zcbenz

6 ms·

1.
by zcbenz 8mo ago
It is a bug in MLX that has been fixed a few days ago: https://github.com/ml-explore/mlx/pull/3083
2.
by zcbenz 1y ago
In the absence of hardware unified memory, CUDA will automatically copy data between CPU/GPU when there are page faults.
3.
by zcbenz 1y ago
Most features are already supported on x86 CPUs, you can pip install mlx on Linux , and you can even use it on Windows (no official binary release yet but it is building and tests are passing).
4.
by zcbenz 1y ago
For learning internals of ML frameworks I recommend reading the source code of MLX: https://github.com/ml-explore/mlx . It is a modern and clean codebase without legacies, and I could understand most things without see
5.
by zcbenz 2y ago
You can use any serialization that supports numbers, in the repo I used BSER (a binary format of JSON), you can even use raw JSON but it would be very inefficient for float numbers though.
6.
Show HN: You need a few lines of JS, not a vector database(github.com)
5 points by zcbenz 2y ago|3 comments
7.
Show HN: Chrohime – cross-platform C library wrapping Chromium's native UI code(github.com)
2 points by zcbenz 2y ago|0 comments
8.
by zcbenz 2y ago
No it is not normal, I only tested x64/arm64 macs, I will try on linux.
9.
by zcbenz 2y ago
That is sad, the name sisi comes from the sisi empress: https://en.m.wikipedia.org/wiki/Empress_Elisabeth_of_Austria
10.
Show HN: Sisi – Semantic Image Search CLI tool, locally without third party APIs(github.com)
128 points by zcbenz 2y ago|41 comments
11.
V8 is damn fast, faster than C++ when it comes to numerics(github.com)
1 points by zcbenz 2y ago|1 comments
12.
V8 is damn fast, faster than C++ when it comes to numerics(twitter.com)
8 points by zcbenz 2y ago|1 comments
13.
Show HN: Train a language model from scratch with JavaScript(github.com)
1 points by zcbenz 2y ago|0 comments
14.
by zcbenz 2y ago
A few months ago I ported ~15k lines of python code (10k are tests) to typescript, using GPT4. It cost me ~$70. The python project is https://github.com/ml-explore/mlx and the converted project is https://gi
15.
Show HN: node-mlx – Machine learning framework for Node.js(github.com)
2 points by zcbenz 2y ago|0 comments
16.
JavaScript's garbage collector is not very ready for machine learning(github.com)
9 points by zcbenz 2y ago|2 comments
17.
Show HN: Inference Llama2 with High-Level C++(github.com)
1 points by zcbenz 3y ago|0 comments
18.
by zcbenz 3y ago
I wrote a Node.js module that implemented similar feature years ago. https://github.com/atom-archive/node-runas (An updated fork can be found at https://www.npmjs.com/package/runas-redux )
19.
by zcbenz 3y ago
Shameless to insert my app here, also open source and provides quick access from menu bar. https://chie.app/posts/minimalist-approach-to-using-chatgpt/
20.
by zcbenz 3y ago
I would argue it is not a large part a webview. The webview related code is 900 lines vs the whole project 10k lines.
21.
by zcbenz 3y ago
I actually asked ChatGPT what chie might mean in other languages, for multiple times, but it did not tell me this :(
22.
Show HN: Chie – a cross-platform, native, and extensible desktop client for LLMs(chie.app)
67 points by zcbenz 3y ago|23 comments
23.
by zcbenz 3y ago
Suppose one is writing a desktop app doing searches of embeddings, is there a lightweight vector database that runs on all major desktop platforms? I'm aware of redis and sqlite-vss, but none runs natively on Windows.
24.
by zcbenz 3y ago
Adding fields to context will break ABI compatibility. To work around the problem, some libraries add padding fields, and some require passing the size of struct to determine ABI version.
25.
by zcbenz 3y ago
Will you provide APIs for this service? I'm writing an extensive desktop app for all ChatGPT alike services, and this service is definitely a thing I want to include.
26.
by zcbenz 3y ago
Since several years ago I was starting to have health issues that if I sat still for too long my eyes/arms/legs/waist would feel pain, sometimes fiercely. So I set a timer to remind me to stop staring at the monitor and stand
27.
by zcbenz 4y ago
When I was teen I lent some high value items in online game, and of course they never got back. I’m still thankful that I learnt the lesson online instead of in real life.
28.
by zcbenz 4y ago
I have no relationship with Apple, but from my experience, the usual way big corp contributes to open source is to have their engineers sending patches, instead of sponsoring money/hardware directly. For example Apple sent patches to s
29.
by zcbenz 4y ago
From my memory GNOME 3 and Unity were created because there was a trend to make all interfaces both desktop and mobile friendly, Ubuntu tried to ship a phone using Unity, and GNOME 3 was advertised for its tablet friendly interface. For Win
30.
China uses Covid health code to restrict movement of financial scandal victims(scmp.com)
4 points by zcbenz 4y ago|0 comments

More ›