Video yükleniyor...

Video Yüklenemedi

Bu video yüklenirken bir sorun oluştu. Bu geçici bir ağ sorunundan kaynaklanıyor olabilir veya video kullanılamıyor olabilir.

Ana Sayfaya Dön

🏴󠁧󠁢󠁳󠁣󠁴󠁿⚽️| 𝙏𝙤𝙢𝙢𝙮 𝘾𝙤𝙣𝙬𝙖𝙮! Middlesbrough's Scotland duo combine! An excellent chipped cross by Ben Doak who finds Tommy Conway at the back post and his header makes it 2-0 to Boro' v Hull City That’s Conway's seventh Championship goal this season

Football Park Scotland

41,557 subscribers

110,599 görüntüleme • 1 yıl önce •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Yorum

Yorum bulunmuyor

Orijinal gönderinin yorumları burada görünecek

Benzer Videolar

“That was fun, but not as fun as tackling people.” Preston Lobliner had these words of wisdom after his first day of kicking camp with James Wilhoit. Without any training, he was the only kid in his league last year who could kick extra points, and that was HUGE since they are worth 2 points since nobody does it. With Middle School Football upon us, we are excited to see him contribute to his team as an accurate and powerful kicker and an all out DAWG on offense and defense. This was 30 yards, and after this camp and learning how to actually kick, I expect to hit 40+ by the time the season hits. Also the best news ever for us, he is starting back in wrestling next week daily after injuring his elbow at practice the day after winning state so we will start working on that 3rd state championship! SUPER PROUD of this kid!

“That was fun, but not as fun as tackling people.” Preston Lobliner had these words of wisdom after his first day of kicking camp with James Wilhoit. Without any training, he was the only kid in his league last year who could kick extra points, and that was HUGE since they are worth 2 points since nobody does it. With Middle School Football upon us, we are excited to see him contribute to his team as an accurate and powerful kicker and an all out DAWG on offense and defense. This was 30 yards, and after this camp and learning how to actually kick, I expect to hit 40+ by the time the season hits. Also the best news ever for us, he is starting back in wrestling next week daily after injuring his elbow at practice the day after winning state so we will start working on that 3rd state championship! SUPER PROUD of this kid!

Marc Lobliner - IFBB Pro

28,108 görüntüleme • 1 yıl önce

May be the most difficult hurdle for a wrestler to clear—trusting that the same offense that built the lead is what protects it. The instinct is to protect the lead, but the approach that creates better odds, less regret, and is simply more logical is to keep wrestling the same way that created the lead. Stay on the attack, stay in position, and keep applying pressure. The one who needs to change and create opportunity is the opponent who is losing. So why would you go ahead and assist them in this by back peddling and lead protecting, opening up two windows that weren’t there before? 1 – Stall points 2 – Allowing your opponent to solely focus on their offense as they no longer have to respect your attacks. You hear wrestlers in post-match interviews “We do this all the time in the room! Down 2 with 30 seconds left and working to find a takedown.” Not so much the opposite… “up 2 with 30 seconds left and just have to find ways to back up to protect the lead!” So why resort to something you don’t practice in the most crucial moments of a match—or your season? This is from the excellent series “The Climb” by Stilly Boys on YouTube — a segment from Episode 2: Road to the Big 12s.

May be the most difficult hurdle for a wrestler to clear—trusting that the same offense that built the lead is what protects it. The instinct is to protect the lead, but the approach that creates better odds, less regret, and is simply more logical is to keep wrestling the same way that created the lead. Stay on the attack, stay in position, and keep applying pressure. The one who needs to change and create opportunity is the opponent who is losing. So why would you go ahead and assist them in this by back peddling and lead protecting, opening up two windows that weren’t there before? 1 – Stall points 2 – Allowing your opponent to solely focus on their offense as they no longer have to respect your attacks. You hear wrestlers in post-match interviews “We do this all the time in the room! Down 2 with 30 seconds left and working to find a takedown.” Not so much the opposite… “up 2 with 30 seconds left and just have to find ways to back up to protect the lead!” So why resort to something you don’t practice in the most crucial moments of a match—or your season? This is from the excellent series “The Climb” by Stilly Boys on YouTube — a segment from Episode 2: Road to the Big 12s.

Cornell Kevin

42,096 görüntüleme • 6 ay önce

Repost from @gritadelphia_ • PHILLY SPORTS FANS HAVE HAD ENOUGH OF SPIKE ESKIN 😤🗑️🚮 This is a man who has made a career out of hating on Philadelphia’s biggest moments. When Bryce Harper hit for the cycle and went 7-for-9 with four RBI in one of the most dominant weekends in franchise history, Spike Eskin’s response was to call him “selfish.” Not celebrate it. Not acknowledge one of the great individual stretches in Phillies history. Attack it. When Jalen Hurts led this franchise to a Super Bowl and won Super Bowl MVP, Eskin still couldn’t bring himself to call him a top-five quarterback, spending an entire season riding the Hurts hate train while the rest of the city moved on to celebrating a championship. And when LeBron James, the greatest player of this generation, signed with the Sixers, Eskin went on the radio and called it “one of the worst days” of his fandom, then spent days trying to convince everyone he wasn’t just doing it for clicks. There’s a pattern here, and Philly sports fans see it clearly. Every time this city has a reason to celebrate, Spike Eskin finds a way to make it about himself and his manufactured contrarian takes. Hold Bryce Harper accountable after an MVP-caliber weekend. Dismiss a Super Bowl MVP quarterback. Trash the biggest free agency addition in Sixers history. It’s not analysis, it’s rage bait dressed up as “honesty,” and fans have caught on to the act a long time ago. Philadelphia doesn’t need a hater masquerading as a fan. We’ve got championship banners, championship quarterbacks, and now the greatest player alive suiting up for the Sixers, and no amount of forced negativity from Spike Eskin changes any of it.

