fire

closest i've seen so far is https://llamahub.ai, but afaict it's only for loaders atm

yeah, the pope thing just drove way more traffic than they were prepped to handle

are you assuming they decided this or did they state it somewhere?

hey man, hope things have been going well for you

on the main topic, have you heard about RWKV ( rnn based gpt style network )? the project's actively working on implementing "infinite" context length support, which would probably pair very well with a project like yours

now that's a name I haven't seen in years, hi mpoon!

also this is slick as hell

idk maybe reddit's nootropics subreddit

replaceable key "frame" had me hopeful, but no, there is no ortholinear support - only staggered layouts. Lame.

this just seems to be a new attempt at the ol' Optimus Maximus

Unfortunately Dust.tt doesn't use a modern email validation method and so fails on "modern" TLD's like .media, making it impossible to sign up for an activation key with it.

I think a github jekyll or similar static site blog is probably the easiest option?

iirc aws exists mainly because amazon kept needing capacity, and after solving that problem it then had a bunch of excess capacity

I love that the name makes this feel like typing a sentence to your terminal

have you been able to contact OpenAI about this? It sounds like they're actively adding load to your CS ops with this

I dislike most mono fonts, but I found a monospace version of Atkinson Hyperlegible a while back and was using it until the author made a new monospace font they call Unolexic[1]; I've been using it since. They also have done posts comparing different parts of different fonts[2], making quasi-monospace fonts[3], etc., which are fascinating to read.

1: http://thedarnedestthing.com/unolexic

2: http://thedarnedestthing.com/monolexic%20timeline

3: http://thedarnedestthing.com/quasimono

As far as I can tell, all this does is mitm yourself so the proxy server can change the data being sent to the server, which is like, the most basic of possible ways to cheat a game, and only works on games that fully trust the client and have absolutely no server-side validation. Yikes

like I probably don't understand this in the right way, but I could have sworn we had the ability to probe latent space on models like these and make mappings based on them? Or was that only for diffusers?

the full segment also includes disclaimers on this so it isn't quite as cut and dry:

> 4. Defense of Third Party Claims.

  Notwithstanding any other language in your Agreement, GitHub will defend you against any claim by an unaffiliated third-party that your use of GitHub Copilot misappropriated a trade secret or directly infringes a patent, copyright, trademark, or other intellectual property right of a third party, up to the greater of $500,000.00 USD or the total amount paid to GitHub for the use of GitHub Copilot during the 12 months preceding the claim.  
  GitHub’s defense obligations do not apply if (i) the claim is based on Code that differs from a Suggestion provided by GitHub Copilot, (ii) you fail to follow reasonable software development review practices designed to prevent the intentional or inadvertent use of Code in a way that may violate the intellectual property or other rights of a third party, or (iii) you have not enabled all filtering features available in GitHub Copilot.

and as I understand it:

  • i) means you can't modify copilot suggestions without losing this protection,
  • ii) isn't actually defined; this might mean they can use it as a crutch to avoid providing protection,
  • iii) means you aren't protected if you fail to enable settings that are not on by default but in theory allow you to avoid generations that reference "radioactive" licenses like gpl.

this seems like it covers non-business copilot as well

I don't understand why they aren't tagging data with license information and allowing users to use models that don't include certain licenses - seems like it would be the middle ground given the stance they've taken; like, "we don't think it's a problem, but if this makes you feel better you can use these other models that specifically don't train on gpl code, or whatever"

I would prefer to see full license attributions included in generated responses, though. Something that then also wouldn't be that difficult to generate a licenses file from?

Amazon's CodeWhisperer has a "reference tracker" that tells you the license of training data code if the generated response is within some similarity threshold, but that's still not good enough imo.

this is neat, do you have any docs/posts about it? I presume it isn't on github?

I honestly thought Glimpse was the rename, but I guess they're different projects

thanks for this, I've been needing to factory reset my L16 but I couldn't locate any of the firmware

if memory serves camelcamelcamel gets their pricing data directly from amazon via a source they aren't allowed to disclose, and bloomberg has previously reported that amazon gives camel^3 the data they have; given that, i'm not too certain they're the cause of amazon's consistent reduction in transparency to consumers - a lot of it feels like cruft built into their marketplace allowing for bad actors to effectively play the marketplace

Waez

not much longer 'till it straight up tells me to take the highway to the danger zone

unfortunately I put in "sonic the hedgehog" and the result was...

... nothing? like, there's no playback at all. Is that expected?

right? Does linking to reddit qualify since there's a subreddit for elonjet as well?

It looks like they may have tweeted something about the elonjet account that still exists on mastodon, and were suspended under the new version / enforcement of Twitter's doxxing rules.

is the entertaining ads channel perhaps Aging Wheels? His ad skits have been great

> *Basically, we aren't targeting support for graphical applications running on Linux because there is very little demand for this - but we cover everything else. You CAN run graphical applications on Windows vs. a Linux server.

Ah, disappointing; I was hoping to try this with a Steam Deck as an alternative to using Moonlight and streaming the entire game from the Windows machine.

I'm at the server limit and can't join additional servers without leaving something else. Can you add this stuff to the github repo?