Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
liuliu
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
45 ms
·
631.
▲
by
liuliu
5y ago
We release carbons from dead trees accumulated over millions of years on Earth though.
632.
▲
by
liuliu
5y ago
Combining transferrability and making patent dirt-cheap to acquire means that most patents will end up with companies that won't put it in productive use (most recent example, see Lecun's account for convnet patent: https:/&
633.
▲
by
liuliu
5y ago
One thing helps me: 100ºF is around body temperature. That helps to set a mid-point between the other two temperature points like you described.
634.
▲
by
liuliu
5y ago
People expect a positive investigation result. Asking yourself, what kind of evidence would be "enough" to say it is not a lab leak? This is not obvious. Any negative result would beg deeper questions on trustworthiness of the giv
635.
▲
by
liuliu
5y ago
Sensor fusion is hard, especially with data at various qualities (different generations of LIDAR, radar, probably images are more portable actually). I am actually on sensor-fusion side, and think a transformer can merge everything and gene
636.
▲
by
liuliu
5y ago
LIDAR operates by timing how a photon reflects from a surface, it doesn't guarantee see everything. As you stated already, it cannot see in snow or rain. I am actually on sensor-fusion side, but I don't think we should jump to the
637.
▲
by
liuliu
5y ago
You get sparse point cloud from LIDAR sensors, not accurate 3D maps. This is the main reason why some people think LIDAR may not work well (mostly, only comma.ai and Tesla folks). Vision can also get you 3D maps, either in active manner (IR
638.
▲
by
liuliu
5y ago
You say it as if it is regulated. They will simply put a stop for real customers and pump some orders in the market to keep the peg. I fail to see how they can fail except massive incompetence in managing this scheme.
639.
▲
by
liuliu
5y ago
Alternative theory: USDT is actually down by 40% but the current peg is artificially supported by restricting trades (with USD).
640.
▲
by
liuliu
5y ago
Interesting in-depth discussion with the parallel parentheses matching solution. Back to the clipping problem though, in each intermediate step (when we still incorporating prefixed results), do we compute clipping for each elements progres
641.
▲
by
liuliu
5y ago
If it is $25 a token with 100k of them, these are not exactly non-fungible, isn't it :)
642.
▲
by
liuliu
5y ago
For these cases, you can just use a modern build system (Bazel) and incremental linker (lld / gold).
643.
▲
by
liuliu
5y ago
Because the people who answer the survey cannot freely express their mind, otherwise another mysterious group of people will start to kill them and their families. Everyone knows CCP controls these assassins through their seat at the High T
644.
▲
by
liuliu
5y ago
Then just do a sanction on business with mainland like U.S. did with Iran. As it is, these are very targeted moves. Taiwan to decoupling from China would only be good for them.
645.
▲
by
liuliu
5y ago
The language also had some false starts (in 1.0 / 2.0) and some interoperability baggage (with ObjC). But arguably, it is in a steady path now. If Apple continue the current path, in 3 or 4 years, it will be much more interesting. Of c
646.
▲
by
liuliu
6y ago
VNC let you choose between slow or low-quality. Low-latency (in terms of <10ms) high-fidelity (4k@60fps) video streaming is not a solved problem. It is hard engineering.
647.
▲
by
liuliu
6y ago
Both are in the same sweet spot. But both are big languages with a lot of features and requires a lot of support. Luckily, so far, they are supported by big companies (JetBrain is no Apple, but has a good track record of supporting this kin
648.
▲
by
liuliu
6y ago
Do you want to take a look at https://github.com/liuliu/s4nnc/ ? Nothing to serious though.
649.
▲
by
liuliu
6y ago
Maybe sodium? It tastes unusually salty.
650.
▲
by
liuliu
6y ago
Yeah, I believe Seam Carving paper did this. Pretty simple, find the low-energy path and do linear interpolation between the two neighboring pixels (I think it also introduced some tricks to find n low-energy paths at once to avoid insertin
651.
▲
by
liuliu
6y ago
Torch (Lua) predates TensorFlow and is Lecun's pet project for a few years at that point already. But Lua as a language is unpopular at a time. PyTorch would be a welcome addition then. But even if no PyTorch (nor Caffe2) in an alterna
652.
▲
by
liuliu
6y ago
It surely will pivot. Low-latency high-fidelity streaming doesn't come handy, and a lot of hard engineering work. But if worked, a lot of applications. Like you said, 100Gbps NIC between your data and video editor is much better than 1
653.
▲
by
liuliu
6y ago
Got some time to try this now. It worked as expected, I have vgpu_vfio. However, it doesn't perfectly fit my needs. Particularly, my host system is "heavy", I need it to run CUDA etc, while the VM just to run games. However,
654.
▲
by
liuliu
6y ago
Register for evaluation: https://nvid.nvidia.com/dashboard/#/dashboard
655.
▲
by
liuliu
6y ago
One thing I want to figure out (because I don't have a dedicated Windows gaming desktop), and the documentation on the internet seems sparse: it is my understanding that if I want to use PCIe passthrough with Windows VM, these GPUs can
656.
▲
by
liuliu
6y ago
Ha, thanks! I mostly use Apollo iOS and didn't notice this, will check it out!
657.
▲
by
liuliu
6y ago
One thing I want to call out, and seems to be the case (based on my understanding of the client code?), Apollo client doesn't try to understand the query and make intelligent use of the normalized cache based on the actual query going
658.
▲
by
liuliu
6y ago
It can be quickly validated / disproved by doing unsupervised learning on RAW images. I believe there are a few large RAW image dataset available nowadays.
659.
▲
by
liuliu
6y ago
Dflat's compiler requires Bazel to build because it is easier to reference to flatbuffers' source code with Bazel than with Swift Package Manager. The runtime itself (which you delivers with your app) can be built with either Baze
660.
▲
by
liuliu
6y ago
I may omit details in that paragraph, sorry! Autoincrement was used to prevent rowid reuse (and it is explicit primary key, where the properties you marked as primary in flatbuffers schema is just unique index), SQLite documentation claims
More ›