| Name | sdkit JSON |
| Version |
2.0.18
JSON |
| download |
| home_page | |
| Summary | sdkit (stable diffusion kit) is an easy-to-use library for using Stable Diffusion in your AI Art projects. It is fast, feature-packed, and memory-efficient. It bundles Stable Diffusion along with commonly-used features (like SDXL, ControlNet, LoRA, Embeddings, GFPGAN, RealESRGAN, k-samplers, custom VAE etc). It also includes a model-downloader with a database of commonly used models, and advanced features like running in parallel on multiple GPUs, auto-scanning for malicious models etc. |
| upload_time | 2024-03-13 13:48:28 |
| maintainer | |
| docs_url | None |
| author | |
| requires_python | >=3.8.5 |
| license | |
| keywords |
stable diffusion
ai
art
|
| VCS |
 |
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
# sdkit
**sdkit** (**s**table **d**iffusion **kit**) is an easy-to-use library for using Stable Diffusion in your AI Art projects. It is fast, feature-packed, and memory-efficient.
[](https://discord.com/invite/u9yhsFmEkB)
*New: Stable Diffusion XL, ControlNets, LoRAs and Embeddings are now supported!*
This is a community project, so please feel free to contribute (and to use it in your project)!
# Why?
The goal is to let you be productive quickly (at your AI art project), so it bundles Stable Diffusion along with commonly-used features (like ControlNets, LoRAs, Textual Inversion Embeddings, GFPGAN and CodeFormer for face restoration, RealESRGAN for upscaling, k-samplers, support for loading custom VAEs, NSFW filter etc).
Advanced features include a model-downloader (with a database of commonly used models), support for running in parallel on multiple GPUs, auto-scanning for malicious models, etc. [Full list of features](https://github.com/easydiffusion/sdkit/wiki/Features)
# Installation
Tested with Python 3.8. Supports Windows, Linux, and Mac.
**Windows/Linux:**
1. `pip install torch torchvision --extra-index-url https://download.pytorch.org/whl/cu116`
2. Run `pip install sdkit`
**Mac:**
1. Run `pip install sdkit`
# Example
### Local model
A simple example for generating an image from a Stable Diffusion model file (already present on the disk):
```python
import sdkit
from sdkit.models import load_model
from sdkit.generate import generate_images
from sdkit.utils import log
context = sdkit.Context()
# set the path to the model file on the disk (.ckpt or .safetensors file)
context.model_paths['stable-diffusion'] = 'D:\\path\\to\\512-base-ema.ckpt'
load_model(context, 'stable-diffusion')
# generate the image
images = generate_images(context, prompt='Photograph of an astronaut riding a horse', seed=42, width=512, height=512)
# save the image
images[0].save("image.png") # images is a list of PIL.Image
log.info("Generated images!")
```
### Auto-download a known model
A simple example for automatically downloading a known Stable Diffusion model file:
```python
import sdkit
from sdkit.models import download_models, resolve_downloaded_model_path, load_model
from sdkit.generate import generate_images
from sdkit.utils import save_images
context = sdkit.Context()
download_models(context, models={'stable-diffusion': '1.5-pruned-emaonly'}) # downloads the known "SD 1.5-pruned-emaonly" model
context.model_paths['stable-diffusion'] = resolve_downloaded_model_path(context, 'stable-diffusion', '1.5-pruned-emaonly')
load_model(context, 'stable-diffusion')
images = generate_images(context, prompt='Photograph of an astronaut riding a horse', seed=42, width=512, height=512)
save_images(images, dir_path='D:\\path\\to\\images\\directory')
```
Please see the list of [examples](https://github.com/easydiffusion/sdkit/tree/main/examples), to learn how to use the other features (like filters, VAE, ControlNet, Embeddings, LoRA, memory optimizations, running on multiple GPUs etc).
# API
Please see the [API Reference](https://github.com/easydiffusion/sdkit/wiki/API) page for a detailed summary.
