I couldn't get it working with lmstudio. I have bonsai-8B running with llama-cpp and am attempting to build a harness for it. Looking good so far, I just got it started but Forge made tool calling work pretty quickly!
Very impressive. Unsure what is going on with the tool icons, they're a bit strangely named. "Bessel", "pour corner" (twice), "From the cross section".
Looks like some work needs to be done to get this into familiar CAD terms with the English translation.
Difficult to use right now, it just switched into the Chinese translation when trying to make a cube and there doesn't appear to be an easy way to change that.
I always see articles about the decreasing cost of solar, but where are these costs collected from? Is it just not available at a consumer level? Maybe I missed the sources in the article somewhat...
I think this just illustrates why hidden state is so difficult for the user. It's hard to replicate old calculators because it was poor UI that was making do with what was yet available.
I've read this many times over the years, sort of enamored by how such a strange phenomenon popped out of a factory setting.
In the most 2016 update the relative says it's common to see weird effects from the spools. If it's so common it should be reproducible I would think, yet I've never seen it done.
I like to preface negativity with a positive note. Maybe I am influenced in my word choice but my intent was to point out that this is a very, very impressive feat and I don't want to undermine it.
While it is impressive and I like to follow the advancements in this field, it is incredibly frustrating to listen to. I can't put my finger on why exactly. It's definitely closer to human-sounding, but the uncanny valley is so deep here that I find myself thinking "I just want the point, not the fake personality that is coming with it". I can't make it through a 30s demo.
Guardrails seems to be a common argument. It feels heavy handed though to use an entire library/ecosystem to ensure you don't have typos, your tests should catch them anyhow.
And migrations are a separate feature that a number of ORMs have, you can have clean migrations with pure SQL using a tool like gomigrate.
I think the real utility is that the ORM is doing the marshalling for you.
Agreed. I think the earlier days of Angular put a bad taste in many people's mouths. Many people who complain about Angular being verbose seem to only have experience working on smaller apps.
I thought this was a joke at first. He says he "feels great" but I can't help but be skeptical about that claim. Maybe I'm soft because I think a 40 is reasonable and prudent.
To say it another way, you should not expect more code to be merged in the same amount of time as less code. 250 lines is 5x more code, so what is the purpose of a statistic showing 1/5 of the code gets merged faster? That's not a good or a bad thing.
Unless the metric is actually 40% faster per LOC but that's not specified.
I am curious if there has been any research on temporal attention in humans. I'm not sure how you'd quantify it. But in myself I know that I'm constantly predicting where something will be or what it will look like based on how it did a second ago. It's probably the root of reflexes.
This looks like the video equivalent of Dall-E 1. Hard to believe how far we've come so quickly.
The paper talks about "pseudo 3D attention layers" that are used in place of temporal attention layers for each dimension due to memory consumption. It seems like AI research is vastly outpacing GPU development.
I feel it's the typical HN reaction to be negative about generated content. I enjoyed it. I think it could have used another pass on editing but I felt the idea was interesting and I wanted to hear it through.
This is something I have noticed with a lot of models. I'm not sure what the technical term for it is, but when there is a repeated sequence of human input with model generation following (like a chatbot) it seems to be unable to focus. When you prod it to regain focus and come back to the topic being discussed it starts making up lies.
If you use GPT3 for a large amount of content generation the issue of focus doesn't seem to be so prevalent, but it has zero guarantee of truth.
This is my major gripe with the college courses I took. They start at the molecular level and build to the planetary scale. This works for some people, but for me I had the hardest time finding a personal interest until I got to the application. It was the application I wanted to build, and to do actual work you don't manipulate the molecules directly.
Build an application, then as you need to learn more to get things done you naturally pick up a deeper knowledge of the tech you're using.