🎵 Purr 4 Me (Official Music Video) 🎵 Watch FULL Music video on YouTube! https://t.co/c1yYPS5BCF Music/Producer/Director: Sat Grewal

Friends, do not try this at home. This stream is for observation only, not medical advice, legal advice, or encouragement. The activity is in a legal setting, with a professional facilitator and being medically supervised. Psilocybin is illegal in many places and carries real

I’m taking magic mushrooms and livestreaming it on sunday. + 6 hours Live + 10am PST, Sun 30th + 5.24 g mushrooms, dried + 28 mg of active psilocybin + Strain B+ (psilocybin cubensis) + 2nd dose (3 total across 3 months) We’re measuring 249 biomarkers to determine the effect on

Cancel everything you were doing Sunday

If you went all out yesterday and feel like shit today, I invite you to go on a walk with me right now. See if you can break away from whatever you're doing, or if you can't, save this and read it later when you can be out walking. Evidence shows that walking and talking

You are one of my biggest inspiration in Generative AI work @Casey. 🫡🙏

Casey Neistat worked the hardest in beginning. Made vlogs, every, single, day! Did not monetise his channel for long time, resulting in no ads to show on his videos. Excellent quality, faster delivery, absolutely enjoyable vlog content. He is my prime example of not caring how

Source:

📁 Casey Neistat says the rise of AI is unstoppable and compares it to a giant wave that cannot be held back. He predicts that full AI-generated films will arrive within a couple of years and that quality will escalate quickly. He notes that early attempts will look rough, but

The update commit can be found here

The training adapter is active while training and is removed while inferencing / sampling. This makes it so your fine-tuning is done a de-distilled version of the model but your new information is carried over to the distilled version of the model

The de-distill training adapter can be found here https://t.co/QPDJxnVcGZ . It was trained by slowly training the turbo model on images generated by the same model in order to slowly break down the distillation without adding any new information.

AI Toolkit now supports training LoRAs for @Ali_TongyiLab Z-Image Turbo by utilizing a de-distill training adapter. This allows you to finetune the model, while preserving turbo inference capabilities. More in 🧵

Based and Z-Pilled https://t.co/JVPwi2xmFm

If you are in this AI space you need to watch this video!!!!

I've now watched the entire @GoogleDeepMind documentary "The Thinking Game." I actually wasn't aware until now that "AGI" (Action for Intelligent Applications) has been DeepMind's stated goal since its founding. One thing you have to give Google credit for: making the

Cars to $TSLA are what books were to $AMZN Very few understand this

What if the world's most famous photos were taken today? Well, let's find out! 1. Marilyn in Times Square holding her iPhone in her hand

@JamianGerard I know some great artists who don’t even bother to take awards in biggest awards ceremonies in history. It’s the audience who ultimately matters. If your content is consumed and liked, that’s it. Nothing else matters. Be it AI or not, only matters is how enjoyable it is. Keep

Impressive color grading, but plastic skin! No problem, Magnific to the rescue! Just use these settings: - Magnific Skin Enhancer - Creative - Sharpen: 0% - Smart Grain: 8%

1. Enhancing Midjourney Images Midjourney is amazing, but it has the problem of producing plastic looking skin as soon as the scene gets even slightly complex (if it’s a very closeup face, there’s no issue). Let’s say you generate an image with v7 using this prompt: "face