Broadly, the API contains 5 modules:
```python
sdkit.models # load/unloading models, downloading known models, scanning models
sdkit.generate # generating images
sdkit.filter # face restoration, upscaling
sdkit.train # model merge, and (in the future) more training methods
sdkit.utils
```
And a `sdkit.Context` object is passed around, which encapsulates the data related to the runtime (e.g. `device` and `vram_optimizations`) as well as references to the loaded model files and paths. `Context` is a thread-local object.
# Models DB
Click here to see the [list of known models](https://github.com/easydiffusion/sdkit/tree/main/sdkit/models/models_db).
sdkit includes a database of known models and their configurations. This lets you download a known model with a single line of code. (You can customize where it saves the downloaded model)
Additionally, sdkit will attempt to automatically determine the configuration for a given model (when loading from disk). E.g. if an SD 2.1 model is being loaded, sdkit will automatically know to use `fp32` for `attn_precision`. If an SD 2.0 v-type model is being loaded, sdkit will automatically know to use the `v2-inference-v.yaml` configuration. It does this by matching the quick-hash of the given model file, with the list of known quick-hashes.
For models that don't match a known hash (e.g. custom models), or if you want to override the config file, you can set the path to the config file in `context.model_paths`. e.g. `context.model_paths['stable-diffusion'] = 'path/to/config.yaml'`
# FAQ
## Does it have all the cool features?
It was born out of a popular Stable Diffusion UI, splitting out the battle-tested core engine into `sdkit`.
**Features include:** SD 2.1, SDXL, ControlNet, LoRAs, Embeddings, txt2img, img2img, inpainting, NSFW filter, multiple GPU support, Mac Support, GFPGAN and CodeFormer (fix faces), RealESRGAN (upscale), 16 samplers (including k-samplers and UniPC), custom VAE, low-memory optimizations, model merging, safetensor support, picklescan, etc. [Click here to see the full list of features](https://github.com/easydiffusion/sdkit/wiki/Features).
📢 We're looking to add support for *Lycoris*, *AMD*, *Pix2Pix*, and *outpainting*. We'd love some code contributions for these!
## Is it fast?
It is pretty fast, and close to the fastest. For the same image, `sdkit` took 5.5 seconds, while `automatic1111` WebUI took 4.95 seconds. 📢 We're looking for code contributions to make `sdkit` even faster!
`xformers` is supported experimentally, which will make `sdkit` even faster.
**Details of the benchmark:**
Windows 11, NVIDIA 3060 12GB, 512x512 image, sd-v1-4.ckpt, euler_a sampler, number of steps: 25, seed 42, guidance scale 7.5.
No xformers. No VRAM optimizations for low-memory usage.
| | Time taken | Iterations/sec | Peak VRAM usage |
| --- | --- | --- | --- |
| `sdkit` | 5.5 sec | 6.0 it/s | 5.1 GB |
| `automatic1111` webui | 4.95 sec | 6.15 it/s | 5.1 GB |
## Does it work on lower-end GPUs, or without GPUs?
Yes. It works on NVIDIA/Mac GPUs with at least 2GB of VRAM. For PCs without a compatible GPU, it can run entirely on the CPU. Running on the CPU will be *very* slow, but at least you'll be able to try it out!
📢 We don't support AMD yet on Windows (it'll run in CPU mode, or in Linux), but we're looking for code contributions for AMD support!
## Why not just use diffusers?
You can certainly use diffusers. `sdkit` is in fact using `diffusers` internally, so you can think of `sdkit` as a convenient API and a collection of tools, focused on Stable Diffusion projects.
`sdkit`:
1. is a simple, lightweight toolkit for Stable Diffusion projects.
2. natively includes frequently-used projects like GFPGAN, CodeFormer, and RealESRGAN.
3. works with the popular `.ckpt` and `.safetensors` model format.
4. includes memory optimizations for low-end GPUs.
