slewis

5 ms·

1.
by slewis 8mo ago
Can you try a simpler less programmery version? "are any of my recent blog drafts unpublished and nearly ready to go?"
2.
Serverless RL: Faster, Cheaper and More Flexible RL Training(openpipe.ai)
9 points by slewis 11mo ago|3 comments
3.
by slewis 1y ago
OpenAI created a benchmark for this: https://openai.com/index/paperbench/
4.
by slewis 2y ago
It would be really useful to see these evaluated across some of the same evals that the original R1 and deepseek's distills were evaluated on.
5.
by slewis 2y ago
Hey, that's me! Happy to answer any questions about how this works if folks are interested.
6.
by slewis 2y ago
I've spent tons of time evaluating o1-preview on SWEBench-Verified. For one, I speculate OpenAI is using a very basic agent harness to get the results they've published on SWEBench. I believe there is a fair amount of headroom to
7.
by slewis 2y ago
Is it stateful? Like can I do a run, read the results, and then do another run from that point?
8.
by slewis 2y ago
This 100% matches my experience. I like to jokingly call founder mode: "fine-grained multi-level oversight". Others might call it the derogatory "micromanagement". That doesn't mean I control every decision, or that
9.
by slewis 2y ago
Creator describing how this works: https://youtu.be/PHQweR1z7pI?si=BpRlWxtxnRmYaeKG
10.
by slewis 3y ago
Overplay one’s hand: spoil one's chance of success through excessive confidence in one's position
11.
by slewis 3y ago
The memorization use case is brilliant. Put your talk track for a presentation in and say “help me memorize this by quizzing me”. Thanks!
12.
by slewis 3y ago
I call this "keep the fingers moving".
13.
by slewis 3y ago
Clicked the youtube link. An hour and 20 minutes later here I am. Thanks for sharing it that was awesome.
14.
by slewis 3y ago
I like the ambition! You can do some amount of iterative and visual data transformation in Weave now. But maybe like 10% of what you can do with Pandas. Pandas is awesome and lingua franca in data science, unseating it would be an incredibl
15.
by slewis 3y ago
Hi! I'm Shawn, founder/CTO at Weights & Biases. We've been working on Weave for a couple years now, and it powers core parts of wandb.ai. It's a UI toolkit built for programmers that can be reprogrammed from the UI i
16.
by slewis 3y ago
How dare you! This is a serious place. ^
17.
by slewis 4y ago
Wow! What a joy to click on the comments and find a positive comment at the top. Thanks for writing this and thanks HN upvoters for expressing your gratitude.
18.
by slewis 4y ago
It’s also surprising at first that infinite series can add up to a specific, finite number. If I go halfway to X, halfway again, and so on, where do I end up? All next realities are weighted based on their probability given the current one.
19.
by slewis 4y ago
Finally! All you ungrateful users who expect your search engines to pore through billions of pages in milliseconds without complaining are going to have to show some humility for once.
20.
by slewis 4y ago
The response here makes me think most commenters don’t have experience with this particular footgun. To clarify: Python can gc your task before it starts, or during its execution if you don’t hold a ref to it yourself. I can’t think of any
21.
by slewis 4y ago
For one thing: if you’re going to make a post like this, put your contact info in your profile!
22.
by slewis 4y ago
nbdev2, which this article is about, is a solution to this problem. It makes notebooks testable, composable, versionable, and more.
23.
by slewis 4y ago
“Everybody already knows that code running on a computer cannot become sentient,” said the temporarily coherent chemical blob.
24.
by slewis 4y ago
Very cool! This might be a naive question. Are your arithmetic patterns equivalent to dependent types, like those found in Idris?
25.
by slewis 4y ago
Yes, I can certainly see how this is confusing. We will work with our legal team to clarify the terms.
26.
by slewis 4y ago
Thank you for responding! I am certainly not a lawyer, however, 3b and 3c (from the terms link you posted) state that user content, including specifically Models, are property of the user. Are you saying you think there is a conflict betwee
27.
by slewis 4y ago
Oh, great! HN as bug resolution mechanism++.
28.
by slewis 4y ago
Founder of Weights & Biases here (wandb). There are a couple issues raised in this thread: api key shouldn’t be required to download a public model, cache in home directory is annoying for this case. We will fix them.
29.
by slewis 4y ago
Founder of Weights & Biases here (wandb). We don’t forbid anything, models are property of the people who created them. Why do you think that? EDIT: I'll edit respond, since you did. Look at sections 3b and 3c in the terms, they co
30.
by slewis 4y ago
Switching to the "latest tweets" view instead of the algorithmic "home" view helps.

More ›