Repost from @gritadelphia_ • PHILLY SPORTS FANS HAVE HAD ENOUGH OF SPIKE ESKIN 😤🗑️🚮 This is a man who has made a career out of hating on Philadelphia’s biggest moments. When Bryce Harper hit for the cycle and went 7-for-9 with four RBI in one of the most dominant weekends in franchise history, Spike Eskin’s response was to call him “selfish.” Not celebrate it. Not acknowledge one of the great individual stretches in Phillies history. Attack it. When Jalen Hurts led this franchise to a Super Bowl and won Super Bowl MVP, Eskin still couldn’t bring himself to call him a top-five quarterback, spending an entire season riding the Hurts hate train while the rest of the city moved on to celebrating a championship. And when LeBron James, the greatest player of this generation, signed with the Sixers, Eskin went on the radio and called it “one of the worst days” of his fandom, then spent days trying to convince everyone he wasn’t just doing it for clicks. There’s a pattern here, and Philly sports fans see it clearly. Every time this city has a reason to celebrate, Spike Eskin finds a way to make it about himself and his manufactured contrarian takes. Hold Bryce Harper accountable after an MVP-caliber weekend. Dismiss a Super Bowl MVP quarterback. Trash the biggest free agency addition in Sixers history. It’s not analysis, it’s rage bait dressed up as “honesty,” and fans have caught on to the act a long time ago. Philadelphia doesn’t need a hater masquerading as a fan. We’ve got championship banners, championship quarterbacks, and now the greatest player alive suiting up for the Sixers, and no amount of forced negativity from Spike Eskin changes any of it.

Tiffany Paige

26,188 görüntüleme • 1 ay önce

This is facts and I'd even give two clear examples a lot of Arsenal fans may remember. Last season when we played Brentford at home, Saka was on an attack for us and he did not have a clear chance to shoot the ball, Gyokeres was down in the middle and squaring that pass would have resulted in an easier goal for us but Ska chose to shoot, luckily the ball went in but Kelleher had a bad punch on the ball which is why the ball even went in. Chelsea vs Arsenal Carabao cup semi final, Saka was on an attack again, he could have played the cross first time in for Gyokeres but he decided to cut back before now passing to Ben White who eventually squared it to Gyokeres, luckily for Gyokeres, Robert Sanchez let the ball slip underneath him which resulted in Gyokeres tapping the ball in. If you see Gyokeres's reaction in the buildup to the goal, he got frustrated Saka did not release the ball to him on time. There are many more occassions similar to this from Saka, Martinelli, Odegaard and co where they could have released Gyokeres quickly but chose not to. So yeah, this tweet I quoted is absolutely spot on.

This is facts and I'd even give two clear examples a lot of Arsenal fans may remember. Last season when we played Brentford at home, Saka was on an attack for us and he did not have a clear chance to shoot the ball, Gyokeres was down in the middle and squaring that pass would have resulted in an easier goal for us but Ska chose to shoot, luckily the ball went in but Kelleher had a bad punch on the ball which is why the ball even went in. Chelsea vs Arsenal Carabao cup semi final, Saka was on an attack again, he could have played the cross first time in for Gyokeres but he decided to cut back before now passing to Ben White who eventually squared it to Gyokeres, luckily for Gyokeres, Robert Sanchez let the ball slip underneath him which resulted in Gyokeres tapping the ball in. If you see Gyokeres's reaction in the buildup to the goal, he got frustrated Saka did not release the ball to him on time. There are many more occassions similar to this from Saka, Martinelli, Odegaard and co where they could have released Gyokeres quickly but chose not to. So yeah, this tweet I quoted is absolutely spot on.

Senator Fowosere

74,915 görüntüleme • 8 gün önce

NEW WORLD MODEL: Yann LeCun's team is back with an efficient model! This project involves Yann LeCun, Lukas Kuhn, Lucas Maes, Quentin Le Lidec, and Randall Balestriero. A couple definitions first: - DINO: self-DIstillation with NO labels. A self-supervised image model (Meta, 2021) where a student network learns to match a teacher (an EMA copy of itself) across two crops of the same image, with no labels and no negatives. - SIGReg: a regularizer that prevents embedding collapse by forcing the embeddings to match an isotropic Gaussian, tested with a normality test (Epps–Pulley) on many random 1-D projections instead of in full dimension. LeVJEPA is a self-supervised video pretraining method, released with open code, weights, and checkpoints. It learns a video representation by pushing the embeddings of global and local crops of the same clip together (an invariance loss), while a regularizer called SIGReg forces the embeddings toward an isotropic Gaussian to provably prevent representation collapse. Unlike V-JEPA and V-JEPA 2 it uses a single shared encoder with a projector and no target network, no predictor and no stop-gradient. It drops 95% of tokens per view, uses block-causal attention (each frame attends only to past frames), and has a single loss weight. It is evaluated purely as a representation learner via frozen probing on ImageNet-1K, Something-Something-v2 and Kinetics-400, not on any robot. What I find interesting, is that V-JEPA and V-JEPA 2 need an EMA target encoder, stop-gradients and a capacity-limited predictor to avoid collapse; LeVJEPA drops all of it for one shared encoder plus projector, preventing collapse instead with the SIGReg regularizer under a provable guarantee and a single hyperparameter. The "P" (predictor) in JEPA is effectively gone. LeVJEPA is also less compute intensive: - 5.6x to 20.8x lower total pretraining compute than V-JEPA 2 - 7.6 points higher on ImageNet-1K at matched FLOPs - trains at batch size 128 within 8GB where V-JEPA 2 saturates at batch size 28 Also worth mentioning: ImageNet-1K accuracy rises monotonically with the token-drop rate, from 33.9% at rho = 0 to 47.6% at rho = 0.95. The aggressive dropping is actually doing regularization work. On the JEPA-versus-DINO debate: - it loses to DINOv2 by 3.1 points on ImageNet-1K (appearance, static) - but wins on Something-Something-v2 by nearly 2x (motion, temporal) - and beats V-JEPA 2 by 1.9 points on ViT-L at 5.6x lower cost. -> optimized for temporal and motion understanding per compute dollar.

