本地生成和训练依赖 ComfyUI 与 ai-toolkit,需按文档单独准备;新版本先向 Patreon 会员开放,非会员只能用已装版本;v1 已冻结,旧 main 安装需迁移到 V2。
lora-dataset-studio
perfectgf活跃维护本地一站式 LoRA 训练工作台,从单张照片或大批图片构建角色、风格、概念数据集,自动打标并用 ai-toolkit 训练,支持多模型 checkpoint 对比。
Free local LoRA training studio: build character, style and concept datasets from one photo or a big image folder, caption them (JoyCaption, prose or booru tags), train LoRAs with ai-toolkit for Krea 2, FLUX.1, FLUX.2 Klein, Qwen-Image 2.1, Z-Image, SDXL and Anima, then compare checkpoints side by side. ComfyUI-powered, self-hosted.
Windows:从 latest release 下载 LoRA-Dataset-Studio-windows.zip,解压到新文件夹后运行 start.bat核心功能无需 GPU,导入、整理和手动打标均可 CPU 运行;本地生成通常需要约 16 GB NVIDIA 显存,训练需求随模型家族和设置变化,详见 requirements 文档。
本地免费 LoRA 训练工作台,从一张照片或图片文件夹建数据集、打标、训练并对比检查点
本地一站式 LoRA 训练工作台,从单张照片或大批图片构建角色、风格、概念数据集,自动打标并用 ai-toolkit 训练,支持多模型 checkpoint 对比。
核心能力
- 从单张参考图生成角色多角度镜头目录并本地补图
- 用 JoyCaption 或本地视觉模型按模型家族写 prose 或 booru 标签
- 为 Z-Image、SDXL、FLUX.1、Qwen-Image 等七类模型提供 ai-toolkit 训练配方
- 固定种子并排对比检查点与强度,含投票和排名
- Image Bank 支持质量打分、去重分组和相似度检索建短名单
容易踩的坑
- v2026.11.12LoRA Dataset Studio V2 — v2026.11.12最新正式版2026-11-09 · 2026-11-09 20:03 UTC
LoRA Dataset Studio V2
V2 is the maintained LDS release. Existing ZIP installations can upgrade from Settings → Maintenance → Update & restart, keeping datasets, media and history. Git installations still on the former
mainbranch must first switch tov2. Optional features are plugins: open Plugins → Store after updating, install the ones you use and review their settings before the first run.Install the core, then choose your plugins
Download
LoRA-Dataset-Studio-windows.zipbelow, extract it into a new folder and double-clickstart.bat. The launcher prepares a compatible Python environment. Complete the core Setup, then open Plugins → Store to install the features you want. Each plugin has its own settings and preparation steps, including required ComfyUI custom nodes. Importing and organising datasets does not require a plugin.Free public plugins are available through the Store. Each plugin describes its features and preparation requirements. Paid API engines and rented GPUs still charge separately when you choose to use them.
The core and these public plugins are available at no charge, with their source published under the project's PolyForm Noncommercial license. New LDS versions reach Patreon members (LDS Member) first, then everyone, and some plugins are members-only; the version you have stays free and keeps working. Usage statistics are optional and off by default; the app asks before sharing, and you can change your choice in Settings → Maintenance.
Two ways to install and update
ZIP installations: Update & restart installs the next published release, preserving your data and local configuration. The ZIP below contains the core; plugins are installed and updated separately through the Store.
Git installations:
v2is the default and maintained branch. Update & restart follows the commits of your checkout's configured branch. The formermainbranch is now the frozenv1branch; it receives no further updates. Follow the migration instructions if your checkout still tracksmain.git clone https://github.com/perfectgf/lora-dataset-studio.git cd lora-dataset-studio start.bat🎁 What's new in v2026.11.12
start.bat starts when python is a version-manager shim
With pyenv-win, or another tool whose python command is a .bat or .cmd file, start.bat could stop right after its Python check: back to the prompt, no message, no app. It now carries on. And when a step does fail, start.bat says which one and waits for a key, so a double-clicked window no longer closes before you can read why.
