buildbot

19 ms·

331.
by buildbot 1y ago
Honestly it's a large enough library with enough weirdness and untested areas, footguns, and bugs that I'd deem it just as valid as React for example. Why did tensor_parallel have output += mod instead of output = output + mod? (T
332.
by buildbot 1y ago
My M1 MacBook Pro turns off the display & backlight that would show the notch as needed, for example right now in full screen Safari you could not tell there is a notch/menu bar area at all. It's actually just as good as it ca
333.
by buildbot 1y ago
I’ve long wanted to do this with an SDR and maybe some simple ML, build a dataset by driving by cars/things with frequencies of interest. Now I wonder if you can fingerprint antennas…
334.
by buildbot 1y ago
One thing to note is that MXFP4 is a block scaled format, with 4.25 bits per weight. This lets it represent a lot more numbers than just raw FP4 would with say 1 mantissa and 2 exponent bits.
335.
by buildbot 1y ago
Yes, as other noted AMD used VLIW for terscale in the 2000-6000 series. https://en.wikipedia.org/wiki/TeraScale_(microarchitecture) They are used in a lot of DSP chips too, where you (hopefully) have very simple branch
336.
by buildbot 1y ago
The law says it’s copyright infringement, not theft.
337.
by buildbot 1y ago
Claude code, being a CLI interface, might be more your style? Expensive though
338.
by buildbot 1y ago
Certainly possible! Or perhaps a block diffusion+autoregressive model or something like GPT 4o's image gen.
339.
by buildbot 1y ago
The math is really that hard? I have no idea what the soil or rock is, what happens when the first bomb hits it, the second, and then the third? Does the timing matter? Does the timing matter if it's 5 minutes between? 1 hour between?
340.
by buildbot 1y ago
Really? I think calculating the achieved overpressure to whatever structure underground after 6x 30k pound bomb impacts is far into the “genius” category. I’d wager you’d need a team of pretty smart people to even begin to get a wrong model
341.
by buildbot 1y ago
I think in theory, given terabytes of bitstreams, you might be able to get an LLM to output valid designs. Excepting hardened IP blocks, a bitstream is literally a sequence of sram configuration bits to set the routing tables and LUTs. Give
342.
by buildbot 1y ago
I’ve lived in Seattle 90% of my life and still finding and learning new stuff! It’s an awesome city (So are Portland and Vancouver BC!).
343.
by buildbot 1y ago
Nothing of technical merit haha. It’s a bit like vinyl vs CD for me, the analog component is cool. I also like the idea of swappable ADCs, not that any camera has actually done that.
344.
by buildbot 1y ago
Nitpick, technically they are CMOS, not CCD sensors ;) A modern CCD camera would be neat though
345.
by buildbot 1y ago
That’s a little uncharitable I think, you could know all those issues and be hoping that marketing and management will hold off on a launch until things are set. And the pricing made a huge difference - at 250, it would have been a differen
346.
by buildbot 1y ago
Yes, I immediately thought of Windows Aero too!!! I wasn’t able to enable it until I got a 9800GX2 a few years later, very cool at the time combined with the ability to have movies as your desktop background. It was a nice vibe.
347.
Emerging Properties in Unified Multimodal Pretraining(arxiv.org)
1 points by buildbot 1y ago|1 comments
348.
by buildbot 1y ago
New Multimodal in input & output 14B param LLM from Bytedance. Supports chain of thought reasoning and image generation.
349.
by buildbot 1y ago
That could be an interesting project in itself, take a simple 8 but microcontroller, a thermal camera, and some code that does different kinds of operations, see if you can train a classification model at least, or even generate the code ru
350.
by buildbot 1y ago
Not necessarily new - first ternary computer was around in 1959! https://en.wikipedia.org/wiki/Setun
351.
by buildbot 2y ago
Yep, I’ve only volunteered in commercial kitchens for events and stuff before, and the ones I’ve been in have essentially an assembly line for dishes from sinks to a dishwasher box that took 2-3 plastic cubes of dishes, and washed them in l
352.
by buildbot 2y ago
This paper has a bit about it, not very in depth though: https://www.imaging.org/common/uploaded%20files/pdfs/Papers/...
353.
by buildbot 2y ago
Microsoft has the MAIA 100 as well. No comment on their scale/plans though.
354.
by buildbot 2y ago
Oh nice, I didn't know that TIFF could handle that as well!
355.
by buildbot 2y ago
The algorithms for getting a useable image from a Foveon sensor are very non trivial from what I understand - the different layers don’t separate light perfectly into red, green, and blue bands, so there is some fancy cross layer processing
356.
by buildbot 2y ago
Oh interesting! This seems like it would be a good fit ;) Especially for really old setups that had RGB color wheels and multiple exposures, exactly like a multispectral astro image might. Phase one also has a multispectral capture system f
357.
by buildbot 2y ago
Maybe? I’m not familiar enough with DNG to say - possibly that wasn’t a thing when Phase One first started using IIQ? I doubt it was around when Sinar was - in fact the last two (esprit 65 & S30|45 ) Sinar backs do use DNG as an option!
358.
by buildbot 2y ago
I don’t think that’s true in general. As a sibling comments points out, this is not true for some DNGs - for example, the output of an iPhone is in DNG, but with many, many transforms already baked in. A DNG might even be debayered already.
359.
by buildbot 2y ago
Typically they are not to my knowledge! Though I am also not an expert. Most camera makers apply a fixed sensor profile, and possibly a dark frame to remove noise before writing out the values to whatever file. Some of them may apply lens o
360.
by buildbot 2y ago
I think that might be why a lot of camera makers don't care to use DNG - it's easier to make their own format and easy enough for others to reverse engineer it. One thing that open source libraries do tend to miss is that very imp

More ›