Lesson 22: Deep Learning Foundations to Stable Diffusion

preview_player
Показать описание
Oops I say it's "Lesson 21" at the start of the video -- but actually this is lesson 22!

The lesson covers various sampling techniques, such as the Euler sampler, Ancestral Euler sampler, and Heuns method. Jeremy explains the concepts behind these methods and demonstrates how they can be used to improve the sampling process. He emphasizes the importance of understanding the underlying concepts and techniques in research papers and demonstrates how these can be applied to improve model performance.

00:00 - Intro
00:30 - Cosine Schedule (22_cosine)
06:05 - Sampling
09:37 - Summary / Notation
10:42 - Predicting the noise level of noisy Fashion MNIST images
12:57 - Why .logit() when predicting alpha bar t
14:50 - Random baseline
16:40 - mse_loss why .flatten()
17:30 - Model & results
19:03 - Why are we trying to predict the noise level?
20:10 - Training diffusion without t - first attempt
22:58 - Why it isn’t working?
27:02 - Debugging (summary)
29:29 - Bug in ddpm - paper that cast some light on the issue
38:40 - Karras (Elucidating the Design Space of Diffusion - Based Generative Models)
49:47 - Picture of target images
52:48 - Scaling problem - (scalings)
59:42 - Training and predictions of modified model
1:03:49 - Sampling
1:06:05 - Sampling: Problems of composition
1:07:40 - Sampling: Rationale for rho selection
1:09:40 - Sampling: Denosing
1:15:26 - Sampling: Heun’s method fid: 0.972
1:19:00 - Sampling: LMS sampler
1:20:00 - Kerras Summary
1:23:00 - Comparison of different approaches
1:25:00 - Next lessons

Timestamps thanks to Piotr Czapla. Transcript thanks to Francisco Mussari
  1. Jeremy Howard
  2. deep learning
  3. fastai
Рекомендации по теме
Lesson 22: Deep 1:26:44

Lesson 22: Deep Learning Foundations to Stable Diffusion

Lesson 21: Deep 1:55:47

Lesson 21: Deep Learning Foundations to Stable Diffusion

Lesson 9: Deep 2:15:16

Lesson 9: Deep Learning Foundations to Stable Diffusion

Lesson 24: Deep 1:55:48

Lesson 24: Deep Learning Foundations to Stable Diffusion

Lesson 18: Deep 2:05:57

Lesson 18: Deep Learning Foundations to Stable Diffusion

Lesson 10: Deep 1:49:14

Lesson 10: Deep Learning Foundations to Stable Diffusion, 2022

Deep Learning | 0:05:52

Deep Learning | What is Deep Learning? | Deep Learning Tutorial For Beginners | 2023 | Simplilearn

Lung inflation in 0:00:16

Lung inflation in Science Lesson #science #teacher #biology

What Is AI? 0:05:28

What Is AI? | Artificial Intelligence | What is Artificial Intelligence? | AI In 5 Mins |Simplilearn

Simple explanation of 0:23:54

Simple explanation of convolutional neural network | Deep Learning Tutorial 23 (Tensorflow & Pyt...

Neural Network In 0:05:45

Neural Network In 5 Minutes | What Is A Neural Network? | How Neural Networks Work | Simplilearn

What is BERT? 0:23:03

What is BERT? | Deep Learning Tutorial 46 (Tensorflow, Keras & Python)

What is Machine 0:08:23

What is Machine Learning?

Tutorial 22-Univariate, Bivariate 0:13:11

Tutorial 22-Univariate, Bivariate and Multivariate Analysis- Part1 (EDA)-Data Science

All in One 11:56:33

All in One Artificial Intelligence (AI) Full Course 2024 | AI Tutorial for Beginners

Data Science In 0:04:38

Data Science In 5 Minutes | Data Science For Beginners | What Is Data Science? | Simplilearn

Lesson '0': Practical 0:55:31

Lesson '0': Practical Deep Learning for Coders (fast.ai)

What is Deep 0:38:06

What is Deep Learning? | Introduction to Deep Learning | Deep Learning Tutorial | Simplilearn

Machine Learning Tutorial 0:06:51

Machine Learning Tutorial Python -1: What is Machine Learning?

Derivatives | Deep 0:12:35

Derivatives | Deep Learning Tutorial 9 (Tensorflow Tutorial, Keras & Python)

Lesson 12: Deep 2:17:55

Lesson 12: Deep Learning Part 2 2018 - Generative Adversarial Networks (GANs)

Natural Language Processing 0:05:29

Natural Language Processing In 5 Minutes | What Is NLP And How Does It Work? | Simplilearn

Machine Learning | 0:07:52

Machine Learning | What Is Machine Learning? | Introduction To Machine Learning | 2024 | Simplilearn

Golang Tutorial #22 0:12:19

Golang Tutorial #22 - Interfaces

Комментарии
Автор

I have a preliminary chapters. Let's see if YouTube let me add them so that it is easier to improve on them.

Chapters
00:00 - Intro
00:30 - Cosine Schedule (22_cosine)
06:05 - Sampling
09:37 - Summary / Notation
10:42 - Pedicting the noise level of noisy Fashion MNIST images (22_noise-pred)
12:57 - Why .logit() when predicting alpha bar t
14:50 - Random baseline
16:40 - mse_loss why .flatten()
17:30 - Model & results
19:03 - Why are we trying to predict the noise level?
20:10 - Training diffiusion without t - first attempt
22:58 - Why it isn’t working?
27:02 - Debugging (summmary)
29:29 - Bug in ddpm - paper that cast some light on the issue
38:40 - Kerras (Elucidating the Design Space of Diffusion - Based Generative Models)
49:47 - Picture of target images
52:48 - Scaling problem - (scalings)
59:42 - Training and predictions of modified model
1:03:49 - Sampling
1:06:05 - Sampling: Problems of composition
1:07:40 - Sampling: Rationale for rho selection
1:09:40 - Sampling: Denosing
1:15:26 - Sampling: Heun’s method fid: 0.972
1:19:00 - Sampling: LMS sampler
1:20:00 - Kerras Summary
1:23:00 - Comparison of different approaches
1:25:00 - Next lessons

piotr.czapla