Full changelog: https://github.com/perfectgf/lora-dataset-studio/compare/v2026.11.11...v2026.11.12
- v2026.11.11LoRA Dataset Studio V2 — v2026.11.112026-11-08 · 2026-11-08 10:25 UTC
查看更新说明
LoRA Dataset Studio V2
V2 is the maintained LDS release. Existing ZIP installations can upgrade from Settings → Maintenance → Update & restart, keeping datasets, media and history. Git installations still on the former
mainbranch must first switch tov2. Optional features are plugins: open Plugins → Store after updating, install the ones you use and review their settings before the first run.Install the core, then choose your plugins
Download
LoRA-Dataset-Studio-windows.zipbelow, extract it into a new folder and double-clickstart.bat. The launcher prepares a compatible Python environment. Complete the core Setup, then open Plugins → Store to install the features you want. Each plugin has its own settings and preparation steps, including required ComfyUI custom nodes. Importing and organising datasets does not require a plugin.Free public plugins are available through the Store. Each plugin describes its features and preparation requirements. Paid API engines and rented GPUs still charge separately when you choose to use them.
The core and these public plugins are available at no charge, with their source published under the project's PolyForm Noncommercial license. New LDS versions reach Patreon members (LDS Member) first, then everyone, and some plugins are members-only; the version you have stays free and keeps working. Usage statistics are optional and off by default; the app asks before sharing, and you can change your choice in Settings → Maintenance.
Two ways to install and update
ZIP installations: Update & restart installs the next published release, preserving your data and local configuration. The ZIP below contains the core; plugins are installed and updated separately through the Store.
Git installations:
v2is the default and maintained branch. Update & restart follows the commits of your checkout's configured branch. The formermainbranch is now the frozenv1branch; it receives no further updates. Follow the migration instructions if your checkout still tracksmain.git clone https://github.com/perfectgf/lora-dataset-studio.git cd lora-dataset-studio start.bat🎁 What's new in v2026.11.11
The diagnostic report copies in Firefox
In Firefox, 📋 Copy diagnostic report could fail with "Clipboard write was blocked due to lack of user activation" when the report took more than a few seconds to build. The copy now starts with your click and finishes once the report is ready. If a browser still refuses, the report appears under the button so you can copy it by hand.
start.bat works from a PowerShell 7 prompt
Launched from PowerShell 7, start.bat could not fetch its own Python ("Get-FileHash is not recognized") and warned that the ML extras would not install. The download now works from any prompt, and a Python 3.10–3.12 already on your PATH is used as is, even without the py launcher.
Full changelog: https://github.com/perfectgf/lora-dataset-studio/compare/v2026.11.10...v2026.11.11
- v2026.11.10LoRA Dataset Studio V2 — v2026.11.102026-11-07 · 2026-11-07 23:28 UTC
查看更新说明
LoRA Dataset Studio V2
V2 is the maintained LDS release. Existing ZIP installations can upgrade from Settings → Maintenance → Update & restart, keeping datasets, media and history. Git installations still on the former
mainbranch must first switch tov2. Optional features are plugins: open Plugins → Store after updating, install the ones you use and review their settings before the first run.Install the core, then choose your plugins
Download
LoRA-Dataset-Studio-windows.zipbelow, extract it into a new folder and double-clickstart.bat. The launcher prepares a compatible Python environment. Complete the core Setup, then open Plugins → Store to install the features you want. Each plugin has its own settings and preparation steps, including required ComfyUI custom nodes. Importing and organising datasets does not require a plugin.Free public plugins are available through the Store. Each plugin describes its features and preparation requirements. Paid API engines and rented GPUs still charge separately when you choose to use them.
The core and these public plugins are available at no charge, with their source published under the project's PolyForm Noncommercial license. New LDS versions reach Patreon members (LDS Member) first, then everyone, and some plugins are members-only; the version you have stays free and keeps working. Usage statistics are optional and off by default; the app asks before sharing, and you can change your choice in Settings → Maintenance.
