Completed
Introduction speech
Class Central Classrooms beta
YouTube playlists curated by Class Central.
Classroom Contents
How to Do Stable Diffusion LORA Training by Using Web UI on Different Models
Automatically move to the next video in the Classroom when playback concludes
- 1 Introduction speech
- 2 How to install the LoRA extension to the Stable Diffusion Web UI
- 3 Preparation of training set images by properly sized cropping
- 4 How to crop images using Paint .NET, an open-source image editing software
- 5 What is Low-Rank Adaptation LoRA
- 6 Starting preparation for training using the DreamBooth tab - LoRA
- 7 Explanation of all training parameters, settings, and options
- 8 How many training steps equal one epoch
- 9 Save checkpoints frequency
- 10 Save a preview of training images after certain steps or epochs
- 11 What is batch size in training settings
- 12 Where to set LoRA training in SD Web UI
- 13 Explanation of Concepts tab in training section of SD Web UI
- 14 How to set the path for training images
- 15 Classification Dataset Directory
- 16 Training prompt - how to set what to teach the model
- 17 What is Class and Sample Image Prompt in SD training
- 18 What is Image Generation settings and why we need classification image generation in SD training
- 19 Starting the training process
- 20 How and why to tune your Class Prompt generating generic training images
- 21 Why we generate regularization generic images by class prompt
- 22 Recap of the setting up process for training parameters, options, and settings
- 23 How much GPU, CPU, and RAM the class regularization image generation uses
- 24 Training process starts after class image generation completed
- 25 Displaying the generated class regularization images folder for SD 2.1
- 26 The speed of the training process - how many seconds per iteration on an RTX 3060 GPU
- 27 Where LoRA training checkpoints weights are saved
- 28 Where training preview images are saved and our first training preview image
- 29 When we will decide to stop training
- 30 How to resume training after training has crashed or you close it down
- 31 Lifetime vs. session training steps
- 32 After 30 epochs, resembling images start to appear in the preview folder
- 33 The command line printed messages are incorrect in some cases
- 34 Training step speed, a certain number of seconds per iteration IT
- 35 How I'm picking a checkpoint to generate a full model .ckpt file
- 36 How to generate a full model .ckpt file from a LoRA checkpoint .pt file
- 37 Generated/saved file name is incorrect, but it is generated from the correct selected .pt file
- 38 Doing inference generating new images using the text2img tab with our newly trained and generated model
- 39 The results of SD 2.1 Version 768 pixel model after training with the LoRA method and teaching a human face
- 40 Setting up the training parameters/options for SD version 1.5 this time
- 41 Re-generating class regularization images since SD 1.5 uses 512 pixel resolution
- 42 Displaying the generated class regularization images folder for SD 1.5
- 43 Training of Stable Diffusion 1.5 using the LoRA methodology and teaching a face has been completed and the results are displayed
- 44 The inference text2img results with SD 1.5 training
- 45 You have to do more inference with LoRA since it has less precision than DreamBooth
- 46 How to give more attention/emphasis to certain keywords in the SD Web UI
- 47 How to generate more than 100 images
- 48 How to check PNG info to see used prompts and settings
- 49 How to upscale using AI models
- 50 Fixing face image quality, especially eyes, with GFPGAN visibility
- 51 How to batch post-process
- 52 Where batch-generated images are saved
