From f47b9c7e819ea5f84648cee4be87ea583e6b0756 Mon Sep 17 00:00:00 2001 From: Vladimir Mandic Date: Sun, 9 Feb 2025 10:17:19 -0500 Subject: [PATCH] update Signed-off-by: Vladimir Mandic --- AMD-ROCm.md | 5 +---- CLI-Arguments.md | 2 +- Control-Settings.md | 4 ++-- Control-Technical.md | 2 +- DirectML.md | 4 ++-- FAQ.md | 2 +- FLUX.md | 10 +++++----- Features.md | 6 +++--- Installation.md | 2 +- Intel-ARC.md | 4 ++-- Locale.md | 5 +++++ Models.md | 1 - ONNX-Runtime.md | 2 +- OpenVINO.md | 2 +- Quantization.md | 4 ++-- SD3.md | 6 +++--- 16 files changed, 31 insertions(+), 30 deletions(-) diff --git a/AMD-ROCm.md b/AMD-ROCm.md index c3bf2da..f3cea46 100644 --- a/AMD-ROCm.md +++ b/AMD-ROCm.md @@ -83,9 +83,6 @@ This guide is for users with AMD GPUs lacking official ROCm/[HIP SDK](https://ww If you already have the libraries, you can skip this section! -To build your own rocBLAS, follow this guide. A simplified version is available on GitHub: -[ROCm-HIP-SDK-Windows-Support](https://github.com/vladmandic/sdnext/wiki/ROCm-HIP-SDK-Windows-Support). - **Prerequisites:** Ensure the following software is installed on your PC. `python`, `git`, and the `HIP SDK`are essential. The script `rdeps.py` will automatically download any missing dependencies when you run it. @@ -115,7 +112,7 @@ essential. The script `rdeps.py` will automatically download any missing depend ## Patching Tensile for ROCm (For Advanced Users, Not-a-must-Do) These steps are necessary for specific configurations of ROCm and may not be required in all cases. -If you had a optimized logic for you gpu arche or [do the necessary edit as guide here](https://github.com/vladmandic/sdnext/wiki/Rocm-Support#note-editing-tensilecommonpy),you may skip this steps.Especily build libs for xnack- features. +If you had a optimized logic for you gpu arche,you may skip this steps.Especily build libs for xnack- features. ### Determine Your ROCm Version: diff --git a/CLI-Arguments.md b/CLI-Arguments.md index 855dab6..0a180e4 100644 --- a/CLI-Arguments.md +++ b/CLI-Arguments.md @@ -142,7 +142,7 @@ ### Troubleshooting Options -**Also see [Troubleshooting Wiki page](https://github.com/vladmandic/sdnext/wiki/Troubleshooting)** +**Also see [Troubleshooting Wiki page](Troubleshooting)** - `--safe`: Run in safe mode with no user extensions. Safe mode can be useful when troubleshooting or when you want to restrict the execution of potentially unsafe user-provided code. Enabling this flag (`True`) runs the application in safe mode, disabling user extensions. The default is `False`, and you can set it to `True` using the environment variable `SD_SAFE`. **Try this first.** diff --git a/Control-Settings.md b/Control-Settings.md index 05f02e0..75f83c9 100644 --- a/Control-Settings.md +++ b/Control-Settings.md @@ -12,7 +12,7 @@ This guide will attempt to explain how to use it so that anyone can understand i ### Input -The Input control is exactly what it sounds like, it controls what input images (or videos) are contributing to your image generation, by default that is just the image in [Control input pane](https://github.com/vladmandic/sdnext/wiki/control-guide#control-input), however if you select `Separate init image`, another image pane will appear below, allowing you to use that as well. +The Input control is exactly what it sounds like, it controls what input images (or videos) are contributing to your image generation, by default that is just the image in the Control input pane, however if you select `Separate init image`, another image pane will appear below, allowing you to use that as well. **Note:** When using a Control input image as well as a Init input image, the Init input dominates. Adjusting denoise to >=0.9 is recommended, as that will allow the Control input to balance with the Init input. Higher values will increase the strength of Control input further, giving it dominance. @@ -61,7 +61,7 @@ The Mask controls are where we start getting into the real meat of Control, not - `Auto-mask`: There are three options here, Threshold, Edge, and Greyscale. Each provides a different method of auto-masking your images. -- `Auto-segment`: Just like Auto-mask, we have provided an extensive list of [Auto-segmentation models](https://github.com/vladmandic/sdnext/wiki/control#segmentation-models), they