sorenjan

23 ms·

331.
by sorenjan 3y ago
I'm referring to using the modeled scene to bind gaussian splats to an existing mesh. > Binding New 3D Gaussians to the Mesh > This binding strategy also makes possible the use of traditional mesh-editing tools for editing a Gaus
332.
by sorenjan 3y ago
Yes, it's possible to create gaussian splats from a mesh. See for example step 3 in SuGaR: https://imagine.enpc.fr/~guedona/sugar/
333.
by sorenjan 3y ago
You'll note that AMD (and Nvidia, but that's expected) is unfortunately missing from the members list for UXL, which seems to be based on Intel's oneAPI which in turn is based in part on Khrono's SYCL: > Our mission:
334.
by sorenjan 3y ago
Well the point is that ä is one character, not two. It shouldn't be "a with two dots on it", it should be ä. It's its own letter with its own key on Swedish keyboards. MacOS apparently normalizes it to be two characters,
335.
by sorenjan 3y ago
I sometimes see texts where ä is rendered as a¨, i.e. with the dots next to the a instead of above it even though it's a completely different letter and not a version of a. I managed to track the issue down to MacOS' normalization
336.
by sorenjan 3y ago
There's something strange going on, the comment you replied to was posted three days ago according to the user's profile, but it says 43 minutes ago here. You probably read it when it was new, I recognized it as well. https:&#x2F
337.
by sorenjan 3y ago
Does anyone have any recommendations for good local PDF readers for Windows? I've been reading a lot of various papers recently, and clicking on a citation in Acrobat reader is very frustrating. The document scrolls to show the citatio
338.
by sorenjan 3y ago
What does the learned downsampler add? Isn't the output of the algorithm the upsampled features that's fed into the downsampler?
339.
by sorenjan 3y ago
Humans will use LLMs to produce long texts from a brief description, and then other humans will use LLMs to make a summary of that text. Maybe eventually we can make this communication more effective.
340.
by sorenjan 3y ago
"Among adults, those ages 18 to 29 are most likely to say they regularly get news on TikTok. About a third of Americans in this age group (32%) say they regularly get news there" https://www.pewresearch.org/short-r
341.
by sorenjan 3y ago
You can drop similar frames with ffmpeg: https://stackoverflow.com/questions/37088517/remove-sequenti... The most obvious use is of course to reencode the video after you did your screen recording with your usual
342.
by sorenjan 3y ago
67.8% of the respondents was positive to a Swedish Nato membership in 2023. https://www.scb.se/hitta-statistik/statistik-efter-amne/demo...
343.
by sorenjan 3y ago
That does make it a lot better, but at the same time makes the footnote even more of a deliberate statement that could have been left out.
344.
Spotify is about to put Apple's new App Store policies to the test(theverge.com)
3 points by sorenjan 3y ago|0 comments
345.
by sorenjan 3y ago
Changing a quote to change "his" to "theirs" seem like a very Rust community thing to do. > Updated to use modern pronouns. https://dada-lang.org/docs/about/
346.
by sorenjan 3y ago
I don't understand why AMD doesn't cooperate with Intel to push SYCL as the standard GPGPU and heterogeneous programming method. Intel is good with software, SYCL is an open standard so both companies would benefit from the same c
347.
by sorenjan 3y ago
Youtube recommended a similar video where they tested various patterns. Spreading paste over the the whole heat spreader resulted in a small air pocket, while the same method with an additional paste dot in the middle didn't have such
348.
by sorenjan 3y ago
I've seen reports and debates that modern heat spreaders are too large for the pea method, and that an X pattern is more common now. Here's one of the test videos you can find on Youtube about the subject: https://youtu
349.
by sorenjan 3y ago
I've seen ads in embedded Youtube videos. https://support.google.com/youtube/answer/132596?hl=en
350.
by sorenjan 3y ago
It should be relatively straightforward to build something with Visualequation, latex2sympy, SymPy, and maybe IPython's QTConsole. https://github.com/daniel-molina/visualequation https://pypi.org/p
351.
by sorenjan 3y ago
Finally treating small battery powered devices as embedded devices rather than small PCs. Still no match for Garmins's battery life though.
352.
by sorenjan 3y ago
I know Germany dub most video, but wouldn't a seven year old be able to read subtitles? It's a great way for her to learn English, it's how most Swedes learn it before starting school. I think there's a pretty strong cor
353.
by sorenjan 3y ago
What's the use case of models this small? Can you use the "knowledge" encoded in them and ask them questions and get relevant answers, or are they used as text processors to summarize documents etc?
354.
by sorenjan 3y ago
That's an even bigger issue if it's a Polish plane, you don't want too many poles on the right side of the plane.
355.
by sorenjan 3y ago
I would love a video series from him where he makes a text2img diffusion model. I found the fast.ai course a bit unfocused and annoying.
356.
by sorenjan 3y ago
I just read the article, nice write up! I think it would benefit from a short explanation of what the magnitude vector (m) and the directional matrix (V) are, I'm not familiar with that kind of decomposition. Not related to the article
357.
by sorenjan 3y ago
Speaking of LoRA, what happened with ZipLoRA? It's supposed to be a better way of merging multiple LoRAs, and the results look good in their examples. Is it being used anywhere? https://ziplora.github.io/
358.
by sorenjan 3y ago
Sometimes I forget how fast modern computers are. PCIe v4 x16 has a transfer speed of 31.5 GB/s, so theoretically it should take less than 100 ms to transfer stage B and A. Maybe it's not so bad after all, it will be interesting t
359.
by sorenjan 3y ago
The large C model uses 3.6 billion parameters which is 6.7 GiB if each parameter is 16 bits.
360.
by sorenjan 3y ago
Thanks. I guess this means that fewer people will be able to use it on their own computer, but the improved efficiency makes it cheaper to run on servers with enough VRAM. Maybe running stage C first, unloading it from VRAM, and then do B a

More ›