NEW WORLD MODEL: Yann LeCun's team is back with an efficient model! This project involves Yann LeCun, Lukas Kuhn, Lucas Maes, Quentin Le Lidec, and Randall Balestriero. A couple definitions first: - DINO: self-DIstillation with NO labels. A self-supervised image model (Meta, 2021) where a student network learns to match a teacher (an EMA copy of itself) across two crops of the same image, with no labels and no negatives. - SIGReg: a regularizer that prevents embedding collapse by forcing the embeddings to match an isotropic Gaussian, tested with a normality test (Epps–Pulley) on many random 1-D projections instead of in full dimension. LeVJEPA is a self-supervised video pretraining method, released with open code, weights, and checkpoints. It learns a video representation by pushing the embeddings of global and local crops of the same clip together (an invariance loss), while a regularizer called SIGReg forces the embeddings toward an isotropic Gaussian to provably prevent representation collapse. Unlike V-JEPA and V-JEPA 2 it uses a single shared encoder with a projector and no target network, no predictor and no stop-gradient. It drops 95% of tokens per view, uses block-causal attention (each frame attends only to past frames), and has a single loss weight. It is evaluated purely as a representation learner via frozen probing on ImageNet-1K, Something-Something-v2 and Kinetics-400, not on any robot. What I find interesting, is that V-JEPA and V-JEPA 2 need an EMA target encoder, stop-gradients and a capacity-limited predictor to avoid collapse; LeVJEPA drops all of it for one shared encoder plus projector, preventing collapse instead with the SIGReg regularizer under a provable guarantee and a single hyperparameter. The "P" (predictor) in JEPA is effectively gone. LeVJEPA is also less compute intensive: - 5.6x to 20.8x lower total pretraining compute than V-JEPA 2 - 7.6 points higher on ImageNet-1K at matched FLOPs - trains at batch size 128 within 8GB where V-JEPA 2 saturates at batch size 28 Also worth mentioning: ImageNet-1K accuracy rises monotonically with the token-drop rate, from 33.9% at rho = 0 to 47.6% at rho = 0.95. The aggressive dropping is actually doing regularization work. On the JEPA-versus-DINO debate: - it loses to DINOv2 by 3.1 points on ImageNet-1K (appearance, static) - but wins on Something-Something-v2 by nearly 2x (motion, temporal) - and beats V-JEPA 2 by 1.9 points on ViT-L at 5.6x lower cost. -> optimized for temporal and motion understanding per compute dollar.

Léo

90,490 görüntüleme • 29 gün önce

Batch Normalization by hand ✍️ ~ 7 steps walkthrough below Batch normalization is common practice for improving training and achieving faster convergence. It sounds simple. But it is often misunderstood. 🤔 Does batch normalization involve trainable parameters, tunable hyper-parameters, or both? 🤔 Is batch normalization applied to inputs, features, weights, biases, or outputs? 🤔 How is batch normalization different from layer normalization? So I drew and calculated one entirely by hand. Goal: normalize a mini-batch of 4 examples to mean 0 and variance 1, then let the network scale it back. = 1. Given = A mini-batch of 4 training examples, each with 3 features. = 2. Linear layer = Let us multiply by the weights and add the biases. Batch norm sits after this, which answers the second question: what gets normalized is features, not inputs, weights or biases. = 3. ReLU = We apply the activation, and -2 becomes 0. Negative values are suppressed before any statistic is taken. = 4. Batch statistics = Let us compute the sum, mean, variance and standard deviation, one row at a time. A row is a feature and the four columns are the four examples, so every number here measures one feature against the rest of the batch. That is the "batch" in batch normalization, and it is exactly what layer normalization does not do. The statistics are rounded to whole numbers, which is what keeps the rest of the page doable in pen. = 5. Shift to mean 0 = We subtract the mean, in green. The four values in each feature now average to zero. = 6. Scale to variance 1 = Let us divide by the standard deviation, in orange. Each feature now has variance one, whatever scale it arrived at. = 7. Scale and shift = We multiply by a linear transformation and pass the result on. The diagonal and the last column are trainable, so having just forced every feature to mean 0 and variance 1, we hand the network the means to undo it. The outputs: Mean of each feature = [2, 1, 2] Std dev of each feature = [1, 1, 2] To the next layer = [2, -2, 2, 0], [-3, 3, 6, -3], [2, 0, 1, 2] The answers: 🤔 Both. The scale and shift are trainable, the statistics are not. Epsilon and the momentum on the running statistics are the hyper-parameters, and one mini-batch by hand needs neither. 🤔 Features, after the linear layer, not inputs, weights or biases. 🤔 Batch norm measures across the batch, one feature at a time. Layer norm measures across the features, one example at a time. 💾 Save this post!

Batch Normalization by hand ✍️ ~ 7 steps walkthrough below Batch normalization is common practice for improving training and achieving faster convergence. It sounds simple. But it is often misunderstood. 🤔 Does batch normalization involve trainable parameters, tunable hyper-parameters, or both? 🤔 Is batch normalization applied to inputs, features, weights, biases, or outputs? 🤔 How is batch normalization different from layer normalization? So I drew and calculated one entirely by hand. Goal: normalize a mini-batch of 4 examples to mean 0 and variance 1, then let the network scale it back. = 1. Given = A mini-batch of 4 training examples, each with 3 features. = 2. Linear layer = Let us multiply by the weights and add the biases. Batch norm sits after this, which answers the second question: what gets normalized is features, not inputs, weights or biases. = 3. ReLU = We apply the activation, and -2 becomes 0. Negative values are suppressed before any statistic is taken. = 4. Batch statistics = Let us compute the sum, mean, variance and standard deviation, one row at a time. A row is a feature and the four columns are the four examples, so every number here measures one feature against the rest of the batch. That is the "batch" in batch normalization, and it is exactly what layer normalization does not do. The statistics are rounded to whole numbers, which is what keeps the rest of the page doable in pen. = 5. Shift to mean 0 = We subtract the mean, in green. The four values in each feature now average to zero. = 6. Scale to variance 1 = Let us divide by the standard deviation, in orange. Each feature now has variance one, whatever scale it arrived at. = 7. Scale and shift = We multiply by a linear transformation and pass the result on. The diagonal and the last column are trainable, so having just forced every feature to mean 0 and variance 1, we hand the network the means to undo it. The outputs: Mean of each feature = [2, 1, 2] Std dev of each feature = [1, 1, 2] To the next layer = [2, -2, 2, 0], [-3, 3, 6, -3], [2, 0, 1, 2] The answers: 🤔 Both. The scale and shift are trainable, the statistics are not. Epsilon and the momentum on the running statistics are the hyper-parameters, and one mini-batch by hand needs neither. 🤔 Features, after the linear layer, not inputs, weights or biases. 🤔 Batch norm measures across the batch, one feature at a time. Layer norm measures across the features, one example at a time. 💾 Save this post!

