It’s 7 billion parameters big.
oktoberpaard
joined 2 years ago
Revolution!
I used to have everything set to English (my second language), but nowadays I use Spanish when available (third language). I use my native language only for a handful of local apps and websites if Spanish is unavailable.
I find it incredibly cringe anyway that all these parties just copy the same slogan. Some weird form of international nationalism, where they all just copy whatever the others are doing. It apparently works very well.
I guess in Western Europe it’s largely focused around anti-immigration and anti-EU sentiments, but with this movement being more and more international, I do notice an uptick in rhetoric concerning sexual minorities and women’s rights, with a lot of anti science and elitism/wokeism sprinkled in. It’s very scary. I’m happy that we don’t have a political system where the winner takes it all in my country, as it’s pretty bad already as it is right now.
Some features/plugins can be quite taxing on the system and in extreme cases it can slow the editor down to the point of being unusable. I’m a happy Neovim user with a LazyVim setup, but I experience this extreme slowdown for some JSON files and I haven’t looked into it yet to see what causes it.
You can let your editor do the same compute intensive or memory hogging things that a GUI editor does. The fact that it runs in your terminal doesn’t make it lightweight by definition.
A general remark rather than an answer to your question: in general, more than not equal better with vitamins, and too much vitamin B6 can lead to nerve problems. I would only supplement when you have a good reason to believe that you’re deficient. If you insist, at least try to stick to normal doses.
I disagree, but I do agree that there are better options available than JPEG. Lossy compression is actually what allows much of the modern internet to function. 4K HDR content on Netflix wouldn’t be a thing without it. And lossy compression can be perceptually lossless for a broader range of use cases. Many film productions use high quality lossy formats in their production pipelines in order to be able to handle the vast amounts of data.
Of course it all depends on the use case. If someone shares some photos or videos with me to keep, I’d like them to send the originals, whatever format they might be in.
The Dutch brought Sinterklaas (Saint Nicholas) to New Amsterdam, current day NYC, which has developed into Santa Claus as we know it know. Santa Claus is a phonetic derivation of Sinterklaas.
Apparently he was depicted as a thick-bellied Dutch sailor with a pipe and a green outfit in Washington Irving’s History of New York book, so maybe that’s where you got that idea from.
The way I read it, they’re not asked to follow an open standard, but to provide a specification for their own standard. That means that they could dictate the pace of development and others could decide for themselves wether or not to implement it for their devices. It’s similar to WhatsApp being forced to provide third parties with API access.
Also, they currently don’t support alternative protocols with cross-platform support and they’re not going to, unless forced.
Looking at their repo, they’ve tested this with LLM models that have not been trained to generate chain of thought outputs, by varying the system prompts. It’s therefore more of a proof of concept, but I can imagine that if you train a model to do this natively it could work.
Using the same prompt with QwQ made no difference for me (the chain of thought was still very long and quite verbose), while using it with Qwen2.5 Coder made the output extremely terse and not very useful for open-ended questions.