alyxya

9 ms·

121.
by alyxya 1y ago
I made my own single controller PyTorch extension [1], though mines doesn't yet support cross node communication. I found it interesting to compare how Monarch makes things performant. I believe Monarch also uses cloudpickle for code t
122.
by alyxya 1y ago
I don't know if or how you've been using AI for coding, but Claude Code is almost certainly being used to develop itself, and I have a hard time imagining their devs not having the best setup possible to make it as easy as possibl
123.
by alyxya 1y ago
I tried understanding the gist of the paper, and I’m not really convinced there’s anything meaningful here. It just looks like a variation of the transformer architecture inspired by biology, but no real innovation or demonstrated results.
124.
by alyxya 1y ago
I think Claude Code itself is mostly vibe coded if you consider that impressive. In general I’m not aware of any vibe coded projects with a substantial amount of iteration put into it rather than being close to one shot, which is why I shar
125.
by alyxya 1y ago
I created a pytorch extension using AI generated code. It wasn’t really created solely off of vibes though as I put significant effort into designing the architecture and all the features, but I had Claude Code write all of the code. https
126.
by alyxya 1y ago
The most important advice is at the end. > Undergrads tend to have tunnel vision about their classes. They want to get good grades, etc. The crucial fact to realize is that noone will care about your grades, unless they are bad. For exam
127.
by alyxya 1y ago
It doesn’t apply to empirical science because there’s a lot more variation in observations. The variation of ideas in ML model architecture is limited by being theoretical.
128.
by alyxya 1y ago
The notion of inventing or creating something in ML doesn't seem very important as many people can independently come up with the same idea. Conversely, you can create novel results just by reviewing old literature and demonstrating it
129.
by alyxya 1y ago
Noting down here that the installation issue should be fixed now.
130.
by alyxya 1y ago
I think there may be a bug that breaks the installation of mycelya-torch if you're not on the latest version of pytorch. I'm working on fixing this, but mentioning it here for now in case anyone runs into this issue.
131.
Show HN: Run PyTorch locally with a remote GPU backend(github.com)
5 points by alyxya 1y ago|3 comments
132.
by alyxya 1y ago
The thing about this that’s interesting to me is that it can be used as a foundation for products they or other people make that combine real time RL rewards and fine tuning to improve the model. I see a lot of potential here compared to th
133.
by alyxya 1y ago
The sentiment of the comments here seems rather pessimistic. A perspective that balances both sides might be that the rate of mass adoption of some technology often lags behind the frontier capabilities, so I wouldn’t expect AI to take over
134.
by alyxya 1y ago
Then this is your new AGI benchmark to beat.
135.
by alyxya 1y ago
It's on a shirt.
136.
by alyxya 1y ago
Here's how I solved it: Let the square be a unit square with vertices at (0, 0), (1, 0), (1, 1), (0, 1), and let the center of the small circle be (x, y) and the radius be r. Each of the 3 circular arcs can be used to form an equation
137.
by alyxya 1y ago
I watched it for an hour, and at some point the black ball crossed the boundary onto the black side, so eventually the whole circle became black.
138.
by alyxya 1y ago
I haven’t tried marimo, so I’m not sure how it currently works, but I think instead of disabling autorun for slow and expensive computations, it sounds like it would be nicer if there were heuristics or benchmarks to automatically determine
139.
by alyxya 1y ago
It’s worth pointing out that the statement is about how much of Claude Code is written with it and not how much of the codebase of the whole company. In the more critical parts of the codebase where bugs can cause bigger problems, I expect
140.
by alyxya 1y ago
I skimmed over a couple of the papers referenced to get an idea of what optimizations LMCache is doing. * KV cache compression - compressing the bytes of the KV cache, taking advantage of patterns in the KV cache and with dynamic levels of