unixpickle

18 ms·

1.
by unixpickle 7mo ago
I think we are missing the bigger point here. Licensing only matters on things that take real effort or money to produce. Who will care about licenses on software when software is free and infinite? It would be like licensing each ounce of
2.
by unixpickle 7mo ago
The `try_roots` example here is actually a _counterexample_ to the author's main argument. They explicitly ignore the "negative discriminant" case. What happens if we consider it? If we take their "parse" approach,
3.
I tried zoo.dev, and found it to be frustrating(twitter.com)
1 points by unixpickle 8mo ago|0 comments
4.
An AI agent for unsubscribing from spam(blog.aqnichol.com)
2 points by unixpickle 1y ago|0 comments
5.
by unixpickle 2y ago
To optimize for fast nearest neighbors, I chose 256 dims. Notably, this actually hurt some of the pre-training classification losses pretty severely compared to 2k dims, so it definitely has a quality cost. The site uses cosine distance. Th
6.
by unixpickle 2y ago
The "shop for random products" direction was actually fun for me too. Reminds me of amazon.com/stream a bit.
7.
by unixpickle 2y ago
Probably the same complaint as https://news.ycombinator.com/item?id=43375415
8.
by unixpickle 2y ago
You definitely highlighted a shortcoming of the feature vector model in this case. Indeed it's quite a small model trained on a single Mac for about a week, so it's not very "smart". I'd expect that this is a proble
9.
by unixpickle 2y ago
I think it would be a useful feature. For the sake of being a fun project, I didn't use CLIP because I only wanted to use models that I trained myself on a single Mac. However, to make this more useful, text search would be quite helpf
10.
by unixpickle 2y ago
Yup, it's a small model I trained on my Mac mini! The model itself just classifies product attributes like keywords, price, retailer, etc. The features it learns are then used as embeddings
11.
by unixpickle 2y ago
Ideally pose and lighting wouldn't matter as much as it currently does. I think using a better model to produce feature vectors could achieve this, or perhaps even finetuning the feature model to match human preferences.
12.
Show HN: Fashion Shopping with Nearest Neighbors(vibewall.shop)
76 points by unixpickle 2y ago|22 comments
13.
by unixpickle 2y ago
This should just be called "why VPNs are useful", i think?
14.
by unixpickle 2y ago
This seems to be pretty much exactly a standard Bayesian deep learning approach, albeit with a heavily engineered architecture.
15.
by unixpickle 2y ago
Wait what? Who actually calls trainable params "hyperparameters"? Nobody at OpenAI does, as far as I know.
16.
by unixpickle 2y ago
> I asked Nicole Avena, a professor of neuroscience at Mount Sinai who studies sugar addiction, if she believed it could be possible for food companies to engineer, intentionally or not, compounds that would make GLP-1 drugs less effecti
17.
by unixpickle 2y ago
These lines from the diff linked above are the fix: - /\* Log error and exit. \*/ - sigdie("Timeout before authentication for %s port %d", - ssh_remote_ipaddr(the_active_state), - ssh_remote_por
18.
by unixpickle 4y ago
Before one epoch, both the train and eval curves look pretty much identical. Quite curious
19.
by unixpickle 4y ago
OP here. Interesting, I had not researched the used car market enough to know about the price inflation. Can you think of any other sites that have more realistic prices?
20.
by unixpickle 4y ago
These are good ideas! It should be possible to test how photographic quality correlates with price, though one complication is that I train with data augmentation (a commonly used technique to make better use of the training data). Some of
21.
by unixpickle 4y ago
OP here. Yeah, it's quite odd that the median price prediction doesn't fall into the most likely bucket, even when that bucket has more than a 50% confidence. I'm not sure how to interpret this. One theory: maybe the median p
22.
by unixpickle 4y ago
Before writing this post, I asked chatgpt for examples of positive transfer from auxiliary losses in the literature. It pointed me to this paper: https://arxiv.org/abs/1705.07115
23.
by unixpickle 4y ago
It's definitely not just a regularizer in my case, because the gap appears even before a single epoch. The gap does also appear for two very different model architectures. One explanation is that price labels are super noisy. If there
24.
by unixpickle 4y ago
Apparently my router's built-in DNS server doesn't like the domain name `ai.` or `ai`. Switching to Google's fixes the issue for me.
25.
by unixpickle 4y ago
> We can inject targeted taps continuously with a standard deviation of as low as 14.6 x 19.2 pixels from the target area, a delay of less than 0.5s and a distance of up to 40mm Seems like 40mm isn't a very far distance. I understan
26.
by unixpickle 4y ago
I also wrote a similar thing years ago... https://github.com/unixpickle/ffmpego
27.
by unixpickle 5y ago
I think you are conflating "Transformers" and "autoregressive models". Transformers are a general purpose architecture for transforming sequences into other sequences with self-attention. AR models / GANs are framew
28.
by unixpickle 6y ago
Author here! This project includes a boatload of 3D models I've created in code and 3D printed over the past half year or so. It also includes a more recent ray tracer, which can produce some cool pictures like [1]. I've finally s
29.
Show HN: Create and Render 3D models in Go(github.com)
6 points by unixpickle 6y ago|1 comments
30.
by unixpickle 7y ago
A neural network would have some nice properties for sure, although it'd also have plenty of bad ones. For one thing, I'd guess that the NN wouldn't be great for rendering in real-time, since it's liable to be orders of

More ›