Two ways to install and update
ZIP installations: Update & restart installs the next published release, preserving your data and local configuration. The ZIP below contains the core; plugins are installed and updated separately through the Store.
Git installations:
v2is the default and maintained branch. Update & restart follows the commits of your checkout's configured branch. The formermainbranch is now the frozenv1branch; it receives no further updates. Follow the migration instructions if your checkout still tracksmain.git clone https://github.com/perfectgf/lora-dataset-studio.git cd lora-dataset-studio start.bat🎁 What's new in v2026.11.10
A negative prompt for Krea test runs above CFG 1
The Test Studio offered no negative prompt for Krea, even on a Raw base that samples at CFG 4, where it changes the image. The Negative section now appears as soon as a Krea run uses a CFG above 1. At CFG 1 the sampler ignores it.
Full changelog: https://github.com/perfectgf/lora-dataset-studio/compare/v2026.11.09...v2026.11.10
- v2026.11.09LoRA Dataset Studio V2 — v2026.11.092026-11-07 · 2026-11-07 21:15 UTC
查看更新说明
LoRA Dataset Studio V2
V2 is the maintained LDS release. Existing ZIP installations can upgrade from Settings → Maintenance → Update & restart, keeping datasets, media and history. Git installations still on the former
mainbranch must first switch tov2. Optional features are plugins: open Plugins → Store after updating, install the ones you use and review their settings before the first run.Install the core, then choose your plugins
Download
LoRA-Dataset-Studio-windows.zipbelow, extract it into a new folder and double-clickstart.bat. The launcher prepares a compatible Python environment. Complete the core Setup, then open Plugins → Store to install the features you want. Each plugin has its own settings and preparation steps, including required ComfyUI custom nodes. Importing and organising datasets does not require a plugin.Free public plugins are available through the Store. Each plugin describes its features and preparation requirements. Paid API engines and rented GPUs still charge separately when you choose to use them.
The core and these public plugins are available at no charge, with their source published under the project's PolyForm Noncommercial license. New LDS versions reach Patreon members (LDS Member) first, then everyone, and some plugins are members-only; the version you have stays free and keeps working. Usage statistics are optional and off by default; the app asks before sharing, and you can change your choice in Settings → Maintenance.
Two ways to install and update
ZIP installations: Update & restart installs the next published release, preserving your data and local configuration. The ZIP below contains the core; plugins are installed and updated separately through the Store.
Git installations:
v2is the default and maintained branch. Update & restart follows the commits of your checkout's configured branch. The formermainbranch is now the frozenv1branch; it receives no further updates. Follow the migration instructions if your checkout still tracksmain.git clone https://github.com/perfectgf/lora-dataset-studio.git cd lora-dataset-studio start.bat🎁 What's new in v2026.11.09
Your Bank shows HEIC and AVIF photos too
A Bank over a folder of iPhone photos or AVIF images used to skip them. They are now inventoried, shown, scored, face-sorted and promoted like any other image, through a lossless PNG copy the app keeps in its own data folder. Your files are never touched.
iPhone photos (HEIC) and AVIF images import into your datasets
HEIC/HEIF photos and AVIF images were refused at import. They now import from the dropzone, a ZIP or a folder, with or without auto head-crop, and work as a reference too. The trainers cannot read them, so the default original-file mode keeps each one as a lossless PNG.
A queued training that cannot start says why
When a training waiting in the queue failed to start when its turn came, it left the queue without a word. The Training panel now names it and gives the reason.
A queued training waits for the GPU instead of vanishing
A training in the queue could disappear when its turn came while ComfyUI was generating or a vision pass held the GPU. It now keeps its place, the queue says what it is waiting for, and it starts on its own once the GPU is free.
