Recent content on Easy Diffusion
Cut down the startup time of Easy Diffusion by 4 seconds on warm start, and atleast 10 seconds on cold start. Removed a couple of unnecessary imports and ran a few things in parallel. Released this to the beta channel. Implemented and tested another change (re-ordered startup process) that will improve the startup time by another 4 seconds. Will release that next week, because I want to watch for…
tl;dr - Right now, what annoys me the most with Easy Diffusion is the startup time and friction, and how Easy Diffusion isn’t actually “easy” anymore - the default model (SD 1.5) is bad, users take too long to get productive, and new models require a lot of configuration. That’s what v4 aims to bring under control. Easy Diffusion v4 needs to ship soon, this has gone on long…
Added support for Krea2 in Easy Diffusion v4, in the beta channel.
Upgraded stable-diffusion.cpp in Easy Diffusion (and sdkit) to the latest version. This adds support for two new models: Lens and PiD , along with a bunch of bug fixes (for e.g. Chroma rendering has been fixed). It also brings back --diffusion-fa by default, which speeds up rendering. And Chroma seems to be rendering faster than before.
Just completed another round of fixing support issues in Easy Diffusion. Today’s evidence continues to convince me to move Easy Diffusion away from the Python ecosystem for AI inference on end-user PCs. And definitely to stay away from conda, it’s so leaky. Both are excellent in their own right (especially for training and research), but I don’t think they make sense for end-user…
Released Easy Diffusion v4.3 (which updates to sdkit v3.2). This adds support for Ernie Image (and Ernie Turbo), as well as improved support for Anima models. It also includes a bunch of bug fixes in the rendering engine (i.e. stable-diffusion.cpp), and a few community-contributed bug fixes to the UI.
Development update for Easy Diffusion: the beta branch has been merged into main , so this releases v3.5 (webui) and v4 to everyone. This shouldn’t affect existing users who’re on the main branch, i.e. people using the v3 engine will continue doing so. The two engines (v3.5 and v4) are marked as optional, so new users will continue to get and use v3 by default. The main purpose of this…
Got Easy Diffusion v4 working on Apple and Intel Macs. The performance difference ratio (vs ED v3) is similar to the ratio on Windows (with CUDA) and other deployment targets. So that indicates optimization opportunities in sd.cpp. It’s currently about 1.5x slower than diffusers-based Stable Diffusion. In other news, easyinstaller is also out with its first release, which means that Easy…
Started the long-pending rewrite of Easy Diffusion’s server code. v4 intends to replace the Python (and PyTorch) based server with a simple C++ version. The reason for rewriting the server in C++ is to achieve sub-second startup time for the UI, and to reduce the download size (won’t need to distribute Python along with Easy Diffusion) or mess with conda/venv etc. And it’s also…
For Z-Image, the performance of the stock version of chromaForge is poorer than sd.cpp’s. Mainly because chromaForge isn’t able to run the smaller gguf quantized models that sd.cpp is able to run (chromaForge fails with the errors that I was fixing yesterday). If I really want to push through with this, it would be good to fix the remaining issues with gguf models in chromaForge. Only…
Worked on fixing Z-Image support in ED’s fork of chromaForge (a fork of Forge WebUI). Fixed a number of integration issues. It’s now crashing on a matrix multiplication error, which looks like an incorrectly transposed matrix (mostly due to reading the weights in the wrong order). I’ll try to install a stock version of chromaForge to see its raw performance with Z-Image (and…
Collecting the worklog over the past few weeks. Enabled Flash-Attention and CPU offloading by default in sdkit3 (i.e. Easy Diffusion v4). Added optional VAE tiling (and VAE tile size configuration) via config.yaml in Easy Diffusion v4. Created Easy Diffusion’s fork of Forge WebUI, in order to apply the patches required to run with ED. And also to try adding new features like Z-Image (which…
The new engine that’ll power Easy Diffusion’s upcoming v4 release (i.e. sdkit3 ) has now been integrated into Easy Diffusion. It’s available to test by selecting v4 engine in the Settings tab (after enabling Beta ). Please press Save and restart Easy Diffusion after selecting this. It uses stable-diffusion.cpp and ggml under-the-hood, and produces optimized, lightweight builds…
Managed to get stable-diffusion.cpp integrated into sdkit v3 and Easy Diffusion. sdkit v3 wraps stable-diffusion.cpp with an API server. For now, the API server exposes an API compatible with Forge WebUI. This saves me time, and allows Easy Diffusion to work out-of-the-box with the new C++ based sdkit. It compiles and runs quite well. Ran it with Easy Diffusion’s UI. Tested with Vulkan and…
Following up to the previous post on sdkit v3’s design: The initial experiments with generating ggml from onnx models were promising, and it looks like a fairly solid path forward. It produces numerically-identical results, and there’s a clear path to reach performance-parity with stable-diffusion.cpp with a few basic optimizations (since both will eventually generate the same…
