rao-v

12 ms·

241.
by rao-v 1y ago
That's awesome! Yeah it seems strange to me that we don't have a standard way of tracking how a model is built up and parameterized (basically Fusion 360's history mode). I'd naively assumed that this was a solved proble
242.
by rao-v 1y ago
Isn’t the python based build123d the current best CAD in code solution? The problem with OpenSCAD is that it cannot export solid geometry, just a final mesh. More broadly, I was genuinely shocked to realize, when I was playing with it, that
243.
by rao-v 1y ago
I’d love a pointer to this when it’s shareable!
244.
by rao-v 1y ago
Just to add a nit to this, “10x faster” is noise and fluff. Save it for when you actually have a real measured metric that someone might care about. Notice also 10x faster generic event creation is not what the value is! Rather say “Make mu
245.
by rao-v 1y ago
I’ve been playing a bit with toy robots and vision models and I can’t shake the sense that somebody has got to be close to simplifying the control game with high fps cameras. Why are there so many systems to control motors and ground positi
246.
by rao-v 1y ago
Why are we still launching models without simple working python example code (or llama.cpp support)?
247.
by rao-v 1y ago
Absolutely! I remember playing and struggling with Cygwin back in the day… I meant original in the sense of the original vision for WSL.
248.
by rao-v 1y ago
I'm with you - after years of messing with dualboot Linux, including (foolishly) running multiday Gentoo builds, WSL + Windows now gives me everything I want from Linux with zero friction. In fact, I'm a little annoyed that I ca
249.
by rao-v 2y ago
This is terrific and it’s tempting to turn into a little python package. +1 for notation to say it’s ~20,2 to mean 18~22
250.
by rao-v 2y ago
This seems like a near perfect use of coding LLMs and a useful way to implement reinforcement learning. “Add a major bug to this file that is not covered by existing tests” vs “Find the bug in this file” vs “write a sensible test in this fi
251.
by rao-v 2y ago
The Guardian feels like the last good normal newspaper at this point. Great book and movie reviews, normal detailed circa 2005 coverage, and none of the NYT’s wierd if we didn’t break the story we won’t talk about it.
252.
by rao-v 2y ago
Thank you for this. I’ve been baffled by how many of the usual tools are a poor fit for persisting work over SSH across multiple servers.
253.
by rao-v 2y ago
I’m interested in understanding what happened to the debt. My simplistic understanding was that a bunch of the $13B debt raised for the twitter aq was secured by Tesla stock, with some sort of callable mechanism (which got a bit scary when
254.
by rao-v 2y ago
I thought this effect would dominate too, but it turns out the paper’s assumptions is that the PBH moves at 100 km/s so we’re talking about 5 joules while it’s inside you and maybe 1000 joules total while it’s in your vicinity. Hardly
255.
Ask HN: How do I learn more about LLM-assisted transpilation (e.g. Python->C)?
2 points by rao-v 2y ago|1 comments
256.
by rao-v 2y ago
Got a link or two?
257.
by rao-v 2y ago
Is the thought that it can't be done efficiently or that GPUs are missing instructions to make it possible at all?
258.
by rao-v 2y ago
Is there a transpiler / compiler that can take basic Python / Rust / C code and run it exclusively on GPU (I'm assuming GPUs support some sort of branching these days)? It would be nice to be able to take random bits of