AIWF Studio
A local-first Stable Diffusion workspace under active development. The public repo documents image, inpaint, model browsing, video routes, local runtime folders, and the difference between stable sharing and experimental work.
AI Without Fear open lab
Local AI work people can inspect.
AIWF is Shawn O'Hagan's public lab for local AI tools, grounded retrieval, workflow guides, and experiments that run on consumer hardware. This page collects the work that can help people learn, test, and decide what to try.
// local AI
AIWF Studio
Atlas
RAG
ComfyUI notes
consumer GPUs
Open lab
Guides and experiments
AI Without Fear Atlas
A public field manual for local AI workflows, ComfyUI notes, grounded retrieval, Gradio patterns, model serving, and setup problems that assistants often get wrong.
Atlas LoRA Adapter
A research and evaluation repo for testing whether a small adapter can follow source rules, lane labels, compact evidence packs, and refusal behavior without treating the corpus as memorized knowledge.
What visitors can try
Visitors can read the guides, review the repositories, copy public workflow patterns, and test local setup ideas. The stable download lane stays closed until a release branch is ready.
Hardware notes
What people need to run it
Read the guides
A browser is enough for the public guides and repo notes. No GPU is needed to learn from the written material.
Run local image tools
AIWF Studio targets local Windows workflows. A recent NVIDIA GPU, Python, Git, disk space for models, and patience with model setup are the practical baseline.
Consumer GPU range
Smaller image tests can fit on modest cards. Heavier SDXL, Flux, video, and post-processing routes need more VRAM and more disk. Treat 8 GB as a narrow test lane and 16 GB or more as the safer local target.
Training experiments
Training and QLoRA tests need stricter setup than guide reading. Plan for an NVIDIA GPU, enough VRAM for the base model, clean datasets, saved eval records, and fallback to rented compute when local hardware is too small.
Downloads
Listed now, linked after release
Downloads are intentionally not linked yet. The public download path will point to a stable branch or release artifact so day-to-day app work cannot break what visitors receive.
AIWF Studio stable package
Planned download for the stable sharing branch.
Gate: branch and smoke receipt
AIWF PDF guide pack
Planned public guide bundle for local AI setup and workflow notes.
Gate: copy review and version label
Workflow examples
Planned examples for public experiments that are safe to share.
Gate: license and file review
Support
Open-source support stays simple
Tip the open-source work
Small tips support the public AIWF tools, guides, testing, and experiments. The current tip path lives on Shawn's GitHub profile.
Investor and sponsor conversations
Funding conversations now have their own page. That keeps AIWF focused on public guides and downloads while giving serious company backing a cleaner path.