mikeravkine

5 ms·

1.
by mikeravkine 2y ago
The Internet is 90% mobile users that can't hover.
2.
by mikeravkine 3y ago
Thank you for your efforts on behalf of the GPU poor! It's getting tougher to use older, cheaper GPUs (Pascal/Maxwell) with modern quantization schemes so anything you can do to keep kernels compatible with SM52 and SM61 would be
3.
by mikeravkine 3y ago
Ubuntu 18 went EOL on May 31, 2023 so about 8 months ago. What CentOS are you running that's still in support and was affected by this change? I had to upgrade all my systems last year after the EOL as most python packages dropped 3.
4.
by mikeravkine 3y ago
The challenge to Mastodon isn't adoption, its absense is a symptom of a deeper issue: it's a Very Bad Idea to trust random server operators with both your data and it's sovereignty. You can take your data to another server on
5.
by mikeravkine 3y ago
I just wanted to say thank you, aider (with the new unified diff format) is the first AI tool that has actually changed the way I work.
6.
by mikeravkine 3y ago
The razor doesn't apply to the police.
7.
by mikeravkine 3y ago
Hetzner can and will terminate your account and delete all your data on a whim, never telling you why. There is nothing at all you can do to prove your usage was legitimate. I get that they must deal with a lot of abuse but I wouldn't
8.
by mikeravkine 3y ago
In terms of economics specifically, in what way have things "gotten better"? Purchasing power has collapsed, the poor sentiment here seems the correct sentiment to me.
9.
by mikeravkine 3y ago
There is no difference in download time for 20mb vs 50mb in a datacenter, but the amount of time lost to alpine being "quirky" is incalculable.
10.
by mikeravkine 3y ago
If anyone is looking for a more reasonably cost effective solution, Hetzner has 16 vCPU/32GB RAM ARM VMs for $24E/mo that will run 34b Q4 GGUF at around 4 tok/sec. It's not very fast, but it is very cheap.
11.
by mikeravkine 3y ago
Where do I get one of these tech writing destroying AIs that will both understand how to use my product and produce clear, concise guides for doing so without burdening the existing team?
12.
by mikeravkine 3y ago
This is not about intelligence, it's about agency. Text completion generators don't have agency no matter how big they get.
13.
by mikeravkine 3y ago
Hetzner offers incredibly cheap ARM machines in the Falkenstein DC, for 25Eur a month you can snag the top of the line with 16 vCPU and 32GB RAM. If your usecase fits inside that 32GB (no 70B models, sadly) the price to performance of a GGU
14.
by mikeravkine 3y ago
Pascals := is concise and intuitively obvious, so obviously it never caught on.
15.
by mikeravkine 3y ago
1997 era internet would have scared the shit out of you.
16.
by mikeravkine 3y ago
Why would you need to compete with such an entity? What does "compete" even mean here? Can't instead everyone get one (or 10) of these to.. assist them? I don't understand why we are jumping directly into dystopian sc
17.
by mikeravkine 3y ago
Receiving a PR is a massive success in my books, it means someone not only found it useful but also improved it! The win-win of open source.
18.
by mikeravkine 3y ago
With VS Code remote SSH, there is no "local" you are always on the server so there is also no syncing. They do some tricks to make this seamless and perform and feel as if everything was local.
19.
by mikeravkine 3y ago
I always write the fun bits of code myself, it's usually easier then explaining to the model what I want anyway.. but like 80% of any project is scaffolding or plumbing, The Machine can do that for me with 10x speed and precision witho
20.
by mikeravkine 3y ago
..isn't this just Docker?
21.
by mikeravkine 3y ago
It's now called Xitter (with the X pronounced 'sh')
22.
by mikeravkine 3y ago
Canadian here, just checked and still no access. It's like some kind of bad joke is being played on us, Google can pound sand.
23.
by mikeravkine 3y ago
Hetzner has some weird and very strong opinions on what you can and cannot do on their servers, likely because they are so cheap they attract all sorts of shady customers. I've had good luck with OneProvider.
24.
by mikeravkine 3y ago
OobaBooga supports this kind of load-and-go LORA: https://github.com/oobabooga/text-generation-webui
25.
by mikeravkine 3y ago
I've been working on just such a tool [1] to help me digest podcasts and senate hearings. [1] https://github.com/the-crypt-keeper/tldw
26.
by mikeravkine 3y ago
Mismatched parenthesis, specifically using )) instead of ) is a common failure mode I observed in all of my codellama testing.
27.
by mikeravkine 3y ago
One caveat here is that whisper.cpp does not offer any CUDA support at all, acceleration is only available for Apple Silicon. If you have Nvidia hardware the ctranslate2 based faster-whisper is very very fast: https://github.com&
28.
by mikeravkine 3y ago
I have several sets of quant comparisons posted on my HF spaces, the caveat is my prompts are all "English to code": https://huggingface.co/spaces/mike-ravkine/can-ai-code-compa... The dropdown at the t
29.
by mikeravkine 3y ago
The model card also has prompt formats for context aware document Q/A and multi-CoT, using those correctly improves performance at such tasks significantly.
30.
by mikeravkine 3y ago
I have noticed that on StarLink some sites behind CF go into "prove you are human" loops that are impassable. What causes such loops? Just a challenge over and over.

More ›