Tom Yeh

20,848 görüntüleme • 2 ay önce

Robinson Ortiz...Atta Baby! Ortiz had a HUGE year this past season, and got rewarded by being placed on the Dodgers' 40-man roster. Ortiz posted an ERA of 2.73, an average against of just .207, and had 72 Ks in 59.1 innings. And ALL of that is great. And, while the "execution" piece is still THE focus, keep in mind, he, in essence, hadn't pitched since 2021, due to Tommy John. Ortiz was a HIGHLY thought of prospect back in 2021, before his Tommy John, and came back to start 2024, but was shut down after just 6 outings and 11 innings in May of 2024. So, for him to be able to post 59.1 innings and have 48 appearances was HUGE in terms of showing that he's BACK. Then, to have the success on top of that, after such a long layoff, just goes to show how good his stuff is. Ortiz features a HUGE sweeping slider that starts well behind the back of a left-hander, and breaks from one side of the earth to the other. Trying to decipher whether that pitch is going to land "in the zone" or not is VERY difficult, even for hitters that have a LOT of MLB experience like he faced at OKC. His fastball is firm, and he can hit 6s and 7s, so those 2 pitches play well off of each other. Welcome aboard, Robinson. It's been a LONG road, one you've navigated successfully. Keep killin' it, my man. #dodgers

Robinson Ortiz...Atta Baby! Ortiz had a HUGE year this past season, and got rewarded by being placed on the Dodgers' 40-man roster. Ortiz posted an ERA of 2.73, an average against of just .207, and had 72 Ks in 59.1 innings. And ALL of that is great. And, while the "execution" piece is still THE focus, keep in mind, he, in essence, hadn't pitched since 2021, due to Tommy John. Ortiz was a HIGHLY thought of prospect back in 2021, before his Tommy John, and came back to start 2024, but was shut down after just 6 outings and 11 innings in May of 2024. So, for him to be able to post 59.1 innings and have 48 appearances was HUGE in terms of showing that he's BACK. Then, to have the success on top of that, after such a long layoff, just goes to show how good his stuff is. Ortiz features a HUGE sweeping slider that starts well behind the back of a left-hander, and breaks from one side of the earth to the other. Trying to decipher whether that pitch is going to land "in the zone" or not is VERY difficult, even for hitters that have a LOT of MLB experience like he faced at OKC. His fastball is firm, and he can hit 6s and 7s, so those 2 pitches play well off of each other. Welcome aboard, Robinson. It's been a LONG road, one you've navigated successfully. Keep killin' it, my man. #dodgers

Dodgers Daily

17,525 görüntüleme • 10 ay önce

My 2nd trip to Memphis is almost over and I’ve again really enjoyed my time here. So I felt the need to say something, especially after what LeBron said today. This city often gets a very bad rep and I’ve experienced that myself. Players shitting on it, media saying it’s dangerous, people asking me why I’d go here because I wouldn’t be safe, I’ve heard it all. All those people couldn’t be more wrong. This is a city very different from other “typical American cities”. Memphis has it’s own identity, an identity that makes it special. Memphis is a very warm and welcoming city. From the second my first trip started last year I’ve felt welcomed by the people. In the Fedexforum, on the streets, in the airport, in my hotel. My experiences have been nothing but positive. Not once have I felt unsafe or anything remotely close to it, not in my week long stay last year and also not in my almost 2 week long stay this year. Now of course every city has its weaknesses and room for improvement. And that’s where I have to give a lot of credit to the mayor and the people who run this city. Even in the year between my trips there has been clear improvement. The streets are cleaner, there’s roadwork being done and downtown is noticeably more lively compared to last year. Don’t let people who have never been here tell you what and how this city is. Memphis is a unique and beautiful city, Memphis is Memphis. I have truly loved my time here and I can’t wait to keep coming back in the future. But first one more game against Toronto tomorrow.

My 2nd trip to Memphis is almost over and I’ve again really enjoyed my time here. So I felt the need to say something, especially after what LeBron said today. This city often gets a very bad rep and I’ve experienced that myself. Players shitting on it, media saying it’s dangerous, people asking me why I’d go here because I wouldn’t be safe, I’ve heard it all. All those people couldn’t be more wrong. This is a city very different from other “typical American cities”. Memphis has it’s own identity, an identity that makes it special. Memphis is a very warm and welcoming city. From the second my first trip started last year I’ve felt welcomed by the people. In the Fedexforum, on the streets, in the airport, in my hotel. My experiences have been nothing but positive. Not once have I felt unsafe or anything remotely close to it, not in my week long stay last year and also not in my almost 2 week long stay this year. Now of course every city has its weaknesses and room for improvement. And that’s where I have to give a lot of credit to the mayor and the people who run this city. Even in the year between my trips there has been clear improvement. The streets are cleaner, there’s roadwork being done and downtown is noticeably more lively compared to last year. Don’t let people who have never been here tell you what and how this city is. Memphis is a unique and beautiful city, Memphis is Memphis. I have truly loved my time here and I can’t wait to keep coming back in the future. But first one more game against Toronto tomorrow.

Tim

74,949 görüntüleme • 6 ay önce

🚨 MARK MY WORDS: Manchester United are going 15 games unbeaten to start the season. Save this tweet. Bookmark it. Screenshot it. Do whatever you need to do, because on August 3, 2026, I am telling you exactly how the first phase of the Michael Carrick era unfolds. We aren't just going to start well. We are going to absolutely tear up the Premier League and the Champions League. Our first 15 games in all competitions will be a complete and utter slaughter: 14 WINS. 1 DRAW. 0 DEFEATS. The tactical foundation Carrick laid down in pre-season the intense 72-hour recovery training, the filtering out of lazy player power, and the high-pressing system is built perfectly for a historic run. While our rivals are still trying to figure out their squads, Carrick Ball will be operating at 100% maximum efficiency. The media will call it a fluke. Rival fans will claim we had an easy schedule against teams like Hull City. But by the time November hits and we are sitting clear at the top of the table, everyone will finally admit that Manchester United is officially back. The elite Ferguson-era mentality has returned to Carrington. Am I being way too optimistic, or do you see the exact same vision? If you think I’m crazy, drop your prediction below so I can look back at this in three months and laugh

🚨 MARK MY WORDS: Manchester United are going 15 games unbeaten to start the season. Save this tweet. Bookmark it. Screenshot it. Do whatever you need to do, because on August 3, 2026, I am telling you exactly how the first phase of the Michael Carrick era unfolds. We aren't just going to start well. We are going to absolutely tear up the Premier League and the Champions League. Our first 15 games in all competitions will be a complete and utter slaughter: 14 WINS. 1 DRAW. 0 DEFEATS. The tactical foundation Carrick laid down in pre-season the intense 72-hour recovery training, the filtering out of lazy player power, and the high-pressing system is built perfectly for a historic run. While our rivals are still trying to figure out their squads, Carrick Ball will be operating at 100% maximum efficiency. The media will call it a fluke. Rival fans will claim we had an easy schedule against teams like Hull City. But by the time November hits and we are sitting clear at the top of the table, everyone will finally admit that Manchester United is officially back. The elite Ferguson-era mentality has returned to Carrington. Am I being way too optimistic, or do you see the exact same vision? If you think I’m crazy, drop your prediction below so I can look back at this in three months and laugh

