Butter

@alfa155stick @mikecmorgan you need to see a therapist

@mikecmorgan Armed police can't use their hands in case someone tries to take their gun, they also don't know what they be carrying. Completely reasonable force. And before the wet wipes start asking how I would feel if police kicked in the head, I wouldn't put myself in that situation.

Dear cowardly cops on this cesspool of an app. This is not reasonable force. It’s why we film you and your snide mates without fear or favour…

@alecanalec @Anxiousvids @Contimasque "invades the girl's space" lmao her entire job is to give him the space to play. retard.

@Anxiousvids @Contimasque Acaso ud es ciego? No ve la prepotencia y arrogancia de Jodar que invade el espacio de la chica y hace que ella deba caminar hacia atrás para dejarle el espacio libre, y que al hacerlo se tropieza? Aunque no la tocó fue una agresión. Un desgraciado arrogante y clasista.

Se viralizó un clip de 3 segundos donde el tenista Rafael Jódar supuestamente empuja a una "Ball girl"... Pero en otro video (que tiene más amplitud) se aprecia mejor que la joven tropieza sola y que él, tras lanzar una botella de agua hacia su box, hizo un gesto con la mano izquierda mientras caminaba, señalando a su padre (que estaba en el palco de jugadores) para que le preparara o le diera suplementos que pensaba tomar después de un break 🤷🤷..

@ridemarkets testing, call I made: SQUIREabove$0.01byJun 1, 6:17am how did it end up being; SQUIREabove$0.00696835byMay 30, 6:17am also: - UI showed errors, but call placed in the backend - SQUIRE above $0.00696835 by May 30, 6:17am - it's actually 'at' and not 'by' - very different

what a week (give or take) for Riders - six live funds, mix of private, invite only and public funds - +$40k TVL (MORE funds incoming) - 113 backers - 23 callers some highlights:

@badattrading_ Such a cope post.

$ALTSZN is still bundled See, from time to time they let them run for a very long time, it doesn't mean it's established at all. They are just patient.

@banditxbt kook has shilled probably 10x more rugs tbf

Dave Portnoy RAGEQUITS just 15 minutes into his livestream after Kook ragebaits him by asking about his next rug Kook: "hey dave when is your next rug???? you have launched 50 rugs why stop here" Dave: "And on that note, see ya"

@0xngmi could add a disclaimer instead of removing?

Defillama removed some tvl for a chain that was inflating it and now I see their team writing novels about how tvl is not a good metric to value them Yeah I’m sure it has nothing to do with it being low for their chain

@mikulaja Can't get worse than getting liquidated on 100x leverage.

What could possibly go wrong!

@PINTO03091 @skalskip92 Just 6000 hours you say?

I'm not sure what level of "special thing" you mean, but here's what I did: 1. Spend a lot of time on data design. 2. Draw high-quality rectangles using all my resources. 3. Perform training with a very small amount of data, completely ignoring the evaluation of eval and only checking all the output results, and repeatedly analyzing the variations that the model struggles with. 4. Review the annotation criteria and redo all the annotation from the first image. 5. Repeat steps 3 and 4 until you reach an accuracy you can accept. 6. If the accuracy still doesn't improve, take dozens of extremely difficult images by taking selfies based on the analysis results and add annotations. 7. Repeat steps 3 through 6. 8. Complete training multiple times with the model with the largest parameter size published in the paper's implementation, and fine-tune the hyperparameters with expert skill. By visualizing the training process log, resolve one by one the large number of bugs that are built into the paper's implementation from the beginning. Things I emphasized: 1. When annotating, do not annotate classes with vastly different categories all at once. For example, simplify the work by grouping categories such as upper body only, lower body only, or hands only. 2. Read the source code of the paper's implementation pipeline so thoroughly that you memorize it. 3. Almost completely ignore the architecture of the paper's implementation (because no matter what architecture you choose, there won't be an evolution as dramatic as CNN vs. Transformer). 4. Ignore the eval values ​​(there's no point in comparing them since it's based on unique data). 5. Visually check the final output of the model for every single image (to understand the weaknesses of the architecture, the weaknesses of the data, and the true potential of the architecture). 6. For example, if there is a problem with the segmentation mask, visually check the quality of the training data for every image. Then, thoroughly analyze what image variations are lacking and what perspectives are missing from the original training data. 7. Since everyone is aware of the low quality of publicly available datasets, we will design a custom CNN/Transformer architecture for generating training data to address the issues thoroughly analyzed in the previous steps, and refine the model for several months until satisfactory output is obtained. 8. Using the generated model with satisfactory performance, we will output a segmentation mask to be used as training data. 9. Augmentation and hyperparameter tuning are only minor improvements, so we will avoid making changes as much as possible until all the aforementioned steps are completely finished.