5. built-in support for running on multiple GPUs.
6. can download models from any server.
7. Auto scans for malicious models.
8. includes 16 samplers (including k-samplers).
9. born out of the needs of the new Stable Diffusion AI Art scene, starting Aug 2022.
# Who is using sdkit?
* [Easy Diffusion (cmdr2 UI)](https://github.com/easydiffusion/easydiffusion) for Stable Diffusion.
* [Arthemy AI](https://arthemy.ai/)
If your project is using sdkit, you can add it to this list. Please feel free to open a pull request (or let us know at our [Discord community](https://discord.com/invite/u9yhsFmEkB)).
# Contributing
We'd love to accept code contributions. Please feel free to drop by our [Discord community](https://discord.com/invite/u9yhsFmEkB)!
📢 We're looking for code contributions for these features (or anything else you'd like to work on):
- Lycoris.
- Outpainting.
- Pix2Pix.
- AMD support.
If you'd like to set up a developer version on your PC (to contribute code changes), please follow [these instructions](https://github.com/easydiffusion/sdkit/blob/main/CONTRIBUTING.md).
Instructions for running automated tests: [Running Tests](tests/README.md).
# Credits
* Stable Diffusion: https://github.com/Stability-AI/stablediffusion
* CodeFormer: https://github.com/sczhou/CodeFormer (license: https://github.com/sczhou/CodeFormer/blob/master/LICENSE)
* GFPGAN: https://github.com/TencentARC/GFPGAN
* RealESRGAN: https://github.com/xinntao/Real-ESRGAN
* k-diffusion: https://github.com/crowsonkb/k-diffusion
* Code contributors and artists on Easy Diffusion (cmdr2 UI): https://github.com/easydiffusion/easydiffusion and Discord (https://discord.com/invite/u9yhsFmEkB)
* Lots of contributors on the internet
# Disclaimer
The authors of this project are not responsible for any content generated using this project.
The license of this software forbids you from sharing any content that:
- Violates any laws.
- Produces any harm to a person or persons.
- Disseminates (spreads) any personal information that would be meant for harm.
- Spreads misinformation.
- Target vulnerable groups.
For the full list of restrictions please read [the License](https://github.com/easydiffusion/sdkit/blob/main/LICENSE). By using this software you agree to the terms.
Raw data
{
"_id": null,
"home_page": "",
"name": "sdkit",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.8.5",
"maintainer_email": "",
"keywords": "stable diffusion,ai,art",
"author": "",
"author_email": "cmdr2 <secondary.cmdr2@gmail.com>",
"download_url": "https://files.pythonhosted.org/packages/f3/2b/7671c0389b1c93a74d3de66ef6585ef5c674d8af11fde2a3947b1a8e5ee3/sdkit-2.0.18.tar.gz",
"platform": null,
"description": "# sdkit\n**sdkit** (**s**table **d**iffusion **kit**) is an easy-to-use library for using Stable Diffusion in your AI Art projects. It is fast, feature-packed, and memory-efficient.\n\n[](https://discord.com/invite/u9yhsFmEkB)\n\n*New: Stable Diffusion XL, ControlNets, LoRAs and Embeddings are now supported!*\n\nThis is a community project, so please feel free to contribute (and to use it in your project)!\n\n\n\n# Why?\nThe goal is to let you be productive quickly (at your AI art project), so it bundles Stable Diffusion along with commonly-used features (like ControlNets, LoRAs, Textual Inversion Embeddings, GFPGAN and CodeFormer for face restoration, RealESRGAN for upscaling, k-samplers, support for loading custom VAEs, NSFW filter etc).\n\nAdvanced features include a model-downloader (with a database of commonly used models), support for running in parallel on multiple GPUs, auto-scanning for malicious models, etc. [Full list of features](https://github.com/easydiffusion/sdkit/wiki/Features)\n\n# Installation\nTested with Python 3.8. Supports Windows, Linux, and Mac.