Temitayo

43,413 görüntüleme • 1 ay önce

MLP in PyTorch by hand ✍️ ~ 7 steps walkthrough below Goal: fill in every blank in the PyTorch code to build a multi-layer perceptron. 1. Given Let us start with a code template on the left and the network it is supposed to build on the right. Every blank in the code can be worked out from the picture. 2. Linear layer We count: 3 features in, 4 features out. So the weight matrix is 4 by 3. There is an extra column for the biases, which means bias = T. 3. ReLU Let us apply the activation. ReLU crosses out the negatives, so -1 becomes 0. 4. Linear layer The input size is 4, because that is what the previous layer put out. The output size is 2. A 2 by 4 weight matrix, and this time no extra column, so bias = F. 5. ReLU We cross out the negatives again. 6. Linear layer Two features in, five out. A 5 by 2 weight matrix, with a bias column, so bias = T. 7. Sigmoid Let us finish. Sigmoid squashes the raw scores (3, 0, -2, 5, -5) into probabilities between 0 and 1. You have just implemented a three-layer deep neural network by hand. ✍️ == Story == Three years ago I gave this exercise to my students, to connect the code to the math. They found it odd. Every other AI course they were taking lived inside a Jupyter notebook, and here I was handing out paper. Three years later, my colleagues are the ones rushing to move their materials to paper. The exercise has not changed. Paper still asks the one thing a notebook lets you skip: do you actually understand what the code is doing? If you can tell me why the weight matrix is 4 by 3, and why bias is F on the second layer, you understand nn.Linear better than someone who has been copy-pasting it for a year. 💾 Save this post! #AIbyHand #PyTorch #DeepLearning

MLP in PyTorch by hand ✍️ ~ 7 steps walkthrough below Goal: fill in every blank in the PyTorch code to build a multi-layer perceptron. 1. Given Let us start with a code template on the left and the network it is supposed to build on the right. Every blank in the code can be worked out from the picture. 2. Linear layer We count: 3 features in, 4 features out. So the weight matrix is 4 by 3. There is an extra column for the biases, which means bias = T. 3. ReLU Let us apply the activation. ReLU crosses out the negatives, so -1 becomes 0. 4. Linear layer The input size is 4, because that is what the previous layer put out. The output size is 2. A 2 by 4 weight matrix, and this time no extra column, so bias = F. 5. ReLU We cross out the negatives again. 6. Linear layer Two features in, five out. A 5 by 2 weight matrix, with a bias column, so bias = T. 7. Sigmoid Let us finish. Sigmoid squashes the raw scores (3, 0, -2, 5, -5) into probabilities between 0 and 1. You have just implemented a three-layer deep neural network by hand. ✍️ == Story == Three years ago I gave this exercise to my students, to connect the code to the math. They found it odd. Every other AI course they were taking lived inside a Jupyter notebook, and here I was handing out paper. Three years later, my colleagues are the ones rushing to move their materials to paper. The exercise has not changed. Paper still asks the one thing a notebook lets you skip: do you actually understand what the code is doing? If you can tell me why the weight matrix is 4 by 3, and why bias is F on the second layer, you understand nn.Linear better than someone who has been copy-pasting it for a year. 💾 Save this post! #AIbyHand #PyTorch #DeepLearning

Tom Yeh

13,318 görüntüleme • 2 ay önce

Man must be rough living in all jew country, first you’re surrounded by nothing but jews… which is perhaps the most profound piece of evidence that jews are insane,, seriously who would willingly want to live everyday surrounded by jews… you grow up in a place where you wake up in the morning to the sound of kvetching and your siblings bargaining and arguing over who is the actual owner of a toy they stole from some poor Palestinian kid… you’re at risk of being molested 100% of the time by 100% of the population under you hit puberty and then the cycle continues… perhaps that’s why they brainwash their young with holly propaganda to stay in Israel… “if you go to the goyim lands it’s a matter of time before another holocayst happens.. you’re better off here.. where at least you know you’re getting molested but at least it’s but someone you know…. See it’s not so bad” Anyway. Wtf was the point of this post? Oh yeah. Notice the video below. At first glance, it’s preposterous that the rabbi is charging $100 to bless the other Shlomo….. but you think about it maybe that’s just his business model that’s how he makes money fair is fair. The Shlomo agrees to dish out $100 and then he hands over 100 shekels ($30)… basically it’s impossible to discern who’s jewing who. So the next time you wonder why they all don’t just move to a place that they say God promised them, remember, it ain’t so easy, God may have promised it, but comes with one big caveat …it’s filled with other jews

Man must be rough living in all jew country, first you’re surrounded by nothing but jews… which is perhaps the most profound piece of evidence that jews are insane,, seriously who would willingly want to live everyday surrounded by jews… you grow up in a place where you wake up in the morning to the sound of kvetching and your siblings bargaining and arguing over who is the actual owner of a toy they stole from some poor Palestinian kid… you’re at risk of being molested 100% of the time by 100% of the population under you hit puberty and then the cycle continues… perhaps that’s why they brainwash their young with holly propaganda to stay in Israel… “if you go to the goyim lands it’s a matter of time before another holocayst happens.. you’re better off here.. where at least you know you’re getting molested but at least it’s but someone you know…. See it’s not so bad” Anyway. Wtf was the point of this post? Oh yeah. Notice the video below. At first glance, it’s preposterous that the rabbi is charging $100 to bless the other Shlomo….. but you think about it maybe that’s just his business model that’s how he makes money fair is fair. The Shlomo agrees to dish out $100 and then he hands over 100 shekels ($30)… basically it’s impossible to discern who’s jewing who. So the next time you wonder why they all don’t just move to a place that they say God promised them, remember, it ain’t so easy, God may have promised it, but comes with one big caveat …it’s filled with other jews