超透過画像。

@soby0x False equivalence

This is so real. on a family trip to Japan with significant size in stocks and not once did I have to panic about being rugged. if I had even half the size on perps I wouldn’t have left the hotel once

@AltcoinSherpa You can probably count the stocks with 100x potential on your fingertips + no retail normie is getting an entry on the ground floor.

Nearly everyone is in crypto because of the money and the 100x potential. Now, you can do 100x in stocks with real fundamentals, charts that don't jeet to 0 overnight, and have much cleaner charts/more organic moves. All while crypto chops around and continues to underperform

@a1lon9 @Pumpfun Where's my airdrop?

@Pumpfun another step towards making the pump fun app the greatest place to trench on the go!

The pump fun app just got BIGGER! Introducing frictionless multichain trading: trade Ethereum, Base, BNB, and more on the pump fun app now One wallet, one currency, more chains. Join now 👇

@CryptoKaleo mid question, asked every cycle. SBF gone now what? NFTs done now what? Degods migrating now what? Pump multi chain now what?

Honest question: What is the primary bull case for Solana now with pump fun opening the doors to EVM chains & USDC?

@vibhu @drews888 @SolanaFndn @PhoenixTrade can you please keep trading so we can copy trade inverse vib?

@drews888 @SolanaFndn @PhoenixTrade Thanks king. Not even a competition 💀 Per our DMs I donated $1k to St Jude’s instead 🤝

Memecoin traders don’t define Solana, apparently perp traders don’t either. Huge props to @vibhu / @SolanaFndn for taking on this challenge and giving me the opportunity. It’s been fun. Regarding @PhoenixTrade, much better than i thought, there’s still a lot of work that needs to be done, but the team, based on my limited interactions with them, has seemed competent enough to figure it out. I come out of this with more friends, more experience and more cash to buy $HYPE. Thank you all for following along! Hyperliquid.

@VIOtiluke @DerCheapi @m4h007

In my view, prior to the war, Iran’s urban youth were suffering from alienation: they wanted to be part of something bigger, but the only bigger thing the state provided was religion. That made the West and its ideals attractive. Something to attach themselves to, primarily through consumption of Western cultural products, and in some instances creating their own. But the war changed that. Many are still sitting on the fence, but for many others this war is a story they can be part of. A transcendent framework they can participate in. They’re now part of the story of national survival, belonging, sacrifice, and prevailing against mighty enemies. As I have argued before, moving forward, the Islamic Republic will consolidate and become the uncontested manifestation of the Iranian sovereignty.

@x_kouroshXgreat @DerCheapi @m4h007 Please sir muh oil open the strait please

@Butter4sol @DerCheapi @m4h007 Cope is the ex supreme leader getting mauled, new getting half mauled but then having the audacity to declare victory 😂 you lil snail 🐌

@yatakeni Spill the beans bruh

hearing things

@ip_atusah14 @EylonALevy Yeah we still have some spine left in us

@Butter4sol @EylonALevy I would have been surprised to hear differently from someone in Pakistan.

Qatar, Saudi, and Pakistan should join the Abraham Accords. They should recognize the State of Israel, because it has existed for nearly 80 years and they should recognize reality. The world must not continue normalizing this abnormal situation.

@_A_khalifa Stfu cuck

THE ABRAHAM ACCORDS! The entire region will move towards them. Enough with the divisions and terrorism! The UAE is a founding member and knows exactly what it’s doing so follow its lead!

@ip_atusah14 @EylonALevy There will never be a regime change in Iran, get over your delusion

@EylonALevy This can happen only after regime change in Iran, before that it would be simply wishful thinking.

@p_olitburo @m4h007 Copeeee

@m4h007 You have no fucking idea what you're talking about. Are you a sociologist? They were -- still are -- suffering from alienation. What they don't suffer from is Stockholm syndrome. Your post is one of the most retarded, egoistic and ignorant ones I've seen recently. FUCK. YOU.