Full changelog: https://github.com/perfectgf/lora-dataset-studio/compare/v2026.11.08...v2026.11.09
- v2026.11.08LoRA Dataset Studio V2 — v2026.11.082026-11-07 · 2026-11-07 12:15 UTC
查看更新说明
LoRA Dataset Studio V2
V2 is the maintained LDS release. Existing ZIP installations can upgrade from Settings → Maintenance → Update & restart, keeping datasets, media and history. Git installations still on the former
mainbranch must first switch tov2. Optional features are plugins: open Plugins → Store after updating, install the ones you use and review their settings before the first run.Install the core, then choose your plugins
Download
LoRA-Dataset-Studio-windows.zipbelow, extract it into a new folder and double-clickstart.bat. The launcher prepares a compatible Python environment. Complete the core Setup, then open Plugins → Store to install the features you want. Each plugin has its own settings and preparation steps, including required ComfyUI custom nodes. Importing and organising datasets does not require a plugin.Free public plugins are available through the Store. Each plugin describes its features and preparation requirements. Paid API engines and rented GPUs still charge separately when you choose to use them.
The core and these public plugins are available at no charge, with their source published under the project's PolyForm Noncommercial license. New LDS versions reach Patreon members (LDS Member) first, then everyone, and some plugins are members-only; the version you have stays free and keeps working. Usage statistics are optional and off by default; the app asks before sharing, and you can change your choice in Settings → Maintenance.
Two ways to install and update
ZIP installations: Update & restart installs the next published release, preserving your data and local configuration. The ZIP below contains the core; plugins are installed and updated separately through the Store.
Git installations:
v2is the default and maintained branch. Update & restart follows the commits of your checkout's configured branch. The formermainbranch is now the frozenv1branch; it receives no further updates. Follow the migration instructions if your checkout still tracksmain.git clone https://github.com/perfectgf/lora-dataset-studio.git cd lora-dataset-studio start.bat🎁 What's new in v2026.11.08
A stopped local training keeps its checkpoints in Runs
Stopping a local training showed it as a crash, with an error and ↻ Retry, and the Runs page offered no ◉ Graph for a local run on its own. A Stop now reads as a stop, and every local run opens its ◉ Graph with the checkpoints it saved.
Full changelog: https://github.com/perfectgf/lora-dataset-studio/compare/v2026.11.07...v2026.11.08
用户点评
登录用户的真实使用记录,可按版本和硬件筛选还没有人点评。第一个说出真实体验的人最有价值。
[!WARNING] No affiliation with loradataset.com. That website is not operated, endorsed, or supported by the LoRA Dataset Studio team. Any payments made to that service do not support this project. Use this repository and the links provided here to find our official downloads and community channels.
LoRA Dataset Studio V2
Free local LoRA training studio: turn one reference photo or a whole image folder into a captioned LoRA dataset, train character, style and concept LoRAs for Krea 2, FLUX.1, FLUX.2 Klein, Qwen-Image 2.1, Z-Image, SDXL and Anima, then pick the best checkpoint side by side. LDS runs on your machine in one browser tab and uses ai-toolkit for training and ComfyUI for local generation and checkpoint testing, without a node graph to wire.
Install · Documentation · Plugins · Releases · Discord
[!IMPORTANT] New LDS versions reach Patreon members first. Each new version is available early to Patreon members (LDS Member), then to everyone; exclusive plugins stay members-only. The version you have stays free and keeps working.
The current core and public plugins are free under the PolyForm Noncommercial license, and running them needs no account; optional usage statistics are off by default. External APIs and rented GPUs have their own charges. A Patreon membership is needed for new versions and exclusive plugins, not to keep using the version you have.
A complete Character LoRA workflow in seven minutes:
https://github.com/user-attachments/assets/d51ff89c-34e9-41a9-b47d-08939a8c867b
People shown in the demo and screenshots are AI-generated.