Copperberg

11,405 görüntüleme • 1 ay önce

Transformer by hand ✍️ ~ 6 steps walkthrough below Open the hood of a transformer and the parts list is overwhelming: embeddings, positional encoding, attention weighting, self-attention, cross-attention, multi-head attention, layer norm, skip connections, softmax, linear, Nx, shifted right, query, key, value, masking. Which of those actually make the car run? Two of them. Attention weighting and the feed-forward network. Everything else is an enhancement to make it run faster and longer, which is how we got from a car to a truck, and to the word "large" in large language model. So I drew and calculated those two parts entirely by hand. Goal: push five features through one transformer block, filling in every cell yourself. 1. Given Five positions of input features, arriving from the previous block. 2. Attention matrix Let us feed all five features to a query-key module (QK) and read back an attention weight matrix, A. The details of that module are a post of their own. 3. Attention weighting We multiply the input features by A to get the attention weighted features, Z. Still five positions. The effect is to combine features *across positions*, horizontally: X1 becomes X1 + X2, X2 becomes X2 + X3, and so on. 4. First layer Let us feed all five weighted features into the first layer of the FFN. Multiply by the weights and biases. This time the combining happens *across feature dimensions*, vertically, and each feature grows from 3 numbers to 4. Note that every position goes through the same weight matrix. That is what "position-wise" means. 5. ReLU We cross out the negatives. They become zeros. 6. Second layer Let us bring it back down: 4 dimensions to 3. The output feeds the next block, which has a completely separate set of parameters, and the whole thing runs again. You have just calculated a transformer block by hand. ✍️ The takeaway: the two parts are doing two different jobs, and neither one alone is enough. Attention mixes *across positions*, so a feature can see its neighbours. The FFN mixes *across feature dimensions*, so each position can think about itself. Horizontal, then vertical. Then that pattern repeats N times, each block with its own separate set of weights. That is the Nx from the list up top, and that is what makes the transformer run. 💾 Save this post! #AIbyHand #Transformers #DeepLearning

Transformer by hand ✍️ ~ 6 steps walkthrough below Open the hood of a transformer and the parts list is overwhelming: embeddings, positional encoding, attention weighting, self-attention, cross-attention, multi-head attention, layer norm, skip connections, softmax, linear, Nx, shifted right, query, key, value, masking. Which of those actually make the car run? Two of them. Attention weighting and the feed-forward network. Everything else is an enhancement to make it run faster and longer, which is how we got from a car to a truck, and to the word "large" in large language model. So I drew and calculated those two parts entirely by hand. Goal: push five features through one transformer block, filling in every cell yourself. 1. Given Five positions of input features, arriving from the previous block. 2. Attention matrix Let us feed all five features to a query-key module (QK) and read back an attention weight matrix, A. The details of that module are a post of their own. 3. Attention weighting We multiply the input features by A to get the attention weighted features, Z. Still five positions. The effect is to combine features across positions, horizontally: X1 becomes X1 + X2, X2 becomes X2 + X3, and so on. 4. First layer Let us feed all five weighted features into the first layer of the FFN. Multiply by the weights and biases. This time the combining happens across feature dimensions, vertically, and each feature grows from 3 numbers to 4. Note that every position goes through the same weight matrix. That is what "position-wise" means. 5. ReLU We cross out the negatives. They become zeros. 6. Second layer Let us bring it back down: 4 dimensions to 3. The output feeds the next block, which has a completely separate set of parameters, and the whole thing runs again. You have just calculated a transformer block by hand. ✍️ The takeaway: the two parts are doing two different jobs, and neither one alone is enough. Attention mixes across positions, so a feature can see its neighbours. The FFN mixes across feature dimensions, so each position can think about itself. Horizontal, then vertical. Then that pattern repeats N times, each block with its own separate set of weights. That is the Nx from the list up top, and that is what makes the transformer run. 💾 Save this post! #AIbyHand #Transformers #DeepLearning

Tom Yeh

26,211 görüntüleme • 2 ay önce

Alastor’s refusal to step in and help with chaos at the hotel at the beginning of Season 2 is absolutely a defensive recoil after the physical (and psychological) damage he experienced after his defeat by Adam. It wasn’t just his angelic injury that caused this dismissive nature, the idea of dying an altruist fully destabilizes the narrative Alastor’s constructed for himself. To die a martyr on behalf of the hotel would retroactively rewrite his afterlife as service rather than sovereignty. When Vaggie and Husk ask for Alastor to help, Alastor withdraws entirely, fully posturing as someone who just enjoys being difficult and doesn’t care. Clearly, if he cannot control the outcome, he will refuse participation entirely. Still, there’s a deeper narrative tension building here — the hotel promising (and delivering) on redemption is in direct contrast to Alastor’s worldview. The hotel is no longer a spectacle for him to be entertained by, but a project that threatens the hierarchies of Hell Alastor had worked to establish himself within even before his human death. Alastor cannot tolerate these positions where he must rely on others or act altruistically. Those positions strip him of narrative agency, of the control and self-preservation critical to his character. This refusal to assist the hotel isn’t just Alastor’s classic brattiness, it is a defensive maneuver designed to restore his sense of autonomy.

Alastor’s refusal to step in and help with chaos at the hotel at the beginning of Season 2 is absolutely a defensive recoil after the physical (and psychological) damage he experienced after his defeat by Adam. It wasn’t just his angelic injury that caused this dismissive nature, the idea of dying an altruist fully destabilizes the narrative Alastor’s constructed for himself. To die a martyr on behalf of the hotel would retroactively rewrite his afterlife as service rather than sovereignty. When Vaggie and Husk ask for Alastor to help, Alastor withdraws entirely, fully posturing as someone who just enjoys being difficult and doesn’t care. Clearly, if he cannot control the outcome, he will refuse participation entirely. Still, there’s a deeper narrative tension building here — the hotel promising (and delivering) on redemption is in direct contrast to Alastor’s worldview. The hotel is no longer a spectacle for him to be entertained by, but a project that threatens the hierarchies of Hell Alastor had worked to establish himself within even before his human death. Alastor cannot tolerate these positions where he must rely on others or act altruistically. Those positions strip him of narrative agency, of the control and self-preservation critical to his character. This refusal to assist the hotel isn’t just Alastor’s classic brattiness, it is a defensive maneuver designed to restore his sense of autonomy.

