One thing that has surprised me (and I should've known that it wasn't great at it), but it is terrible at creating bounding boxes around things it's not trained on (like bounding parts on a PCB schematic.)
Ugh, I don't have it. It was from before I used git.
Basically to do this you have a cups server that exposes itself as a network printer that prints to a specified PDF directory and then you have a program watching that directory for new files and if there's a new one it opens up whatever pdf viewer you want in full screen.
One thing I've been thinking about is if you could use a model like this as the first pass for permitters (Like a GitHub Actions CI/CD) who review blueprints.
Many developers use the regulatory side of various engineering approval processes as a quality control check which costs money and time for the regulator who is tasked with enforcing a standard.
It would also be good to speed up the workflow for developers saying hey, this thing looks weird did you really mean to do this?
And then further on, you could add a way to check it for constructability. My framer friends often get annoyed at whatever engineer because the way the structure is designed is materially inefficient or hard to construct.
Azure charges differently based on deployment zone/latency guarantees, OpenAI doesn't let you pick your zone so it's equivalent to the Global Standard deployment (which is the same cost).
I'd be interested in knowing if anyone is seriously using the assistants API, it feels like such a lock in to OpenAIs platform when your can alternatively just use completions that are much more easily interchanged.
This generally resonates with what we've found. Some colour based on our experiences.
It's worth spending a lot of time thinking about what a successful LLM call actually looks like for your particular use case. That doesn't have to be a strict validation set `% prompts answered correctly` is good for some of the simpler prompts, but especially as they grow and handle more complex use cases that breaks down. In an ideal world
> chain-of-thought has a speed/cost vs. accuracy trade-off
a big one.
Observability is super important and we've come to the same conclusion of building that internally.
> Fine-tune your model
Do this for cost and speed reasons rather than to improve accuracy. There are decent providers (like Openpipe, relatively happy customer, not associated) who will handle the hard work for you.
I use this all the time and you can copy a range of values with multiple cursors and then paste them into the same size range. It's nice for hashes or modifying static lists that all need the same changes, but are annoying to regex for
I'm generally interested, but it would be helpful to understand what the different plans are actually offering? I have no idea what the difference is between:
Also, the music on your demo video is a bit excessive, I don't really expect hype music when I'm looking at a demo. It would be more helpful to have someone talking explaining what's going on (or silence)
I have a cluster using terraform-hcloud-kube-hetzner [0] and I'm quite happy with it.
It uses MicroOS for rolling releases and auto-upgrades for less maintenance. I'm quite happy with it so far, and I like that it integrates with terraform.
A big reason that Ontario can't / doesn't is that the demand centers in Southern Ontario are ridiculously far from the generation centers in Manitoba. Toronto is 1800 km straight from the new Keeyask Dam. This is over unpopulated muskeg that is expensive to build on in the first place and difficult to access for maintenance.
I think it really depends on the application of web scraping. (As someone who does, what is in my mind, ethical web scraping)
- Scraping public information from government websites to do analysis: ethical, it's the public's data
- Scraping to help some companies customers more effectively use that companies product, for example scraping a medical office's insurance claims to help them automate their insurance remittance process: ethical
- Scraping faces to build a surveillance-tech company: disgusting
- Scraping your own website because your internal processes are so broken you can't get it any other way: ethical
- Scraping to just copy someone's data they worked hard to generate to go and resell: unethical
I work for Build Canada and I would love to see some maps from the fur trade and early exploration to tell stories.
If you want to chat my email is brendan at buildcanada.com