Yenrabbit

9 ms·

61.
by Yenrabbit 3y ago
It's slow, palm-sized and worse than a normal RC car in pretty much every way. This is not the mobile gunning platform you're worried about
62.
by Yenrabbit 3y ago
Incredible work by the author stepping through all the nitty-gritty details and showing how easy it is to miss something subtle that could degrade performance.
63.
by Yenrabbit 3y ago
Ooh I also use connections as a benchmark! It tends to favour things with 'chain of thought' style reasoning in the training mix somewhere since directly producing the answer is hard. Do you have public code you could share?
64.
by Yenrabbit 3y ago
We've got a technical post with all the juicy details coming next week. But that bit refers to packing the 4-bit weights into a type FSDP is happy to shard (like float16 or float32) which matches the other non-quantized bits of the mod
65.
by Yenrabbit 3y ago
Note that they shared part two recently: https://samsja.github.io/blogs/rig/part_2/ For those talking about breakeven points and cheap cloud compute, you need to factor in the mental difference it makes runni
66.
by Yenrabbit 3y ago
Neat! How many images are in the dataset out of curiosity?
67.
by Yenrabbit 3y ago
One benefit answer.ai has over some of those labs is that it's building software, so expenses are mostly just salaries and GPU time - no need for massive capex on fabrication equipment or anything like that. And since we're focuse
68.
by Yenrabbit 3y ago
Yup, I just used the default demo of their TTS to convert the article one chunk at a time so I could listen while I swapped some RAM around. https://platform.openai.com/docs/guides/text-to-speech
69.
by Yenrabbit 3y ago
Nice work - I get ChatGPT to write ffmpeg commands quire frequently, this is a logical next step :) One thing: you may want to clarify that "Upload" also does the processing with whatever text is currently entered. I expected to u
70.
by Yenrabbit 3y ago
I tested this out on a topic I'd been discussing with some fellow researchers, and it pulled in the papers we'd chatted about plus a bunch of related ones that look very relevant and interesting. Congrats on a cool project!
71.
by Yenrabbit 3y ago
FWIW I was able to get about 20% accuracy (perfect 4/4 groups) with ~50% of groups correct on average and most mistakes being groups with 3/4 right (so at least on the right track) with my first attempt at 0-shot prompting. The pr
72.
by Yenrabbit 3y ago
I tried with clustering similar embeddings but it did extremely poorly (~0%) since the groupings are often deceiving with words in a group only having one small way in which they're connected and lots of spurious fake groups to throw y
73.
by Yenrabbit 3y ago
3D printers are great general motion platforms! I'm a big fan of things like this, which re-purpose cheap printers in new ways. I've had fun adding a pen to mine to turn it into a plotter, or strapping on a camera with a high-magn
74.
by Yenrabbit 3y ago
This is excellent! I've been turning my family's memorable sayings into smooth jazz. Congratulations
75.
by Yenrabbit 3y ago
One reason this implementation is exciting is that the code that accompanies the Gaussian Splatting paper requires companies wanting to use the tech commercially to pay for a commercial license. There are existing viewer implementations, bu
76.
by Yenrabbit 3y ago
Garbage in garbage out still applies (I tested with some shaky phone video and the results were a mess) but when done right this looks incredibly promising for future VR, film and gaming applications.
77.
Luma labs integrates gaussian splatting for real-time interactive 3D scenes(lumalabs.ai)
2 points by Yenrabbit 3y ago|2 comments
78.
by Yenrabbit 3y ago
Gaussian Splatting is a relatively new technique, somewhere between a NeRF and traditional photogrammetry, which allows for fast rendering of 3D scenes captured from 2D input images. Luma labs are rolling out new features based on this tech
79.
by Yenrabbit 3y ago
For anyone curious to learn more about the incredible lives of seabirds I can highly recommend the book "The Seabird's Cry" by Adam Nicholson. Poignant and poetic, with each chapter covering a different species, their amazing
80.
by Yenrabbit 3y ago
Definitely. One of the time-consuming parts of rendering is sorting the gaussians by distance to the camera, which for two nearby cameras could be optimized. This also goes for adjacent frames - assuming smooth motion im pretty sure there i
81.
A new AI-focused chip claims a 9x performance improvement over Nvidia's H100(notebookcheck.net)
1 points by Yenrabbit 3y ago|0 comments
82.
by Yenrabbit 3y ago
This April Fools joke got me back in 2013, and my family have never let me forget it! I should have checked the article date... Thankfully our microwave started working again once it cooled off and all that was hurt was my pride.
83.
by Yenrabbit 3y ago
Ooh interesting, thanks for sharing!
84.
by Yenrabbit 3y ago
It does start getting worse at some point right?
85.
by Yenrabbit 3y ago
They've written a great explanation of how it works too: https://viberary.pizza/how Semantic search for the win :)
86.
Comma 3X(comma.ai)
101 points by Yenrabbit 3y ago|93 comments
87.
by Yenrabbit 3y ago
Cloud GPU providers are running low on capacity at the moment as people frantically suck up capacity to hop on the AI bandwagon, raising worries about availability. So having guaranteed access is maybe one motivation for local GPUs. But for
88.
by Yenrabbit 3y ago
I tried to test this with nanoGPT in an afternoon, since the code change is pretty minimal. It's hard to get conclusive results at that scale though - to be able to say anything with confidence you'd need to run multiple tests, fi
89.
by Yenrabbit 3y ago
Twitter thread from the author summarizing some key ideas from the article: https://twitter.com/jeremyphoward/status/1678558165712113664...
90.
by Yenrabbit 3y ago
Good to know! I've recently started getting a ton a spam calls (8-10/day) and was assuming the situation was hopeless. Will give your method a go :)

More ›