WalnutLum @ WalnutLum @lemmy.ml

Posts

2

Comments

499

Joined

2 yr. ago

Just start using the yuan

https://www.youtube.com/watch?v=_3gD_lnBNu0

The ATC queries them multiple times if they have visual on the CRJ and they repeatedly respond in the affirmative.

also the published helicopter corridor has them maintain a ceiling of 200 feet, they were well above 350 at the time of the collision.

It's either pilot negligence or instrument failure.

also TCAS is inactive below 800900 feet.

I'm sure imprisoning 30000 innocent people next to highly trained terrorists won't come back to haunt the US.

You're going to have to learn python.

Here's a good overview: https://huggingface.co/docs/transformers/training

Or open source groups can make a fully open repro of it: https://github.com/huggingface/open-r1

Yea except DeepSeek released a combined Multimodal/generation model that has similar performance to contemporaries and a similar level of reduced training cost ~20 hours ago:

https://huggingface.co/deepseek-ai/Janus-Pro-7B

Hell yea +1 for shepherd.

Declarativity on steroids.

Nowadays I'd say that applies more to balatro than anything else

Trying to dodge Chinese traditional medicine in Asian countries when you go to the doctor is such a pain in the ass.

I had to go to three different Japanese doctors before they would give me something other than herbal powder.

More than likely we'll see in/outbox functionality similar to activitypub, so the PDSes could interop with ActivityPub and individually talk to one another

This is not true.

The App view, relay, PDS, and mobile app, code is all open source. Anyone can spin up their own version if they want.

If you want a complete copy of bluesky (reflecting the bluesky post firehose) it's prohibitively expensive, but you can spin up your own version if you want.

The only thing that's missing is bluesky federating anything other than PDSes by default, but you can 100% go in the other direction.

A good overview is here: What bluesky is and what bluesky is not

The fact that it's hit #1 on the US Apple app store is wild to me.

Five eyes agents must be pulling their hair out.

Meanwhile all the TikTok users seem to be moving to a copycat called Red Note

But he went on to say: “We’re not adding any more software engineers next year because we have increased the productivity this year with Agentforce and with other AI technology that we’re using for engineering teams by more than 30% – to the point where our engineering velocity is incredible. I can’t believe what we’re achieving in engineering.”

This announcement is just advertising for agentforce (their AI) they're likely not being serious about it.

This is already implemented on windows

Tools > Preferences > Show settings=All > Video \ Subtitles/OSD: Text rendering module [Speech synthesis for Windows]

It's actually using whisper.cpp

From the README:

Memory usage Model Disk Mem tiny 75 MiB ~273 MB base 142 MiB ~388 MB small 466 MiB ~852 MB medium 1.5 GiB ~2.1 GB large 2.9 GiB ~3.9 GiB

Those are the model sizes

I've never seen a big company like bytedance give PTO for a death in the family, much less for someone's house burning down.

Any of those companies big enough middle management can hide behind "company policy" and then off their hearts and brains.

I mean DMA (direct memory access) devices are only like 170 bucks now:

https://www.dma-cheats.com/dma-cards

These are almost impossible to do anything about even client-side because they operate on a hardware level.

In general state-reading hacks (like invisible walls and Gameworld state information hacks) are almost impossible to do anything about, to the point where when companies are able to find a way to defeat one of these things it's huge news.