\n\n**Windows/Linux:**\n1. `pip install torch torchvision --extra-index-url https://download.pytorch.org/whl/cu116`\n2. Run `pip install sdkit`\n\n**Mac:**\n1. Run `pip install sdkit`\n\n# Example\n### Local model\nA simple example for generating an image from a Stable Diffusion model file (already present on the disk):\n```python\nimport sdkit\nfrom sdkit.models import load_model\nfrom sdkit.generate import generate_images\nfrom sdkit.utils import log\n\ncontext = sdkit.Context()\n\n# set the path to the model file on the disk (.ckpt or .safetensors file)\ncontext.model_paths['stable-diffusion'] = 'D:\\\\path\\\\to\\\\512-base-ema.ckpt'\nload_model(context, 'stable-diffusion')\n\n# generate the image\nimages = generate_images(context, prompt='Photograph of an astronaut riding a horse', seed=42, width=512, height=512)\n\n# save the image\nimages[0].save(\"image.png\") # images is a list of PIL.Image\n\nlog.info(\"Generated images!\")\n```\n\n### Auto-download a known model\nA simple example for automatically downloading a known Stable Diffusion model file:\n```python\nimport sdkit\nfrom sdkit.models import download_models, resolve_downloaded_model_path, load_model\nfrom sdkit.generate import generate_images\nfrom sdkit.utils import save_images\n\ncontext = sdkit.Context()\n\ndownload_models(context, models={'stable-diffusion': '1.5-pruned-emaonly'}) # downloads the known \"SD 1.5-pruned-emaonly\" model\n\ncontext.model_paths['stable-diffusion'] = resolve_downloaded_model_path(context, 'stable-diffusion', '1.5-pruned-emaonly')\nload_model(context, 'stable-diffusion')\n\nimages = generate_images(context, prompt='Photograph of an astronaut riding a horse', seed=42, width=512, height=512)\nsave_images(images, dir_path='D:\\\\path\\\\to\\\\images\\\\directory')\n```\n\nPlease see the list of [examples](https://github.com/easydiffusion/sdkit/tree/main/examples), to learn how to use the other features (like filters, VAE, ControlNet, Embeddings, LoRA, memory optimizations, running on multiple GPUs etc).\n\n# API\nPlease see the [API Reference](https://github.com/easydiffusion/sdkit/wiki/API) page for a detailed summary.\n\nBroadly, the API contains 5 modules:\n```python\nsdkit.models # load/unloading models, downloading known models, scanning models\nsdkit.generate # generating images\nsdkit.filter # face restoration, upscaling\nsdkit.train # model merge, and (in the future) more training methods\nsdkit.utils\n```\n\nAnd a `sdkit.Context` object is passed around, which encapsulates the data related to the runtime (e.g. `device` and `vram_optimizations`) as well as references to the loaded model files and paths. `Context` is a thread-local object.\n\n# Models DB\nClick here to see the [list of known models](https://github.com/easydiffusion/sdkit/tree/main/sdkit/models/models_db).\n\nsdkit includes a database of known models and their configurations. This lets you download a known model with a single line of code. (You can customize where it saves the downloaded model)\n\nAdditionally, sdkit will attempt to automatically determine the configuration for a given model (when loading from disk). E.g. if an SD 2.1 model is being loaded, sdkit will automatically know to use `fp32` for `attn_precision`. If an SD 2.0 v-type model is being loaded, sdkit will automatically know to use the `v2-inference-v.yaml` configuration. It does this by matching the quick-hash of the given model file, with the list of known quick-hashes.