don't require ControlNet to handle the process, but may take a few seconds to process, depending on your GPU. +- `Auto-segment`: Just like Auto-mask, we have provided an extensive list of [Auto-segmentation models](Control-Technical#segmentation-models), they don't require ControlNet to handle the process, but may take a few seconds to process, depending on your GPU. - `Preview`: You can select the preview type here, we have provided 5 modes, Masked, Binary, Greyscale, Color, and Composite, which is the default. diff --git a/Control-Technical.md b/Control-Technical.md index 8329458..4950439 100644 --- a/Control-Technical.md +++ b/Control-Technical.md @@ -5,7 +5,7 @@ Native control module for SD.Next for Diffusers backend Can be used for Control generation as well as Image and Text workflows -For a guide on the options and settings, as well as explanations for the controls themselves, see the [Control Guide](https://github.com/vladmandic/sdnext/wiki/Control-Guide) page. +For a guide on the options and settings, as well as explanations for the controls themselves, see the [Control Howto](Control-HowTo) and [Control Settings](Control-Settings) pages. ## Supported Control Models diff --git a/DirectML.md b/DirectML.md index 5cdb483..a61d5e3 100644 --- a/DirectML.md +++ b/DirectML.md @@ -6,7 +6,7 @@ SD.Next includes support for PyTorch-DirectML. Add `--use-directml` on commandline arguments. -For details, go to [Installation](https://github.com/vladmandic/sdnext/wiki/Installation#general-installation). +For details, go to [Installation](Installation#general-installation). ## Performance @@ -34,4 +34,4 @@ Please report us via GitHub issue or Discord. (please check whether there's a du ## Olive (experimental support) -Refer to [ONNX Runtime](https://github.com/vladmandic/sdnext/wiki/ONNX-Runtime-&-Olive) +Refer to [ONNX Runtime](ONNX-Runtime) diff --git a/FAQ.md b/FAQ.md index f6b9b6a..e2ca61e 100644 --- a/FAQ.md +++ b/FAQ.md @@ -23,7 +23,7 @@ ### How do I use an AMD GPU on Windows? *Add the --use-zluda command-line flag when starting the app.* -Checkout the [Zluda wiki](https://github.com/vladmandic/sdnext/wiki//ZLUDA) for more info. +Checkout the [Zluda wiki](ZLUDA) for more info. ### How can I create large images (e.g., 2048x2048) with limited VRAM? diff --git a/FLUX.md b/FLUX.md index 3463b3b..016095e 100644 --- a/FLUX.md +++ b/FLUX.md @@ -16,17 +16,17 @@ Pick variant that uses less memory as model in original form has very high requi > [!IMPORTANT] > Allow gated access > This is a gated model, you need to accept the terms and conditions to use it -> For more information see [Gated Access Wiki](https://github.com/vladmandic/sdnext/wiki/Gated) +> For more information see [Gated Access Wiki](Gated) > [!IMPORTANT] > Set offloading > Set appropriate offloading setting before loading the model to avoid out-of-memory errors -> For more information see [Offloading Wiki](https://github.com/vladmandic/sdnext/wiki/Offload) +> For more information see [Offloading Wiki](Offload) > [!IMPORTANT] > Choose quantization > Check compatibility of different quantizations with your platform and GPU! -> For more information see [Quantization Wiki](https://github.com/vladmandic/sdnext/wiki/Quantization) +> For more information see [Quantization Wiki](Quantization) > [!TIP] > Use reference models @@ -151,8 +151,8 @@ Link to [Flux Tools](https://blackforestlabs.ai/flux-1-tools/) announcement To use, go to image or control interface and select *Flux Tools* in scripts All models are auto-downloaded on first use -*note*: All models are [gated](https://github.com/vladmandic/sdnext/wiki/Gated) and require acceptance of terms and conditions via web page -*recommended*: Enable on-the-fly [quantization](https://github.com/vladmandic/sdnext/wiki/Quantization) or [compression](https://github.com/vladmandic/sdnext/wiki/NNCF-Compression) to reduce resource usage +*note*: All models are [gated](Gated) and require acceptance of terms and conditions via web page +*recommended*: Enable on-the-fly [quantization](Quantization) or [compression](NNCF-Compression) to reduce resource usage - [Redux](https://huggingface.co/black-forest-labs/FLUX.1-Redux-dev): ~0.1GB works together with existing model and basically uses input image to analyze it and