago
Honestly I think a decent chunk of the complaints are from people who are used to windows which would be using 4-6gb of that ram just on the desktop. That said I still wouldn't recommend 8gb to anyone.
89.
by filterfiber 3y ago
The moment you can't fit the model in VRAM and have to use PCIE the performance impact is MASSIVE. Depending on the type of model/usecase/how much you can fit in VRAM it very well may still be usable, however compared to fitt
90.
by filterfiber 3y ago
Oh nice. For a few projects the pi 4 is plenty (some I can make due with a 3b), and for 15$ that'd be awesome (assuming software is okay which... for non-pis is dicey)

More ›