What it does
| Area | Capabilities |
|---|---|
| Datasets | Character, Concept and Style workflows; image and ZIP/folder import; subject-aware shot catalogs; local Klein and Krea 2 Edit generation; reference editing and exact retries |
| Image Bank | Review folders in place, score quality, group duplicates and people, search by text or similarity, and build balanced shortlists to promote into datasets |
| Curation | Keep/reject, crop, mirror, rotate, face similarity, composition checks, editable watermark masks, text detection and recoverable cleaning |
| Captioning | Local vision models or JoyCaption, family-appropriate prose or tags, appearance policies, bulk editing, targeted re-captioning and external image/.txt round trips |
| Local training | Guided ai-toolkit recipes for Z-Image, SDXL, Krea 2, FLUX.1, FLUX.2 Klein, Anima and Qwen-Image 2.1; queues, advanced settings, checkpoint continuation and experimental slider LoRAs |
| Review and testing | Generate with all seven image training families; fixed-seed checkpoint/strength comparisons, prompt batches, runs, logs, lineage, votes, rankings and a generated-image Gallery |
| Files | Standard training ZIPs and sidecars, backup/restore without API keys, ComfyUI deployment, configurable storage and Trash |
Dependencies vary by feature. Bank search ranks matches; it does not guarantee exclusions. Undo covers specific actions, and Delete rejected can remove source files after confirmation. Video and slider workflows are experimental. See the feature reference, requirements and known limitations for the detailed behavior.
Plugins
Install optional features from Plugins → Store, configure and prepare them in their own settings, and update them through Plugins → Updates. Several plugins can be installed together with one LDS restart. Each is independently installable; model downloads, hardware and provider credentials depend on the feature.
| Plugin | What it does | Main requirements |
|---|---|---|
| API image engines | Generate images with Gemini/Nano Banana, ChatGPT or OpenRouter; experimental ChatGPT subscription connection | Provider credentials; API usage is billed by the provider |
| Camera angles | Create new viewpoints of Gallery or dataset images with controls for direction, height and distance | ComfyUI, Qwen-Image-Edit and Multiple-Angles LoRA |
| Canvas | Arrange runs and checkpoints on a visual board, compare or blend LoRAs, generate images and export layouts | ComfyUI for generation; one model family per generation batch |
| Cloud training | Train image or video models on rented GPUs, monitor runs, recover checkpoints and deliver full Krea 2 models | vast.ai account and API key; Hugging Face access for applicable runs |
| DLSS 5 Neural Rendering | Improve a finished video's lighting and materials, compare with the original and export the result | Windows x86-64, NVIDIA driver, compatible model supplied by you and the prepared bridge |
| Klein Improve | Re-render image detail with instructions, LoRA presets and finishing controls | ComfyUI and Klein models; appearance and color can change |
| Live channels | Continuously render scripted H3 scenes and watch them in a browser or VLC | Compatible local ComfyUI, H3 weights and stream encoder; experimental |
| Model tools | Quantize full models to fp8 or merge weighted LoRAs into a base checkpoint | Python with PyTorch and output disk space; CPU processing, no GPU required |
| Publish to Civitai | Upload checkpoints and generated images to a model page | Civitai API key; checkpoints default to drafts, image posts default to publication |
| Publish to Hugging Face | Export kept images, captions, metadata and a dataset card to the Hub | Write-enabled Hugging Face token; private repository by default and rights confirmation |
| Resource monitor | Show CPU, GPU, RAM, VRAM and temperatures, with guarded memory release | No model download |
| SeedVR2 | Restore and upscale images with high-resolution tiling and finishing controls | ComfyUI, SeedVR2 nodes and models; optional TTP nodes for tiling |
| Video lane | Build video datasets from local files or web imports, curate shots, train LoRAs and test H3 generation, continuation and interpolation | Dependencies vary by task: PyAV/ffmpeg, ai-toolkit, ComfyUI and model weights; Video Bank remains beta |
| Web scraping | Import selected images or clips from searches and supported gallery URLs into datasets or banks | Source-dependent credentials and permissions; Pexels requires explicit dataset/ML authorization |
API providers apply their own billing and content policies. Model licenses also apply, including MiniMax H3's territory restrictions; check the video limits before using it. Plugin authors can start with the SDK and package guide.
Screenshots
Click a screenshot to view it at full size.