⊹ ࣪ ˖ ၊၊||၊ ⋆˙alastor fm 𐂂 (🎙)

18,494 görüntüleme • 6 ay önce

Luigi Mangione was Rich. Brian Thompson, the United Healthcare CEO whom Luigi shot in the back, came from modest means. He attended public schools. So do his kids. He went to a Iowa State, a state college. The tuition at Iowa State was about $2800/year for Brian. Brian worked his way up to becoming a CEO. Everything was given to Luigi. His all-boys elite prep school cost about $40k a year. His Ivy League college cost about $75K per year. I am not begrudging Luigi anything. His grandfather worked hard to build his fortune, and Luigi is the benefactor. That's great. It's the hypocrisy that's hard to stomach. These protesters marched over the sidewalk where Brian was shot dead in the back by Luigi. The movement seems to be shifting from Healthcare to killing CEOs to killing the Rich. What an oxymoron this is. Where does it end? Luigi has hired a top dollar, Rich attorney. And Luigi should because he can afford it. But as I said before, if Luigi truly believed his rhetoric, he would have gone with the public defender. These protestors are fighting for a hypocrite, not a hero. Recall that Luigi didn't even have United Healthcare. He didn't lose a family member due to a claim being denied. He received extensive medical treatment on his back and never seemed to want for anything as he traveled the World, post back surgery. Luigi is no revolutionary for a cause. He is a spoiled rich kid who idolizes Ted Kaczynski, has an inflated ego, and who has completely lost his moral compass. #LuigiMangione #UnitedHealthcare #CEOAssassination

Luigi Mangione was Rich. Brian Thompson, the United Healthcare CEO whom Luigi shot in the back, came from modest means. He attended public schools. So do his kids. He went to a Iowa State, a state college. The tuition at Iowa State was about $2800/year for Brian. Brian worked his way up to becoming a CEO. Everything was given to Luigi. His all-boys elite prep school cost about $40k a year. His Ivy League college cost about $75K per year. I am not begrudging Luigi anything. His grandfather worked hard to build his fortune, and Luigi is the benefactor. That's great. It's the hypocrisy that's hard to stomach. These protesters marched over the sidewalk where Brian was shot dead in the back by Luigi. The movement seems to be shifting from Healthcare to killing CEOs to killing the Rich. What an oxymoron this is. Where does it end? Luigi has hired a top dollar, Rich attorney. And Luigi should because he can afford it. But as I said before, if Luigi truly believed his rhetoric, he would have gone with the public defender. These protestors are fighting for a hypocrite, not a hero. Recall that Luigi didn't even have United Healthcare. He didn't lose a family member due to a claim being denied. He received extensive medical treatment on his back and never seemed to want for anything as he traveled the World, post back surgery. Luigi is no revolutionary for a cause. He is a spoiled rich kid who idolizes Ted Kaczynski, has an inflated ego, and who has completely lost his moral compass. #LuigiMangione #UnitedHealthcare #CEOAssassination

Jennifer Coffindaffer

108,208 görüntüleme • 1 yıl önce

The first time people see this place, they assume it is AI-generated. It is not. This wonder is real and it has stood in the sea for thirteen hundred years... It is called Mont Saint-Michel, a small island off the coast of Normandy, where a medieval abbey rises in tiers of stone straight out of the sea, climbing to a single golden spire nearly a hundred metres above the water. There is nothing around it. It stands alone in a vast bay, and the effect is so strange that the eye struggles to accept it: a whole city of stone, floating between the water and the sky. What makes it stranger still is that the sea around it disappears. Mont Saint-Michel sits in a bay with the highest tides in continental Europe. Twice a day the ocean retreats for kilometres, leaving the abbey stranded on a desert of wet sand, and twice a day it floods back in and turns the rock into an island again. The same place is, within hours, surrounded by ocean and surrounded by emptiness. For more than a thousand years, this rhythm has never once stopped. Victor Hugo, who loved it, described it perfectly in 1884: “Mont-Saint-Michel is to France what the Great Pyramid is to Egypt.” I started this newsletter because our past is extraordinary, and fewer and fewer people are showing us how to truly see it. Every week I try to. If that is something you would like to be part of, you can join at the link below, and if you'd like to support my work, a paid subscription is what makes it possible: Thanks for reading.

The first time people see this place, they assume it is AI-generated. It is not. This wonder is real and it has stood in the sea for thirteen hundred years... It is called Mont Saint-Michel, a small island off the coast of Normandy, where a medieval abbey rises in tiers of stone straight out of the sea, climbing to a single golden spire nearly a hundred metres above the water. There is nothing around it. It stands alone in a vast bay, and the effect is so strange that the eye struggles to accept it: a whole city of stone, floating between the water and the sky. What makes it stranger still is that the sea around it disappears. Mont Saint-Michel sits in a bay with the highest tides in continental Europe. Twice a day the ocean retreats for kilometres, leaving the abbey stranded on a desert of wet sand, and twice a day it floods back in and turns the rock into an island again. The same place is, within hours, surrounded by ocean and surrounded by emptiness. For more than a thousand years, this rhythm has never once stopped. Victor Hugo, who loved it, described it perfectly in 1884: “Mont-Saint-Michel is to France what the Great Pyramid is to Egypt.” I started this newsletter because our past is extraordinary, and fewer and fewer people are showing us how to truly see it. Every week I try to. If that is something you would like to be part of, you can join at the link below, and if you'd like to support my work, a paid subscription is what makes it possible: Thanks for reading.

James Lucas

327,887 görüntüleme • 3 ay önce

Let’s be absolutely clear with the facts from today’s “march against the far right” which I attended from the start. 🇬🇧🏴󠁧󠁢󠁥󠁮󠁧󠁿 1. These were left-wing organisations from all over the UK. Every union under the sun was in attendance. National Education Union are definitely indoctrinating children within the teacher and school environment, just look at their presence today. It is very concerning. Watch what they teach your kids. They, as an education union, along with Unite the union: join a union, were also fighting over who got to be at the front of the march today, thanks to two lesbian powerhouses not having it. Comedy gold 😂 2. The second video shows over 100 coaches being used to bus in all these so-called “concerned citizens” on organised transport. They are all part of a left-wing narrative which just doesn’t work anymore, we see exactly what they are doing. That was the entire left-wing loony left in attendance. That’s it. That’s them. This is a major point! We have millions who know what’s right, and the left wing in this country is about to get the shock of their lives in the next elections, swept off the map. That was a panic march. They know it’s coming! 3. There was no working class at this march! Fact! It was all middle and upper class people who attend the usual Palestine marches and Palestine Action events. They hate the Union Jack flag and they hate the working class. This was a stunt by the well-organised unions to pretend the country is behind them which it is not. And that was clear as day to witness today. Also a clear panic march. 4. We don’t have much to worry about. These people are not in the real world. They are privileged and wealthy. And as I said, that’s the crowd, that’s them. They are in total fantasy land, and it’s now time for the adults to get back in charge. 5. Long live England 🏴󠁧󠁢󠁥󠁮󠁧󠁿✝️🇬🇧

