Anyway, it seems that qdrant is more a for-profit organization. So maybe that was the one of the criteria that was taken into account to exclude it?
My understanding might be wrong/incomplete, please let me know if that's not the case.
So don't start looking for reasons why any organization wasn't accepted; there probably isn't one.
I am currently working on a new website. The old stack is Vite, Svelte and Windi CSS (discontinued unfortunately). So this time maybe Astro + Solid + Tailwind.
And I am also trying to rewrite the whole Rust backend if possible, so there is quite some work to be done. What I want to change most is to make the dsp algorithm of each node clear and easy to understand and contribute to. And I also hope that the entire rust project can have complete bench and test, as well as ci, and get rid of the proc macro.
Generally speaking, what I actually care about is how to compose music, and the new possibilities that live coding brings to improvisation and composition. There is also network cooperation, real-time or non-real-time cooperation, and cooperation with AI. What possibilities can these bring?
Let me know on GH or Discord if you are interested.
It's a good chance to try Rust, WASM, DSP, etc.
> I am currently working on a new website. The old stack is Vite, Svelte and Windi CSS (discontinued unfortunately). So this time maybe Astro + Solid + Tailwind.
It's a bit amusing to read that Vite+Svelte can be considered an "old" stack. :)
I'm curious: what makes you want to move to Astro+Solid?
> Implement a dimension reduction algorithm in Rust and compile to WASM and integrate the WASM code with Qdrant Web UI.
Easy, just use a Rust crate to fit a PCA (https://crates.io/crates/pca), then at runtime do a matmul between the fitted matrix and the embeddings to get it reduced. :P
Speaking of which, there's a surprising spike in downloads for that crate on the date this blog post was made.
It's not as simple in practice, and even popular dimensionality reduction techniques like UMAP require you to reference the original dataset which is infeasible for large datasets. The hacky approach that would be good for production use (maybe not "just want to visualize 2D embeddings because they look cool") would be to train a small Parametric UMAP model (with likely a non-Rust implementation: https://umap-learn.readthedocs.io/en/latest/parametric_umap....), then convert the trained model to ONNX for WASM.
I could be wrong, because it’s not really a field I do anything in but based on what you said it’s sounds about right for a first “real project” task an intern would do at a big tech summer internship (this is as someone who has helped summer interns over the years, but never did one myself so it’s based on seeing what tasks are given to someone in their first ever internship vs people who have interned previously)
I'd be interested if anybody has more insights into this.
Nothing more interesting than copy-paste I'm afraid
It has been a part of most major AI open source projects lately, for some reason: https://sigmoid.social/@minimaxir/110951886465291229
It’s just injecting a little bit of life and visual variance into the text.
They would develop for Google because Google would give an additional value to their CV.
[edit: s/Qurans/qdrant, sigh autocarrot]
Are they even compliant with the law with this post?
Additionally, the product that this Summer of Code is for is open-source so it's a win-win for everyone.
If so, that might explain the lack of invitation.
This is a resounding instance of "tell me you don't know the domain without telling me you don't know the domain" and I think you'll find them interesting if you look into it.