Setup & install
Windows: download LoRA-Dataset-Studio-windows.zip from the latest release, extract it into a new folder and run start.bat. The launcher prepares Python and opens LDS in your browser.
Complete Setup, then create a dataset or install the plugins you need. Importing, organizing and manually captioning images require no GPU or API key.
| Installation | Instructions |
|---|---|
| Git checkout or manual Python environment | Native installation |
| Docker with an existing or fresh ComfyUI | Docker guide |
| Docker without a GPU | API-only setup |
| Pinokio | One-click installation |
| Rented RunPod GPU | RunPod guide |
Updates: new LDS versions reach Patreon members first, then everyone, and install from Settings → Maintenance → Updates. Git and Pinokio checkouts fast-forward to a verified release; ZIP installations swap in a verified archive and retain their datasets, media, settings and history. Docker installations from the Windows ZIP install them from the host with update-docker.bat patreon. Without a membership, the installed version keeps working; v1 is frozen and old main installations must migrate to V2. Pinokio and Docker keep their own update procedures.
Minimum requirements
The core runs without a GPU. Python 3.10–3.12 supports the local ML extras; the Windows launcher downloads Python 3.12 if needed. Local generation typically needs about 16 GB NVIDIA VRAM; training requirements depend on the family and settings. See the hardware and dependency tables before downloading models or renting a GPU.
Documentation
- Getting started and end-to-end workflow
- Task instructions and dataset quality
- Settings, models and paths, external tools and API keys
- Troubleshooting and known limitations
The documentation index includes development and plugin references.
LoRA training FAQ
These are the questions that come up most in LoRA training threads, with what LDS does about each. Each question opens its full answer, and the dataset guide has the details.
How do I train a consistent character LoRA from one or a few photos?
Create a Character dataset and set a reference photo. LDS proposes a shot catalog for the subject (face, bust, body and back framings from several angles) and generates the missing shots locally with Klein or Krea 2 Edit, which restages the reference while preserving identity. InsightFace similarity and score-based auto-triage help you reject shots where the face drifts, and the Camera angles plugin adds new viewpoints. Your own photos work too: import them, then let coverage name what the set never shows, such as profile views or a second outfit.
How many images do I need to train a LoRA?
About 25 varied images make a balanced character LoRA, and they beat 60 near-duplicates. Minimums depend on the family: 12 for Z-Image, 15 for Krea 2, FLUX.1 and FLUX.2 Klein, 20 for SDXL. Balance face, bust, body and back shots, because a set of close-ups gives a LoRA that falls apart on full-body prompts; the composition meter on the dataset page tracks that balance. See how many images, and which ones.
Tags or natural language captions? Where does the trigger word go?
Match the caption style to the family: natural-language prose for Krea 2, FLUX.1, FLUX.2 Klein and Z-Image, booru tags for SDXL, either for Anima. LDS writes captions in that style with JoyCaption or a local vision model and puts the trigger word at the start of every caption on export. Describe what should vary (place, light, outfit, pose) and leave out what should stick to the trigger: per-trait appearance policies decide whether hair, makeup or glasses are described. See captions.
Which base model should I train my LoRA on?
Choose the family before captioning, because it sets the caption style, the image count and the settings. In short: Krea 2 has the highest realism ceiling, FLUX.1 the largest LoRA ecosystem, FLUX.2 Klein 4B trains on mid-range GPUs, Z-Image iterates fastest, SDXL suits booru-native checkpoints and Anima targets anime. Qwen-Image 2.1 is supported too. See the family comparison.
What steps, rank and learning rate should I use?
Start from the guided recipe for your family: LDS fills in the ai-toolkit settings and scales steps with the dataset, about 120 per image for a character LoRA, kept between 1,500 and 3,500, because a fixed 3,000 overcooks small sets. Raise the rank (48 to 64) only for a hard identity and a bigger dataset; a high rank on 15 images memorizes them. Advanced options stay editable, and a run can continue from any saved epoch. See the settings cheat-sheet.
Which LoRA trainer should I use? Do I need the command line?