\n\nFor models that don't match a known hash (e.g. custom models), or if you want to override the config file, you can set the path to the config file in `context.model_paths`. e.g. `context.model_paths['stable-diffusion'] = 'path/to/config.yaml'`\n\n# FAQ\n## Does it have all the cool features?\nIt was born out of a popular Stable Diffusion UI, splitting out the battle-tested core engine into `sdkit`.\n\n**Features include:** SD 2.1, SDXL, ControlNet, LoRAs, Embeddings, txt2img, img2img, inpainting, NSFW filter, multiple GPU support, Mac Support, GFPGAN and CodeFormer (fix faces), RealESRGAN (upscale), 16 samplers (including k-samplers and UniPC), custom VAE, low-memory optimizations, model merging, safetensor support, picklescan, etc. [Click here to see the full list of features](https://github.com/easydiffusion/sdkit/wiki/Features).\n\n\ud83d\udce2 We're looking to add support for *Lycoris*, *AMD*, *Pix2Pix*, and *outpainting*. We'd love some code contributions for these!\n\n## Is it fast?\nIt is pretty fast, and close to the fastest. For the same image, `sdkit` took 5.5 seconds, while `automatic1111` WebUI took 4.95 seconds. \ud83d\udce2 We're looking for code contributions to make `sdkit` even faster!\n\n`xformers` is supported experimentally, which will make `sdkit` even faster.\n\n**Details of the benchmark:**\n\nWindows 11, NVIDIA 3060 12GB, 512x512 image, sd-v1-4.ckpt, euler_a sampler, number of steps: 25, seed 42, guidance scale 7.5.\n\nNo xformers. No VRAM optimizations for low-memory usage.\n\n| | Time taken | Iterations/sec | Peak VRAM usage |\n| --- | --- | --- | --- |\n| `sdkit` | 5.5 sec | 6.0 it/s | 5.1 GB |\n| `automatic1111` webui | 4.95 sec | 6.15 it/s | 5.1 GB |\n\n## Does it work on lower-end GPUs, or without GPUs?\nYes. It works on NVIDIA/Mac GPUs with at least 2GB of VRAM. For PCs without a compatible GPU, it can run entirely on the CPU. Running on the CPU will be *very* slow, but at least you'll be able to try it out!\n\n\ud83d\udce2 We don't support AMD yet on Windows (it'll run in CPU mode, or in Linux), but we're looking for code contributions for AMD support!\n\n## Why not just use diffusers?\nYou can certainly use diffusers. `sdkit` is in fact using `diffusers` internally, so you can think of `sdkit` as a convenient API and a collection of tools, focused on Stable Diffusion projects.\n\n`sdkit`:\n1. is a simple, lightweight toolkit for Stable Diffusion projects.\n2. natively includes frequently-used projects like GFPGAN, CodeFormer, and RealESRGAN.\n3. works with the popular `.ckpt` and `.safetensors` model format.\n4. includes memory optimizations for low-end GPUs.\n5. built-in support for running on multiple GPUs.\n6. can download models from any server.\n7. Auto scans for malicious models.\n8. includes 16 samplers (including k-samplers).\n9. born out of the needs of the new Stable Diffusion AI Art scene, starting Aug 2022.\n\n# Who is using sdkit?\n* [Easy Diffusion (cmdr2 UI)](https://github.com/easydiffusion/easydiffusion) for Stable Diffusion.\n* [Arthemy AI](https://arthemy.ai/)\n\nIf your project is using sdkit, you can add it to this list. Please feel free to open a pull request (or let us know at our [Discord community](https://discord.com/invite/u9yhsFmEkB)).\n\n# Contributing\nWe'd love to accept code contributions. Please feel free to drop by our [Discord community](https://discord.com/invite/u9yhsFmEkB)!\n\n\ud83d\udce2 We're looking for code contributions for these features (or anything else you'd like to work on):\n- Lycoris.\n- Outpainting.\n- Pix2Pix.\n- AMD support.\n\nIf you'd like to set up a developer version on your PC (to contribute code changes), please follow [these instructions](https://github.com/easydiffusion/sdkit/blob/main/CONTRIBUTING.md).