tarruda

25 ms·

301.
by tarruda 3y ago
What is even better is the fact that it has 200k weekly downloads. I thought this was supposed to be a troll package.
302.
by tarruda 3y ago
Would be funny if NT Times won this and all commercial LLMs were shut down. Then LLMs would be distributed only via torrents, like most copyright infringing media.
303.
by tarruda 3y ago
> then would you accept that as a sufficient remedy? Probably not until they pay him a hefty copyright fee.
304.
by tarruda 3y ago
I wouldn't touch the generated ifs, but would generate a b-tree database which maps integers to the address of the generated code which has the desired "if" comparison. Then the C program starts by querying the b-tree for the
305.
by tarruda 3y ago
> which I consider hugely positive; AI that is, Why do you think AI is hugely positive? Maybe I'm a pessimist, but I see no good outcomes from the excessive automation that will result from advances in AI. It can make the world bett
306.
by tarruda 3y ago
Have you done it yourself? Is there documentation anywhere?
307.
by tarruda 3y ago
This was a favorite of mine some months ago but it has been surpassed by other Mistral finetunes. To me the best one right now is neuralhermes 2.5, check out how I was able to teach it basic function calling in the system prompt: https:&#x
308.
by tarruda 3y ago
Why would anyone use a LLM with ads injected when there are alternatives?
309.
by tarruda 3y ago
Is the risk any different than that if a programmer who used to work for Rockstar games? Aren't former employees allowed to learn from their experience working on GTA V and develop products based on that knowledge, just as Rockstar pro
310.
by tarruda 3y ago
It could be a great learning material. I'm not a game developer, but I would be very happy to have an AAA game source code I can build locally.
311.
by tarruda 3y ago
> People that were working on this stuff for a long time were not in it for money. It seems like 90% of OpenAI staff was going to join Sam Altman at Microsoft if the didn't return. Can you elaborate on that?
312.
by tarruda 3y ago
If they can extract the game engine code into a reusable framework, yes. No game created from it could ever be legally released though.
313.
by tarruda 3y ago
That the uploader used a Saudi Arabia VPN node?
314.
by tarruda 3y ago
Can these downloads be used to build a working version of the game locally?
315.
by tarruda 3y ago
> My favorite thing about picking up a retro emulation handheld is using it to play through these old games An alternative is install something like retroarch on your smartphone and use a telescopic controller.
316.
by tarruda 3y ago
FF VI and Chrono Trigger (both SNES) are two gems which I would definitely replay if I had the time. It is easy to get lost the pretty graphics of modern games, but to me well written story and characters are the best part of JRPGs
317.
by tarruda 3y ago
I mean you can save more depending on where you live due to cost of living and income taxes. You can live very comfortably in a low cost country with a $90,000/year salary and still save half of that.
318.
by tarruda 3y ago
> It's probably good advice for maxing out your compensation, though. It is not only about how much you make, but how much you save.
319.
by tarruda 3y ago
As long as the app only depends on the standard, you have some flexibility. For example, you can outsource authentication to a third party OIDC server on SaaS deployment and run a self hosted OIDC server like Ory Hydra for onprem deployment
320.
by tarruda 3y ago
Where did you get your information from? Just 4 days ago he redid a model after the author updated the EOS token: https://huggingface.co/TheBloke/openchat-3.5-1210-GGUF/discu...
321.
by tarruda 3y ago
I use them for coding, you'd be surprised at how much you can get done with deepseek coder 6.7b Most GPT plugins for code editors will also work for local models since you can have OpenAI API stubs running locally. Clearly it is nowher
322.
by tarruda 3y ago
Just download the GGUF from huggingface. @TheBloke always uploads GGUF versions of the popular models
323.
by tarruda 3y ago
Yesterday I tried mixtral 7bx8 running on the CPU. With an Intel 11th gen chip and 64gb DDR4 at 3200mhz, I got around 2-4 tokens/second in a small context, this gets progressively slower as the context grows. You would get a much bette
324.
by tarruda 3y ago
The main framework for running LLMs was pytorch, which AFAIK has a slow CPU implementation, so only GPU/CUDA users would be covered. Llama.cpp was the project that popularized running LLMs on the CPU due to its very efficient implement
325.
by tarruda 3y ago
I'm blown away by how easy this makes having a boot drive for multiple operating systems while also letting the thumb drive work normally as a data storage medium
326.
by tarruda 3y ago
I also don't know much, but my limited understanding is that every transaction/mutation in a dApp has a cost, so this might be useful to reduce bot incentives.
327.
by tarruda 3y ago
That is a very interesting concept
328.
by tarruda 3y ago
> Website operators can use alternatives, like asking for micro-payments Similarly to how dApps work in ethereum-like blockchains?
329.
by tarruda 3y ago
This reminds me of https://youtu.be/b2F-DItXtZs
330.
by tarruda 3y ago
In contrast to this, "comptime" is one of Zig's innovations which completely removes the need for preprocessors and code generators without leaving language you're programming in.

More ›