Let’s be absolutely clear with the facts from today’s “march against the far right” which I attended from the start. 🇬🇧🏴󠁧󠁢󠁥󠁮󠁧󠁿 1. These were left-wing organisations from all over the UK. Every union under the sun was in attendance. National Education Union are definitely indoctrinating children within the teacher and school environment, just look at their presence today. It is very concerning. Watch what they teach your kids. They, as an education union, along with Unite the union: join a union, were also fighting over who got to be at the front of the march today, thanks to two lesbian powerhouses not having it. Comedy gold 😂 2. The second video shows over 100 coaches being used to bus in all these so-called “concerned citizens” on organised transport. They are all part of a left-wing narrative which just doesn’t work anymore, we see exactly what they are doing. That was the entire left-wing loony left in attendance. That’s it. That’s them. This is a major point! We have millions who know what’s right, and the left wing in this country is about to get the shock of their lives in the next elections, swept off the map. That was a panic march. They know it’s coming! 3. There was no working class at this march! Fact! It was all middle and upper class people who attend the usual Palestine marches and Palestine Action events. They hate the Union Jack flag and they hate the working class. This was a stunt by the well-organised unions to pretend the country is behind them which it is not. And that was clear as day to witness today. Also a clear panic march. 4. We don’t have much to worry about. These people are not in the real world. They are privileged and wealthy. And as I said, that’s the crowd, that’s them. They are in total fantasy land, and it’s now time for the adults to get back in charge. 5. Long live England 🏴󠁧󠁢󠁥󠁮󠁧󠁿✝️🇬🇧

Danny Tommo

84,224 görüntüleme • 6 ay önce

𝑾𝒉𝒆𝒏 𝒊𝒕 𝑹𝑨𝑰𝑵𝑺 𝒊𝒕 𝑷𝑶𝑼𝑹𝑺⛑️ 𝐺𝑢𝑎𝑟𝑑𝑖𝑎𝑛 𝐴𝑛𝑔𝑒𝑙~𝑃𝑎𝑟𝑡 2 😇🙏 After a botched summer assignment to a new collegiate league in Texarkana, TX that saw Grant Scholzen sleeping in his car and no host family provided as promised, Grant came home and waited patiently for a summer league call. The MGF Marshalls (who by the way is one of the best run orgs in the country, more on that in a future post!) called and in his 1st game and 2nd AB with a 1-2 count in a left on left matchup with the shadows in play, Grant took a 92 mph FB off the shoulder and into the right side of his face. The only time in 5 years w/o a face shield on his helmet, it’s the only time he’s been hit above the shoulder. Was watching the game on live stream, and it was hard to tell where it actually hit him? Grant immediately started walking to the dugout and I couldn’t figure out why he was doing that, because he should be either staying at home plate gathering his bearings or going to 1B. What you don’t see is the catcher and pitcher who are close friends of Grant’s growing up facing each other in HS and teammates throughout their CBA summer/fall travel ball experiences, looking on with what just happened? The pitcher and catcher who I’m close with as well, are in shock and the pitcher then walks into the opposing dugout to see if Grant is okay? Grant was okay, but he wasn’t okay?! That helpless feeling again when you’re not there and your son has something almost tragic happen?! I was calling friends who I knew were at the game and they said he had ice on his cheek in the dugout. After the game Grant was going to drive 4 hours home and get home at 3am to attend his cousin’s church service to listen to him speak at his 2 year church missionary farewell. He got home late as I said and woke up yesterday with his face extremely swollen and his eye already black, so we saw his uncle at church, who is a pediatric dentist and he scheduled a CT Scan with an oral surgeon buddy. Grant went in this morning for the scan and to our surprise, NO broken bones. Thank you to the countless people who text and reached out with worry and wanting updates. The baseball circle is small, but big at the same time and is a comforter! Once again his mother Heidi was working overtime to protect her kids, just like my daughter’s car accident 3 weeks ago. When it rains it pours! #SwarmingBees 🐝

𝑾𝒉𝒆𝒏 𝒊𝒕 𝑹𝑨𝑰𝑵𝑺 𝒊𝒕 𝑷𝑶𝑼𝑹𝑺⛑️ 𝐺𝑢𝑎𝑟𝑑𝑖𝑎𝑛 𝐴𝑛𝑔𝑒𝑙~𝑃𝑎𝑟𝑡 2 😇🙏 After a botched summer assignment to a new collegiate league in Texarkana, TX that saw Grant Scholzen sleeping in his car and no host family provided as promised, Grant came home and waited patiently for a summer league call. The MGF Marshalls (who by the way is one of the best run orgs in the country, more on that in a future post!) called and in his 1st game and 2nd AB with a 1-2 count in a left on left matchup with the shadows in play, Grant took a 92 mph FB off the shoulder and into the right side of his face. The only time in 5 years w/o a face shield on his helmet, it’s the only time he’s been hit above the shoulder. Was watching the game on live stream, and it was hard to tell where it actually hit him? Grant immediately started walking to the dugout and I couldn’t figure out why he was doing that, because he should be either staying at home plate gathering his bearings or going to 1B. What you don’t see is the catcher and pitcher who are close friends of Grant’s growing up facing each other in HS and teammates throughout their CBA summer/fall travel ball experiences, looking on with what just happened? The pitcher and catcher who I’m close with as well, are in shock and the pitcher then walks into the opposing dugout to see if Grant is okay? Grant was okay, but he wasn’t okay?! That helpless feeling again when you’re not there and your son has something almost tragic happen?! I was calling friends who I knew were at the game and they said he had ice on his cheek in the dugout. After the game Grant was going to drive 4 hours home and get home at 3am to attend his cousin’s church service to listen to him speak at his 2 year church missionary farewell. He got home late as I said and woke up yesterday with his face extremely swollen and his eye already black, so we saw his uncle at church, who is a pediatric dentist and he scheduled a CT Scan with an oral surgeon buddy. Grant went in this morning for the scan and to our surprise, NO broken bones. Thank you to the countless people who text and reached out with worry and wanting updates. The baseball circle is small, but big at the same time and is a comforter! Once again his mother Heidi was working overtime to protect her kids, just like my daughter’s car accident 3 weeks ago. When it rains it pours! #SwarmingBees 🐝

𝐉𝐞𝐟𝐟 𝐒𝐜𝐡𝐨𝐥𝐳𝐞𝐧

82,230 görüntüleme • 2 yıl önce