\n\nInstructions for running automated tests: [Running Tests](tests/README.md).\n\n# Credits\n* Stable Diffusion: https://github.com/Stability-AI/stablediffusion\n* CodeFormer: https://github.com/sczhou/CodeFormer (license: https://github.com/sczhou/CodeFormer/blob/master/LICENSE)\n* GFPGAN: https://github.com/TencentARC/GFPGAN\n* RealESRGAN: https://github.com/xinntao/Real-ESRGAN\n* k-diffusion: https://github.com/crowsonkb/k-diffusion\n* Code contributors and artists on Easy Diffusion (cmdr2 UI): https://github.com/easydiffusion/easydiffusion and Discord (https://discord.com/invite/u9yhsFmEkB)\n* Lots of contributors on the internet\n\n# Disclaimer\nThe authors of this project are not responsible for any content generated using this project.\n\nThe license of this software forbids you from sharing any content that:\n- Violates any laws.\n- Produces any harm to a person or persons.\n- Disseminates (spreads) any personal information that would be meant for harm.\n- Spreads misinformation.\n- Target vulnerable groups. \n\nFor the full list of restrictions please read [the License](https://github.com/easydiffusion/sdkit/blob/main/LICENSE). By using this software you agree to the terms.\n",
"bugtrack_url": null,
"license": "",
"summary": "sdkit (stable diffusion kit) is an easy-to-use library for using Stable Diffusion in your AI Art projects. It is fast, feature-packed, and memory-efficient. It bundles Stable Diffusion along with commonly-used features (like SDXL, ControlNet, LoRA, Embeddings, GFPGAN, RealESRGAN, k-samplers, custom VAE etc). It also includes a model-downloader with a database of commonly used models, and advanced features like running in parallel on multiple GPUs, auto-scanning for malicious models etc.",
"version": "2.0.18",
"project_urls": {
"Bug Tracker": "https://github.com/easydiffusion/sdkit/issues",
"Homepage": "https://github.com/easydiffusion/sdkit"
},
"split_keywords": [
"stable diffusion",
"ai",
"art"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "a7d6596fe6706fc49bd3fa29044f7c72cc3e8806714d67216d4ca45466770737",
"md5": "735ea34fcb0a62b0220ebad074ef6372",
"sha256": "662bd357707b81123e3f19a275fc04ae27feee48442ae3453301777de91f41d9"
},
"downloads": -1,
"filename": "sdkit-2.0.18-py3-none-any.whl",
"has_sig": false,
"md5_digest": "735ea34fcb0a62b0220ebad074ef6372",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.8.5",
"size": 128635,
"upload_time": "2024-03-13T13:48:23",
"upload_time_iso_8601": "2024-03-13T13:48:23.229777Z",
"url": "https://files.pythonhosted.org/packages/a7/d6/596fe6706fc49bd3fa29044f7c72cc3e8806714d67216d4ca45466770737/sdkit-2.0.18-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "f32b7671c0389b1c93a74d3de66ef6585ef5c674d8af11fde2a3947b1a8e5ee3",
"md5": "3c791fb33b2c17e29a36ddf3aa4c5ad8",
"sha256": "67570eb3bb330ff189e6f2a39708d932652dc688bb3d256ac1b9a5f113ba3e7f"
},
"downloads": -1,
"filename": "sdkit-2.0.18.tar.gz",
"has_sig": false,
"md5_digest": "3c791fb33b2c17e29a36ddf3aa4c5ad8",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.8.5",
"size": 102999,
"upload_time": "2024-03-13T13:48:28",
"upload_time_iso_8601": "2024-03-13T13:48:28.331875Z",
"url": "https://files.pythonhosted.org/packages/f3/2b/7671c0389b1c93a74d3de66ef6585ef5c674d8af11fde2a3947b1a8e5ee3/sdkit-2.0.18.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-03-13 13:48:28",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "easydiffusion",
"github_project": "sdkit",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"requirements": [],
"lcname": "sdkit"
}
