Sensitive content

This media may contain sensitive content.

Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

WE BACK TO REGULARLY SCHEULED PROGRAMMING!!! Single pair and bulk orders talk to me, missed a few DMs being in Vegas! Each pair is unique, and upper rubs off to unveil another color! Sz 8, 8.5, 9, 9.5, 10, 10.5, 11 Available Ships today!!! Random RT gets $44.44

40,236 Aufrufe • vor 3 Jahren •via X (Twitter)

10 Kommentare

Profilbild von Follow Me If Youre Reading This
Follow Me If Youre Reading Thisvor 3 Jahren

Winner 🏆: @AlfredNew10 ✔️

Profilbild von typierce19
typierce19vor 3 Jahren

no big footers? 12, 12.5, or 13?

Profilbild von N1NO
N1NOvor 3 Jahren

Need a 13 or 14- paying high, please dm

Profilbild von BDJ206
BDJ206vor 3 Jahren

Looking for 12.5

Profilbild von Brad
Bradvor 3 Jahren

Every time I see these my mind says too expensive but my body says do it.

Profilbild von Felix 🇲🇽🇲🇽
Felix 🇲🇽🇲🇽vor 3 Jahren

👏🏻🙏🏻👍🏻🔥

Profilbild von Ovrnundr
Ovrnundrvor 3 Jahren

📈

Profilbild von Austin Lafferry
Austin Lafferryvor 3 Jahren

8.5 🫡

Profilbild von FBots
FBotsvor 3 Jahren

Price pls inbox size 9

Profilbild von KreativeGenius🌍
KreativeGenius🌍vor 3 Jahren

Super plug shit 🔥

Ähnliche Videos

Happy Final Space Friday! Got another soundtrack piece "One Final Talk" from Andrew Goodwin and now that I'm hearing more music it's hard for me to imagine this novel without the music. It adds so much emotion for our final ride. Me and Seth went over all the notes from our test readers and are starting to fix the spelling errors and grammar stuff. There are a couple page fixes for three moments to land even harder that I'm going to give Daz Daz Tibbles, today! So the good new is there isn't a lot to fix or add. I'm currently combing the novel to find the right spot to add our embedded soundtrack and blending it into the story as best as possible. There will be some adds with this too. I'm also finishing up the back matter to the novel with will include a written goodbye from me and sketches of pages from ideas to ink to color which will be cool. And I'm hoping to put an audio commentary from Daz and I as we talk over each part. I really like this idea so you all can follow along as we talk about it. Now color were getting there, we have to fix Part 3, Part 4 (we fixed the inking so we need to fix the coloring), were coloring Part 5, Part 7 still needs to be colored and then Part 8. I'm playing with the idea that part of PART 8 will be black and white and then the color comes back and a killer moment (it's apart of the story) but this has been an idea I was playing around with since the beginning and it looks REALLY cool. Anyways lots of fine tuning but it's coming together! Exciting! -Olan PRE-ORDER:

Final Space

27,064 Aufrufe • vor 1 Jahr

[CLIP] by Hand ✍️ The CLIP (Contrastive Language–Image Pre-training) model, a groundbreaking work by OpenAI, redefines the intersection of computer vision and natural language processing. It is the basis of all the multi-modal foundation models we see today. How does CLIP work? Goal: 🟨 Learn a shared embedding space for text and image [1] Given ↳ A mini batch of 3 text-image pairs ↳ OpenAI used 400 million text-image pairs to train its original CLIP model. Process 1st pair: "big table" [2] 🟪 Text → 2 Vectors (3D) ↳ Look up word embedding vectors using word2vec. [3] 🟩 Image → 2 Vectors (4D) ↳ Divide the image into two patches. ↳ Flatten each patch [4] Process other pairs ↳ Repeat [2]-[3] [5] 🟪 Text Encoder & 🟩 Image Encoder ↳ Encode input vectors into feature vectors ↳ Here, both encoders are simple one layer perceptron (linear + ReLU) ↳ In practice, the encoders are usually transformer models. [6] 🟪 🟩 Mean Pooling: 2 → 1 vector ↳ Average 2 feature vectors into a single vector by averaging across the columns ↳ The goal is to have one vector to represent each image or text [7] 🟪 🟩 -> 🟨 Projection ↳ Note that the text and image feature vectors from the encoders have different dimensions (3D vs. 4D). ↳ Use a linear layer to project image and text vectors to a 2D shared embedding space. 🏋️ Contrastive Pre-training 🏋️ [8] Prepare for MatMul ↳ Copy text vectors (T1,T2,T3) ↳ Copy the transpose of image vectors (I1,I2,I3) ↳ They are all in the 2D shared embedding space. [9] 🟦 MatMul ↳ Multiply T and I matrices. ↳ This is equivalent to taking dot product between every pair of image and text vectors. ↳ The purpose is to use dot product to estimate the similarity between a pair of image-text. [10] 🟦 Softmax: e^x ↳ Raise e to the power of the number in each cell ↳ To simplify hand calculation, we approximate e^□ with 3^□. [11] 🟦 Softmax: ∑ ↳ Sum each row for 🟩 image→🟪 text ↳ Sum each column for 🟪 text→ 🟩 image [12] 🟦 Softmax: 1 / sum ↳ Divide each element by the column sum to obtain a similarity matrix for 🟪 text→🟩 image ↳ Divide each element by the row sum to obtain a similarity matrix for 🟩 image→🟪 text [13] 🟥 Loss Gradients ↳ The "Targets" for the similarity matrices are Identity Matrices. ↳ Why? If I and T come from the same pair (i=j), we want the highest value, which is 1, and 0 otherwise. ↳ Apply the simple equation of [Similarity - Target] to compute gradients of for both directions. ↳ Why so simple? Because when Softmax and Cross-Entropy Loss are used together, the math magically works out that way. ↳ These gradients kick off the backpropagation process to update weights and biases of the encoders and projection layers (red borders).

Tom Yeh

67,834 Aufrufe • vor 2 Jahren