use that instead of prompt *recommended*: low denoise strength levels result in more variety diff --git a/Features.md b/Features.md index 95ed017..2100e36 100644 --- a/Features.md +++ b/Features.md @@ -10,7 +10,7 @@ After doing that, we decided that we would add everything else under the sun tha This guide will attempt to explain how to use it so that anyone can understand it and put it to work for themselves. -Be sure to also check out [the Control resource page](https://github.com/vladmandic/sdnext/wiki/Control) which has more technical information as well as some general tips and suggestions. Its usage information will be merged into this page soon™️. +Be sure to also check out [the Control resource page](Control-Technical) which has more technical information as well as some general tips and suggestions. Its usage information will be merged into this page soon™️. We'll start with the... Control Controls! @@ -18,7 +18,7 @@ We'll start with the... Control Controls! #### Input -The Input control is exactly what it sounds like, it controls what input images (or videos) are contributing to your image generation, by default that is just the image in [Control input pane](https://github.com/vladmandic/sdnext/wiki/control-guide#control-input), however if you select `Separate init image`, another image pane will appear below, allowing you to use that as well. +The Input control is exactly what it sounds like, it controls what input images (or videos) are contributing to your image generation, by default that is just the image in the Control input pane, however if you select `Separate init image`, another image pane will appear below, allowing you to use that as well. **Note:** When using a Control input image as well as a Init input image, the Init input dominates. Adjusting denoise to >=0.9 is recommended, as that will allow the Control input to balance with the Init input. Higher values will increase the strength of Control input further, giving it dominance. @@ -69,7 +69,7 @@ The Mask controls are where we start getting into the real meat of Control, not - `Auto-mask`: There are three options here, Threshold, Edge, and Greyscale. Each provides a different method of auto-masking your images. -- `Auto-segment`: Just like Auto-mask, we have provided an extensive list of [Auto-segmentation models](https://github.com/vladmandic/sdnext/wiki/control#segmentation-models), they don't require ControlNet to handle the process, but may take a few seconds to process, depending on your GPU. +- `Auto-segment`: Just like Auto-mask, we have provided an extensive list of [Auto-segmentation models](Control-Technical#segmentation-models), they don't require ControlNet to handle the process, but may take a few seconds to process, depending on your GPU. - `Preview`: You can select the preview type here, we have provided 5 modes, Masked, Binary, Greyscale, Color, and Composite, which is the default. diff --git a/Installation.md b/Installation.md index 615d8bd..44ebe6c 100644 --- a/Installation.md +++ b/Installation.md @@ -124,4 +124,4 @@ The server is finished launching when the console shows an entry for **"Startup #### Python-for-MacOS -[See these instructions for Python on MacOS (and an explanation why it's unique).](https://github.com/vladmandic/sdnext/wiki/MacOS-Python) +[See these instructions for Python on MacOS (and an explanation why it's unique).](MacOS-Python) diff --git a/Intel-ARC.md b/Intel-ARC.md index 2a30563..34e280b 100644 --- a/Intel-ARC.md +++ b/Intel-ARC.md @@ -10,7 +10,7 @@ > [!NOTE] -> iGPUs are **not** supported with IPEX, use [OpenVINO](https://github.com/vladmandic/sdnext/wiki/OpenVINO) if you want to use an iGPU. +> iGPUs are **not** supported with IPEX, use [OpenVINO](OpenVINO) if you want to use an iGPU. > Having an Intel iGPU can cause conflicts and random errors if you want to use a dedicated Intel GPU with IPEX. > Disable your iGPU (if any, e.g. UHD or Iris Xe) in the device manager. @@ -21,7 +21,7 @@ - Install [Intel ARC Drivers](https://www.intel.com/content/www/us/en/download/785597/intel-arc-iris-xe-graphics-windows.html) - Install [Intel OneAPI Deep Learning Essentials](https://www.intel.com/content/www/us/en/developer/tools/oneapi/base-toolkit-download.html?packages=dl-essentials&dl-essentials-os=windows&dl-lin=offline) -- Install [Git and Python](https://github.com/vladmandic/sdnext/wiki/Installation#install-python-and-git) +- Install [Git and Python](Installation#install-python-and-git) - Open CMD in a folder you want to install SD.Next diff --git a/Locale.md b/Locale.md index 29a1aee..0f028d8 100644 --- a/Locale.md +++ b/Locale.md @@ -18,6 +18,11 @@ Those files should not be edited directly as they are auto-generated and will be > [!NOTE] > Translation is performed using `cli/localize.js` script +### Editing Localization + +In case of bad ML translation or you just want to provide a better alternernative, +you can create/edit `hmtl/override_{locale}.json` which only needs to include specific items you want to override + ### Supported Languages - `en`: English diff --git a/Models.md b/Models.md index 9a1c0ae..04091be 100644 --- a/Models.md +++ b/Models.md @@ -1,7 +1,6 @@ # Models List of popular text-to-image generative models with their respective parameters and architecture overview -Original URL: | Publisher | Model | Version | Size | Diffusion Architecture | Model Params | Text Encoder(s) | TE Params | Auto Encoder | Other | | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | diff --git a/ONNX-Runtime.md b/ONNX-Runtime.md index 7e2855a..cfe43d3 100644 --- a/ONNX-Runtime.md +++ b/ONNX-Runtime.md @@ -72,7 +72,7 @@ Olive is an easy-to-use hardware-aware model optimization tool that composes ind ## How to -As Olive optimizes the models in ONNX format, you should set up [ONNX Runtime](https://github.com/vladmandic/sdnext/wiki/ONNX-Runtime-%26-Olive/#how-to) first. +As Olive optimizes the models in ONNX format, you should set up ONNX Runtime first. 1. Go to `System` tab → `Compute Settings`. 2. Select `Model`, `Text Encoder` and `VAE` in `Compile Model`. diff --git a/OpenVINO.md b/OpenVINO.md index 1b4d8c6..7d4836e 100644 --- a/OpenVINO.md +++ b/OpenVINO.md @@ -19,7 +19,7 @@ It is basically a TensorRT / Olive competitor that works with any hardware. ### Preparations - Install the drivers for your device -- Install [Git and Python](https://github.com/vladmandic/sdnext/wiki/Installation#install-python-and-git) +- Install [Git and Python](Installation#install-python-and-git) - Open CMD in a folder you want to install SD.Next > [!NOTE] diff --git a/Quantization.md b/Quantization.md index 2f389de..2cb7d42 100644 --- a/Quantization.md +++ b/Quantization.md @@ -41,7 +41,7 @@ Quantization can be done in multiple ways: > [!TIP] > If you're on Windows with a compatible GPU, you may try **WSL2** for broader feature compatibiliy -> See [WSL Wiki](https://github.com/vladmandic/sdnext/wiki/WSL) for more details +> See [WSL Wiki](WSL) for more details SD.Next supports multiple quantization engines, each with multiple quantization schemes: - `TorchAO`: 4 int-based and 3 float-based quantization schemes @@ -118,7 +118,7 @@ with optional platform-specific compute-optimization (available only on OpenVINO - broad platform and GPU support - enable in *Settings -> Compute -> Compress model weights with NNCF* -- see [NNCF Wiki](https://github.com/vladmandic/sdnext/wiki/NNCF-Compression) for more details +- see [NNCF Wiki](NNCF-Compression) for more details ## Errors diff --git a/SD3.md b/SD3.md index 61c9ec5..e80fb47 100644 --- a/SD3.md +++ b/SD3.md @@ -12,17 +12,17 @@ StabilityAI's Stable Diffusion 3 family consists of: > [!IMPORTANT] > Allow gated access > This is a gated model, you need to accept the terms and conditions to use it -> For more information see [Gated Access Wiki](https://github.com/vladmandic/sdnext/wiki/Gated) +> For more information see [Gated Access Wiki](Gated) > [!IMPORTANT] > Set offloading > Set appropriate offloading setting before loading the model to avoid out-of-memory errors -> For more information see [Offloading Wiki](https://github.com/vladmandic/sdnext/wiki/Offload) +> For more information see [Offloading Wiki](Offload) > [!IMPORTANT] > Choose quantization > Check compatibility of different quantizations with your platform and GPU! -> For more information see [Quantization Wiki](https://github.com/vladmandic/sdnext/wiki/Quantization) +> For more information see [Quantization Wiki](Quantization) > [!TIP] Use reference models > Use of reference models is recommended over manually downloaded models!