Successfully compiled the VAE of Stable Diffusion 1.5 using graph-compiler . The compiled model is terribly slow because I haven’t written any performance optimizations, and it (conservatively) converts a lot of intermediate tensors to contiguous copies. But we don’t need any clever optimizations to get to decent performance, just basic ones. It’s pretty exciting because I was…
PolyBlocks is another interesting ML compiler, written using MLIR. It’s a startup incubated in IISc Bangalore, run by someone (Uday Bondhugula) who co-authored a paper on compiler optimizations for GPGPUs back in 2008 (17 years ago)! Some of the compiler passes to keep in mind: fusion tiling use hardware acceleration (like tensor cores) constant folding perform redundant computation to avoid…
Wrote a simple script to convert ONNX to GGML. It auto-generates C++ code that calls the corresponding ggml functions (for each ONNX operator). This file can then be compiled and run like a normal C++ ggml program, and will produce the same results as the original model in PyTorch. The generated file can work on multiple backends: CPU, CUDA, ROCm, Vulkan, Metal etc, by providing the correct…
Following up to the deep-dive on ML compilers : sdkit v3 won’t use general-purpose ML compilers. They aren’t yet ready for sdkit’s target platforms, and need a lot of work (well beyond sdkit v3’s scope). But I’m quite certain that sdkit v4 will use them, and sdkit v3 will start making steps in that direction. For sdkit v3, I see two possible paths: Use an array of…
This post concludes (for now) my ongoing deep-dive into ML compilers , while researching for sdkit v3 . I’ve linked (at the end) to some of the papers that I read related to graph execution on GPUs. Some final takeaways: ML compilers might break CUDA’s moat (and fix AMD’s ROCm support). A single compiler is unlikely to fit every scenario. The scheduler needs to be grounded in…
A possible intuition for understanding GPU memory hierarchy (and the performance penalty for data transfer between various layers) is to think of it like a manufacturing logistics problem: CPU (host) to GPU (device) is like travelling overnight between two cities. The CPU city is like the “headquarters”, and contains a mega-sized warehouse of parts (think football field sizes), also…
Good post on using MLIR for compiling ML models to GPUs. It gives a good broad overview of a GPU architecture, and how MLIR fits into that. The overall series looks pretty interesting too! Making a note here for future reference - https://www.stephendiehl.com/posts/mlir_gpu/
Wrote a fresh implementation of most of the popular samplers and schedulers used for image generation (Stable Diffusion and Flux) at https://github.com/cmdr2/samplers.cpp . A few other schedulers (like Align Your Steps ) have been left out for now, but are pretty easy to implement. It’s still work-in-progress, and is not ready for public use. The algorithmic port has been completed, and the…
Some notes on machine-learning compilers, gathered while researching tech for Easy Diffusion’s next engine (i.e. sdkit v3). For context, see the design constraints of the new engine. tl;dr summary The current state is: Vendor-specific compilers are the only performant options on consumer GPUs right now. For e.g. TensorRT-RTX for NVIDIA, MiGraphX for AMD, OpenVINO for Intel. Cross-vendor…
The design constraints for Easy Diffusion’s next engine (i.e. sdkit v3) are: Lean: Install size of < 200 MB uncompressed (excluding models). Fast: Performance within 10% of the best-possible speed on that GPU for that model. Capable: Supports Stable Diffusion 1.x, 2.x, 3.x, XL, Flux, Chroma, ControlNet, LORA, Embedding, VAE. Supports loading custom model weights (from civitai etc), and…
Cleared the backlog of stale issues on ED’s github repo. This brought down the number of open issues from ~350 to 74. A number of those suggestions and issues are already being tracked on my task board . The others had either been fixed, or were really old (i.e. not relevant to reply anymore). While I’d have genuinely wanted to solve all of those unresolved issues, I was on a break…
Experimented with TensorRT-RTX (a new library offered by NVIDIA). The first step was a tiny toy model, just to get the build and test setup working. The reference model in PyTorch: import torch import torch.nn as nn class TinyCNN (nn . Module): def __init__ (self): super() . __init__ () self . conv = nn . Conv2d( 3 , 8 , 3 , stride = 1 , padding = 1 ) self . relu = nn . ReLU() self . pool = nn .…
Development update for Easy Diffusion - It’s chugging along in starts and stops. Broadly, there are three tracks: Maintenance: The past few months have seen increased support for AMD, Intel and integrated GPUs. This includes AMD on Windows. Added support for the new AMD 9060/9070 cards last week, and the new NVIDIA 50xx cards in March. Flux to the main branch / release v3.5 to stable: Right…
Upgraded the default version of Easy Diffusion to Python 3.9. Newer versions of torch don’t support Python 3.8, so this became urgent after the release of NVIDIA’s 50xx series GPUs. I choose 3.9 as a temporary fix (instead of a newer Python version), since it had the least amount of package conflicts. The future direction of Easy Diffusion’s backend is unclear right now - there…
Easy Diffusion (and sdkit ) now also support AMD on Windows automatically (using DirectML), thanks to integrating with torchruntime . It also supports integrated GPUs (Intel and AMD) on Windows, making Easy Diffusion faster on PCs without dedicated graphics cards.
