Commit
•
a20d3af
1
Parent(s):
093cacd
Upload PIXART-a-First-Open-Source-Rival-to-Midjourney-Better-Than-Stable-Diffusion-SDXL-Full-Tutorial.md
Browse files
PIXART-a-First-Open-Source-Rival-to-Midjourney-Better-Than-Stable-Diffusion-SDXL-Full-Tutorial.md
ADDED
|
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
+
[](https://discord.com/servers/software-engineering-courses-secourses-772774097734074388) [](https://hits.seeyoufarm.com)
|
| 3 |
+
|
| 4 |
+
[](https://www.patreon.com/SECourses) [](https://www.buymeacoffee.com/DrFurkan) [](https://medium.com/@furkangozukara) [](https://civitai.com/user/SECourses/articles) [](https://www.deviantart.com/monstermmorpg)
|
| 5 |
+
|
| 6 |
+
[](https://www.youtube.com/SECourses) [](https://www.linkedin.com/in/furkangozukara/) [](https://www.udemy.com/course/stable-diffusion-dreambooth-lora-zero-to-hero/?referralCode=E327407C9BDF0CEA8156) [](https://twitter.com/GozukaraFurkan)
|
| 7 |
+
|
| 8 |
+
|
| 9 |
+
[**PIXART-α : First Open Source Rival to Midjourney - Better Than Stable Diffusion SDXL - Full Tutorial**](https://youtu.be/ZiUXf_idIR4)
|
| 10 |
+
|
| 11 |
+
[](https://youtu.be/ZiUXf_idIR4)
|
| 12 |
+
|
| 13 |
+
Introduction to the new PixArt-α (PixArt Alpha) text to image model which is for real better than Stable Diffusion models even from SDXL. PixArt-α is close to the Midjourney level meanwhile being open source and supporting full fine tuning and DreamBooth training. In this tutorial I show how to install and use PixArt-α both locally and on a cloud service RunPod with automatic installers and step by step guidance.
|
| 14 |
+
|
| 15 |
+
The link to download resources ⤵️
|
| 16 |
+
https://www.patreon.com/posts/pixart-alpha-for-93614549
|
| 17 |
+
|
| 18 |
+
Stable Diffusion GitHub repository ⤵️
|
| 19 |
+
https://github.com/FurkanGozukara/Stable-Diffusion
|
| 20 |
+
|
| 21 |
+
SECourses Discord To Get Full Support ⤵️
|
| 22 |
+
https://discord.com/servers/software-engineering-courses-secourses-772774097734074388
|
| 23 |
+
|
| 24 |
+
#PixArt #StableDiffusion #SDXL
|
| 25 |
+
|
| 26 |
+
0:00 Introduction to PixArt-α: Fast Training of Diffusion Transformer for Photorealistic Text-to-Image Synthesis and the tutorial content
|
| 27 |
+
* 2:38 What are the requirements to follow this tutorial and install PixArt Alpha
|
| 28 |
+
* 3:05 How to install PixArt Alpha on your machine and start using it
|
| 29 |
+
* 3:59 Where Hugging Face models are downloaded by default and how to change this default cache folder
|
| 30 |
+
* 5:44 How to return back to using default Hugging Face cache folder
|
| 31 |
+
* 6:08 How to fix corrupted files error during installation
|
| 32 |
+
* 6:29 How to start PixArt Web APP after installation has been completed
|
| 33 |
+
* 7:24 How to use PixArt Web APP and its features
|
| 34 |
+
* 7:59 Comparing a dragon prompt with SDXL base version
|
| 35 |
+
* 8:14 How to use provided styles csv file
|
| 36 |
+
* 8:40 How to start Automatic1111 SD Web UI on your second GPU
|
| 37 |
+
* 8:50 Where the PixArt Web APP generated images are saved
|
| 38 |
+
* 9:30 How to set parameters in your Automatic1111 SD Web UI to generate high quality images
|
| 39 |
+
* 9:49 PixArt generated image vs SDXL generated image for same simple prompt
|
| 40 |
+
* 10:15 Anime style same prompt comparison
|
| 41 |
+
* 10:55 One another strong aspect of the PixArt Alpha model
|
| 42 |
+
* 11:29 Fantasy art style comparison of SDXL vs PixArt-α
|
| 43 |
+
* 11:52 3D style comparison of SDXL vs PixArt-α
|
| 44 |
+
* 12:16 Manga style image generation comparison between SDXL vs PixArt-α
|
| 45 |
+
* 12:44 Comparing PixArt vs SDXL vs Midjourney with same prompt
|
| 46 |
+
* 13:41 How to use LLaVA for captioning and obtaining prompt ideas and generating more amazing images
|
| 47 |
+
* 16:12 Comparison of PixArt vs SDXL prompt following in details
|
| 48 |
+
* 17:29 Getting prompt idea from ChatGPT and comparing SDXL and PixArt prompt following
|
| 49 |
+
* 19:46 PixArt owns hard the SDXL with this new detailed prompt
|
| 50 |
+
* 22:00 How to install PixArt on a RunPod pod / machine
|
| 51 |
+
* 23:54 How to set default Hugging Face cache folder on RunPod / Linux machines
|
| 52 |
+
* 25:05 How to understand RunPod machine / pod is not working correctly and fix it
|
| 53 |
+
* 26:00 How to properly delete files / folders on RunPod machines / pods
|
| 54 |
+
* 26:51 How to connect and use PixArt web UI on a RunPod machine after it was started
|
| 55 |
+
* 28:20 How to download all of the generated images on RunPod with runpodctl very fast
|
| 56 |
+
|
| 57 |
+
The paper introduces PIXART-α, a Transformer-based text-to-image (T2I) diffusion model designed to significantly lower training costs while maintaining high image generation quality, competitive with leading models like Imagen and Midjourney. It achieves high-resolution synthesis up to 1024x1024 pixels at reduced training costs.
|
| 58 |
+
|
| 59 |
+
Key Innovations:
|
| 60 |
+
|
| 61 |
+
Training Strategy Decomposition: The process is divided into three steps focusing on pixel dependency, text-image alignment, and image aesthetic quality. This approach reduces learning costs by starting with a low-cost class-condition model and then pretraining and fine-tuning on data rich in information density and aesthetic quality.
|
| 62 |
+
|
| 63 |
+
Efficient T2I Transformer: Built on the Diffusion Transformer (DiT) framework, it includes cross-attention modules for text conditions and streamlines computation. A reparameterization technique enables loading parameters from class-condition models, leveraging prior knowledge from ImageNet, thus accelerating training.
|
| 64 |
+
|
| 65 |
+
High-informative Data: To overcome deficiencies in existing text-image datasets, the paper introduces an auto-labeling pipeline using a vision-language model (LLaVA) to generate captions on the SAM dataset. This dataset is selected for its diverse collection of objects, aiding in creating high-information-density text-image pairs for efficient alignment learning.
|
| 66 |
+
|
| 67 |
+
Image Quality: The model excels in image quality, artistry, and semantic control, surpassing existing models in user studies and benchmarks.
|
| 68 |
+
|
| 69 |
+
Broader Implications: The paper suggests that PIXART-α's approach allows individual researchers and startups to develop high-quality T2I models at lower costs, potentially democratizing access to advanced AI-generated content.
|
| 70 |
+
|
| 71 |
+
The paper concludes with the hope that PIXART-α will inspire the AIGC community and enable more entities to build their own generative models efficiently and affordably.
|