robko

13 ms·

1.
by robko 2y ago
> Are claudisms different from gptisms? Sorry, no idea. I rarely use Claude. > Why can't these authors tell ChatGPT to write with a different prose and avoid "delve", "crucial", etc? The authors could do this,
2.
by robko 2y ago
I agree. This article is clearly written with a generative language model. A few other telltale signs: 1. Repeating the same thing multiple times with slight variation: * "allowing developers to fine-tune their applications and unlock
3.
by robko 2y ago
> Loras are just as powerful as a finetuned model and you can train one in minutes even on consumer hardware. Do you have some more details on training a LoRA in minutes? Last I tried, it took several hours on an RTX 3090, but I am sure
4.
by robko 8y ago
In my experience, the canvas api is very slow and not well thought-out. For example, to create a native image object from raw pixels, you have to copy the pixels into an ImageData object, draw it to a canvas, create a data URL from the canv
5.
by robko 8y ago
I get 60ms in C. But in your code, the compiler might decide to remove most of the code since b is not used after being calculated. I checked the assembly code and it does not seem to be the case here, but it's still something to be aw
6.
by robko 8y ago
You can use both at once. Usually, you'd have something like 64x64 tiles in cache and use 4x4 or 8x8 tiles for SIMD.
7.
by robko 8y ago
I tried a few micro-optimizations, but they did not make a measurable difference, so I kept the code short instead. But maybe some JIT is particularly bad at loop hoisting, so it might make a difference there.
8.
by robko 8y ago
You are correct. The code is using an inefficient cache access pattern, so most of the time is spent waiting. You probably won't get 100x faster without SIMD, but 10x is certainly doable. Unfortunately, SIMD.js support has been removed
9.
by robko 8y ago
My guess is that the JS implementation of the worst-performing browser is having trouble with the non-1 for-loop steps. Doing 90-degree image rotation with fixed steps and some index calculations should work better (0.18 sec vs 1.5 sec for
10.
by robko 8y ago
rand() [1] is not a secure source of randomness. I strongly recommend against using that password manager since it is not secure. [1] https://github.com/oormicreations/OormiPass/blob/d1d2bf5100f...
11.
by robko 8y ago
Ah, the good old OpenGL fixed function pipeline, deprecated for over 11 years now. For something a bit more modern, I'd recommend [0], but one might argue that old OpenGL is easier to learn since you don't have to setup your own s
12.
by robko 8y ago
Interestingly, many browsers are still susceptible to this attack, for example when used in SVG files (WARNING: might crash your browser and/or operating system): https://jsfiddle.net/e3guLn08/
13.
by robko 8y ago
This method requires a video or at least multiple frames. It won't work for a single photo.
14.
by robko 8y ago
Are you already using GPU acceleration? I'm currently working on alpha matting for my master's thesis and found that it helps quite a lot.