Spent the last week or two getting torchruntime fully integrated into Easy Diffusion, and making sure that it handles all the edge-cases. Easy Diffusion now uses torchruntime to automatically install the best-possible version of torch (on the users’ computer) and support a wider variety of GPUs (as well as older GPUs). And it uses a GPU-agnostic device API, so Easy Diffusion will…
Continued to test and fix issues in sdkit, after the change to support DirectML. The change is fairly intrusive, since it removes direct references to torch.cuda with a layer of abstraction. Fixed a few regressions, and it now passes all the regression tests for CPU and CUDA support (i.e. existing users). Will test for DirectML next, although it will fail (with out-of-memory) for anything but the…
Worked on adding support for DirectML in sdkit. This allows AMD GPUs and Integrated GPUs to generate images on Windows. DirectML seems like it’s really inefficient with memory though. So for now it only manages to generate images using SD 1.5. XL and larger models fail to generate, even though I have a 12 GB of VRAM in my graphics card.
Continued from Part 1 . Spent a few days figuring out how to compile binary wheels of PyTorch and include all the necessary libraries (ROCm libs or CUDA libs). tl;dr - In Part 2, the compiled PyTorch wheels now include the required libraries (including ROCm). But this isn’t over yet. Torch starts now, but adding two numbers with it produces garbage values (on the GPU). There’s probably…
Continued in Part 2 , where I figured out how to include the required libraries in the wheel. Spent all of yesterday trying to compile pytorch with the compile-time PYTORCH_ROCM_ARCH=gfx803 environment variable. tl;dr - In Part 1, I compiled wheels for PyTorch with ROCm, in order to add support for older AMD cards like RX 480. I managed to compile the wheels, but the wheel doesn’t include…
Spent the last few days writing torchruntime , which will automatically install the correct torch distribution based on the user’s OS and graphics card. This package was written by extracting this logic out of Easy Diffusion, and refactoring it into a cleaner implementation (with tests). It can be installed (on Win/Linux/Mac) using pip install torchruntime . The main intention is that…
Spent most of the day doing some support work for Easy Diffusion, and experimenting with torch-directml for AMD support on Windows. From the initial experiments, torch-directml seems to work properly with Easy Diffusion. I ran it on my NVIDIA card, and another user ran it on their AMD Radeon RX 7700 XT. It’s 7-10x faster than the CPU, so looks promising. It’s 2x slower than CUDA on my…
Spent a few days prototyping a UI for Easy Diffusion v4. Files are at this repo . The main focus was to get a simple but pluggable UI, that was backed by a reactive data model, and to allow splitting the codebase into individual components (with their own files). And require only a text editor and a browser to develop, i.e. no compilation or nodejs-based developer experiences. I really want…
Notes on two directions for ED4’s UI that I’m unlikely to continue on. One is to start a desktop app with a full-screen webview (for the app UI). The other is writing the tabbed browser-like shell of ED4 in a compiled language (like Go or C++) and loading the contents of the tabs as regular webpages (by using webviews). So it would load URLs like http://localhost:9000/ui/image_editor…
Worked on a few UI design ideas for Easy Diffusion v4. I’ve uploaded the work-in-progress mockups at https://github.com/easydiffusion/files . So far, I’ve mocked out the design for the outer skeleton. That is, the new tabbed interface, the status bar, and the unified main menu. I also worked on how they would look like on mobile devices. It gives me a rough idea of the Vue components…
Spent some more time on the v4 experiments for Easy Diffusion (i.e. C++ based, fast-startup, lightweight). stable-diffusion.cpp is missing a few features, which will be necessary for Easy Diffusion’s typical workflow. I wasn’t keen on forking stable-diffusion.cpp, but it’s probably faster to work on a fork for now. For now, I’ve added live preview and per-step progress…
Spent a few days getting a C++ based version of Easy Diffusion working, using stable-diffusion.cpp. I’m working with a fork of stable-diffusion.cpp here , to add a few changes like per-step callbacks, live image previews etc. It doesn’t have a UI yet, and currently hardcodes a model path. It exposes a RESTful API server (written using the Crow C++ library), and uses a simple task…
tl;dr - Today, I worked on using stable-diffusion.cpp in a simple C++ program. As a linked library, as well as compiling sd.cpp from scratch (with and without CUDA). The intent was to get a tiny and fast-starting executable UI for Stable Diffusion working. Also, ChatGPT is very helpful! Part 1: Using sd.cpp as a library First, I tried calling the stable-diffusion.cpp library from a simple C++…
tl;dr : Explored a possible optimization for Flux with diffusers when using enable_sequential_cpu_offload() . It did not work. While trying to use Flux (nearly 22 GB of weights) with diffusers on a 12 GB graphics card, I noticed that it barely used any GPU memory when using enable_sequential_cpu_offload() . And it was super slow. It turns out that the largest module in Flux’s transformer…