buildbot

6 ms·

1.
by buildbot 6d ago
Plenty of experience ;)
2.
by buildbot 6d ago
I’ve observed exactly these patterns with Opus and Fable as well - for example, forgetting that they can edit files and instead use python scripts as a patching tool…
3.
Co-Designing AI Models Using Speculative Decoding for Faster LLM Inference(developer.nvidia.com)
2 points by buildbot 14d ago|0 comments
4.
by buildbot 26d ago
Downside of the MTIA chips is no open backend - so quite a lift to get them working if possible at all :(
5.
by buildbot 1mo ago
Most people do not support the destruction of books - it seems like the issue in both cases is copyright forcing the destruction and non-release of the scan.
6.
by buildbot 1mo ago
Very much my own opinion, but this continues comma’s extremely cavalier attitude towards safety. PCIe connectors and GPUs are not vibration rated. What happens when you hit a bad pothole?
7.
by buildbot 1mo ago
Unified memory DGX Spark and RTX Spark laptops are already a thing :)
8.
by buildbot 1mo ago
Nemotron 3 also introduced LatentMoE, which was adopted by Kimi K3 :)
9.
by buildbot 1mo ago
The experience I was referencing was actually Microsoft!
10.
by buildbot 1mo ago
Counterpoint to many responses here - when I have interviewed people, it’s a really positive signal if the candidate says they have already seen a problem- the we can move onto something more interesting! Not solving does not mean failure
11.
by buildbot 1mo ago
MLOps though typically wouldn't been going quantization? It requires some careful testing of accuracy and performance even today. Applied ML Research Engineer or something maybe, not that I have ever seen that title. Maybe just catchal
12.
by buildbot 1mo ago
"The internet" is currently fine, besides Norwegian government services. Also something designed to withstand something does not imply it survives other somethings.
13.
by buildbot 2mo ago
Capture One in my opinion has much higher output quality, better defaults, and specifically for supports Phase One backs better than anything else. I have not done a scientific study or anything comparing the outputs (maybe I should!), but
14.
by buildbot 2mo ago
It’s absolutely wild how Jimothy has taken over the internet. This is easily the most media attention anyone/anything from Ballard has ever gotten. Also; Free Ballard! End the century of Seattle tyranny! https://www.seattle.
15.
by buildbot 2mo ago
Honestly a near millennia long expedition would be very cool, and doesn’t seem too long on the scale of space stuff.
16.
by buildbot 2mo ago
Thanks for the link. No need to be so aggressive. The blog with that detail was not live before; and they removed that language from the original link in this post.
17.
by buildbot 2mo ago
And they have since removed that language…
18.
by buildbot 2mo ago
Amazing to see an open source model already nearing the benchmarks of Fable and GPT 5.6 Sol! Also very cool to see LatentMoE being picked up by more models ( https://arxiv.org/abs/2601.18089 )
19.
by buildbot 2mo ago
Not particularly related, but “whispering earring” made me think of the hemalurgic earring in Brandon Sanderson's later Mistborn books.
20.
by buildbot 2mo ago
It was very fun to attend UBC then immediately start recognizing bits of the campus in many TV shows - Man in High Castle, Stargate, Battlestar, and many others. They have long since fixed it but there used to be an absolutely awful/h
21.
by buildbot 3mo ago
CMD-A then export all X unmodified originals works fine
22.
by buildbot 3mo ago
Things affect LLMs besides tokens, like ECC errors or cosmic rays? …
23.
by buildbot 3mo ago
I worked on a tiny, tiny part of the research for the insertable B layer in ATLAS during long shutdown one as an undergraduate - the 3 year period allowed all the detectors to be upgraded significantly! It will be interesting to see how the
24.
by buildbot 3mo ago
Err holy shit that’s quite the array and FPGA processing backend. I assume the purpose of the system is secret? (VCU128s are relatively large, HBM equipped FPGAs for people not aware)
25.
by buildbot 3mo ago
TIL that PNW Western Red Cedars are not in fact, Cedar trees!
26.
by buildbot 3mo ago
It honestly explains so many issues I have been having, as I used it primarily for ML research (on my personal account, doing things not related to my job I should note). It would literally typo package names and spend huge amounts of time
27.
by buildbot 3mo ago
Or build a park on top of the datacenter as a living roof? Maybe even use the waste heat to help grow things in cold, dark climates? It’s pretty different; but locally they covered a good chunk of a freeway with a very nice park to mollify
28.
by buildbot 3mo ago
The 120B and 20B GPT-OSS models by OpenAI did this last year for what it’s worth; the MoEs where MXFP4
29.
by buildbot 4mo ago
Why is this so good? Probably because someone still cared a lot about the bit! And wow this is really quite good lol.
30.
by buildbot 4mo ago
So the AI agent had privileged access to remove 2FA, ignore the account email, and just hands accounts to whoever asked? Honestly that’s so highly negligent I wonder if the implementation team for that “feature” was intentionally trying to

More ›