LDS trains with ai-toolkit. Once ai-toolkit is installed and set in Settings → Local tools, LDS writes the training config: pick a recipe and start the run from the browser, with the queue and logs on the runs page. Datasets also export as standard ZIPs with .txt caption files, so the same set can feed kohya_ss, OneTrainer or another trainer.
How much VRAM do I need to train a LoRA?
Importing, sorting and manual captioning need no GPU. Auto-captioning takes about 8 GB of VRAM and local generation about 16 GB. LoRA training needs 16 GB or more for Z-Image (quantized, low-VRAM mode) and SDXL, 16 to 24 GB for FLUX.2 Klein 4B (32 to 48 GB for 9B), 24 GB for Krea 2 at 1024 px (under 24 GB, train at 768 px only) and about 24 GB for FLUX.1 and Anima. Short on VRAM? The Cloud training plugin trains supported families on rented vast.ai GPUs, and LDS itself can run on a RunPod GPU. See the hardware tables.
How do I pick the best epoch or checkpoint?
Open the run in Test Studio: it renders the checkpoints you pick at several strengths with the same seed and prompts, so the images differ only by checkpoint and strength. Vote and rank them, compare several LoRAs or stack them, and test several prompts in one launch. See after training.
Can I train a style or concept LoRA?
Yes: Style and Concept datasets have their own workflows. Concept captions must be prose rather than booru tags, steps grow more slowly with the image count (about 475 × √n, from 2,000 to 12,000), and the guide explains how to keep faces out of a concept.
Can I train a video LoRA (Wan 2.2, LTX-2, MiniMax H3)?
Experimentally, with the Video lane plugin (beta). It cuts clips to the frame counts each model accepts (4n+1 for Wan, 8n+1 for LTX), measures silence and audio level, shows burned-in subtitles and letterbox bars before training, and trains with ai-toolkit locally or on a rented GPU. Nine targets are offered across Wan 2.1, Wan 2.2, LTX-2, LTX-2.3 and MiniMax H3, but only Wan 2.2 14B has been through a finished run so far, filling 24 GB with CPU offload.
Do I need to know ComfyUI?
You need a ComfyUI install for local generation, not its node editor. LDS sends its own workflows for dataset shots, edits, upscaling and checkpoint tests, and Setup downloads the generation models. With Docker, the launcher can set up a fresh, isolated ComfyUI.
How do I upscale or restore images locally?
Install SeedVR2 to restore and upscale with high-resolution tiling, or Klein Improve to re-render detail from instructions (appearance and color can change). Both run on your local ComfyUI.
Configuration & network access
Use Settings for normal configuration. Native installs bind to 127.0.0.1 by default. Read the security policy before enabling network access; a protected connection also lets you use LDS from a phone or tablet.
Optional usage statistics are off by default. Update checks, requested model downloads, configured providers and scraping can contact external services. The network and privacy guide describes these connections, the data shared and public-access settings.
Support the project
A Patreon membership (LDS Member) funds development and includes new features, exclusive plugins and LDS updates. GitHub Sponsors also supports development, API testing and rented test GPUs. Bug reports, contributions and sharing the project also help. For support, generate a diagnostic report under Guide → Getting help, then use Discord or GitHub issues.
Affiliate disclosure: the project's vast.ai links pay the project 3% of referred users' spending for the lifetime of their account, at no extra cost. Rentals use your own API key and are billed directly by vast.ai. This is separate from optional usage statistics.
Legal & responsible use
Use material you have the rights and consent to train on. Non-consensual likeness use, impersonation, fraud, and sexual or exploitative content involving minors are prohibited. You remain responsible for datasets and outputs. See responsible use for consent, privacy, copyright, platform terms and warranty details.
Contributing
See CONTRIBUTING.md for development and pull requests, and the Code of Conduct for community rules. Report vulnerabilities privately as described in SECURITY.md.
License
PolyForm Noncommercial 1.0.0. Noncommercial use is permitted; commercial use requires separate permission from the licensor.