g413n

no profile record

comments

5 ay öncediscuss

it's a pretty general policy but this is all super early, it's great at exploring websites so fuzzing was easy, for CAD it has good enough base rates with the few-shot prompt when we do the repetitive stuff, and we gave it checkpoints on each step, the other stuff in the mosaic are just some of our favorite clips from internal evals

5 ay öncediscuss

relevant note is that we finetuned by having the human also use arrow keys which keeps it in-distribution but also slower to collect

5 ay öncediscuss

yeah we actually had some wacky ideas with ctc + a reverse-causal mask but diffusion does just make it all a bit more simple

5 ay öncediscuss

we do exponential binning but fwiw I think we can do way better just hasn't been the main research area initially

5 ay öncediscuss

yeah we've done audio work in the past so we'll def merge the recipes at some point, long term should have full io that a human has (except maybe not generating video for video calls that seems a bit much)

7 ay öncediscuss

so cool :)

7 ay öncediscuss

what's the basis for conversion between hours of neural data to number of tokens? is that counting the paired text tokens?

9 ay öncediscuss

yep this. we just turned off management

9 ay öncediscuss

we would be so down to buy s3 junkyard tbh we were going around begging various storage clouds to offer us this before giving up and building it ourselves

9 ay öncediscuss

yep this was the case for us.

9 ay öncediscuss

I think <2x more drives than needed, not 20x (24 vs 14TB), but the racks holding the drives could've been denser. Around the same cost in any case and our colo doesn't charge for space, so it's not a big deal and we were just going with what we were familiar with, but something to try.

9 ay öncediscuss

yeah we weren't sure about putting that number esp whether it includes all the image attachments, but in any case it's at least around the right reference class for the largest text data operations.

10 ay öncediscuss

def agree on the setup fees, that was just a price crunch to get it done within the weekend. (too short-notice for professional services, too sensitive for craigslist, so basically just paying a bunch of folks we already knew and trusted)

for IPXE do you have any reference material you'd recommend? we had 3 people each with reasonably substantial server experience try for like 6 hours each and for whatever reason it turned out to be too difficult.

10 ay öncediscuss

yeah that's why we started paying people near the second half- not super clearly stated in the blogpost, but the novelty definitely wore off with plenty of drives left to stack, so we switched strategies to get it done in time.

I think everyone who showed up for a couple hours as part of the party had a good time tho, and the engraved hard drives we were giving out weren't cheap :p

10 ay öncediscuss

yeah the cost calculus is very different for gpus, it absolutely makes sense for us to be using cloud there. also hardly any datacenters can support the power density, esp in downtown sf

10 ay öncediscuss

it's not just in sf it's across the street from our office

this has been incredibly nice for our first hardware project, if we ever expand substantially then we'd def care more about the colo costs.

10 ay öncediscuss

yeah it's on the wishlist to try

10 ay öncediscuss

our training stack doesn't make strong assumptions about data integrity, it's chill

10 ay öncediscuss

When I was working at sfcompute prior to this we saw multiple datacenters literally catch on fire bc the industry was not experienced with the power density of h100s. Our training chips just aren't a standard package in the way JBODs are.

10 ay öncediscuss

just general research work. Once the recipes are efficient enough the modality is a smaller detail.

On the product side we're trying to orient more towards 'productive work assistant' rather than the default pull of audio models towards being an 'ai friend'.