• 0 Posts
  • 16 Comments
Joined 1 year ago
cake
Cake day: June 12th, 2023

help-circle


  • Yeah, the ISP cartels sucks. I’ve been stuck paying $170/mo for uncapped 1000/35mbps connection.

    Thankfully, before the end of the year, a local ISP is moving into my area. They offer uncapped symmetrical gigabit, for $75/mo… I’ll be saving $95/mo for BETTER service.

    The longstanding ISP cartels should seriously be punished for the abuse of their market positions and failure to appropriately use government funding they’ve been given.


  • LTT, for better or worse, is a very popular tech/PC gaming entertainment YouTube channel. A lot of the younger folks in the nerd space probably grew up watching his content.

    I started watching a little late, because I thought he seemed kind of annoying back in the day, but I actually do enjoy their content. It’s clear they have some organization and communication problems internally that they need to sort out, though.



  • I like video games, goofy stuff, and tech. So lately, it’s been:

    • My Brother, My Brother, and Me
    • Noclip
    • Besties
    • We Have Concerns
    • Search Engine (New podcast from PJ Vogt of Reply All fame - if you haven’t listened to Reply All, there are a number of great episodes to enjoy)
    • How Did This Get Made
    • and occasionally Swindled (my girlfriend likes true crime stuff and this is about as close as I like to get to that)




  • Sampling is just how many passes Stable Diffusion makes to generate the image. It makes noise based on the seed and then diffuses it with the prior image in “latent space” as many times as you select.

    The first few samples are going to be pretty abstract, but with each pass it can add more detail. At a certain point, additional passes aren’t likely to get much more detail, though, I think primarily due to the limitation of the resolution (512 for 1.5, for example). Instead, the new noise basically pushes the image in a slightly altered direction.

    There’s no right answer to how many steps you should use. It’s going to vary between models, seeds, and your prompt. I find 20-40 works best in Stable Diffusion 1.5 based models.

    If you generate an image that looks great other than some minor oddities, try adjusting the steps up or down a bit and see if it resolves it. If not, use a seed variation (the “Extra” checkbox next to the seed number in AUTO1111) to keep the original seed, but introduce slight alterations.




  • The $160 is a lifetime pass… I pay $20/mo for Netflix. That’s $240/year. So, if you think it’s worth it for even one year, compared to something like Netflix, then it’s a pretty solid value proposition.

    I bought the lifetime pass in 2014 when it was $75. Been more than happy with that decision.



  • Honestly, it seems what Reddit has been doing lately is far worse than anything Digg ever did. 🤷‍♂️ Digg basically just made a bunch of bad design decisions primarily but not entirely to please investors, and the majority of changes were not welcomed by the community.

    Reddit is straight up disrespecting their userbase and telling them to suck a fat one and deal with it, holding to their miserable decisions while openly suggesting their userbase is too weak to go anywhere else… And also doing so primarily to please misguided investors.


  • Light roasted coffee has more caffeine than dark roasted coffee.

    Technically, per bean, more of the caffeine is cooked out of the dark roast. However, other things are also roasted out of a dark roast to the point that the individual beans are also lighter and smaller. When brewing coffee, usually you either weigh your dose of beans out, or you use a scoop for some consistency. Either method will result in more dark roast beans ultimately making it into the brew than would with a (larger, heavier) light roast.

    Typically, this more than cancels out the reduced caffeine content per bean, so a brew of dark roast coffee still typically has more caffeine in it.