install-local-gpu.sh and bring-up-local-gpu.sh now call
prefetch-models.sh --sdxl (host-side download, all models including
SDXL/inpaint, ~13GB) automatically rather than treating it as a manual
DNS-troubleshooting step. install-local-gpu.sh drops back to the
invoking user via sudo -u "$SUDO_USER" -H before running it, since
that script itself runs as root and would otherwise leave downloaded
files root-owned. Both call sites are non-fatal: a prefetch failure
logs a warning and the container still starts, falling back to its
own in-container download attempt.
Updated README to match.
Lets SAM, U2Net, BEN2, and BiRefNet-HR (optionally SDXL via --sdxl) be
downloaded outside Docker into ./data/, which is already bind-mounted
into the GPU container — so a blocked container network no longer blocks
first-run setup. Reuses the existing dual-mode download_sam_model.py and
download_u2net_model.py as-is. For the HuggingFace Hub models, sets
HF_HOME (rather than --cache-dir) so the host-side cache layout matches
the container's default ~/.cache/huggingface resolution exactly, avoiding
a path-nesting mismatch between the two.
Wired into install-local-gpu.sh's completion banner and
bring-up-local-gpu.sh's header, and referenced from the relevant README
troubleshooting sections and the hf_cache bind-mount comment in
docker-compose.gpu.yml.
BEN2 becomes the new default local backend (clean cutouts, strong on
hair/fur edges), with BiRefNet-HR available as a high-res/print
alternate and U2Net kept as the lightweight fallback. Both are
MIT-licensed and download weights from HuggingFace on first use
(cached via the existing hf_cache bind mount), unlike U2Net/SAM which
need an explicit download script.
- config: new BG_REMOVAL_MODEL setting (default "ben2")
- tools.py: remove-background-base64 now tries local backends in
order (request.model override > BG_REMOVAL_MODEL > ben2/u2net),
falling back to rembg's birefnet-general session as a last resort
- requirements.gpu.txt / Dockerfile.gpu: add ben2 + transformers deps
needed for the new backends, with a build-time smoke test for ben2
- frontend: model dropdown in the Remove Background dialog, threaded
through api.js to the new request field
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Ro4PwQKvSc3CH19LSN21Ht
U2Net only ever downloaded lazily on the first Remove Background click,
unlike SAM which retries on every container start. If that one attempt
failed (DNS/firewall) the model was never fetched again, surfacing as
"No background removal method available. Install u2net or rembg."
Mirrors the existing SAM auto-download/AUTO_DOWNLOAD_SAM pattern for
U2Net, and documents manual host-side recovery in the README.
Also deletes backend/app/services/u2net_model.py (hand-written
U2NET/U2NETP PyTorch classes) — unused since tools.py switched to
cv2.dnn.readNetFromONNX for background removal.
The iptables DOCKER-USER rule is lost on reboot; the script re-applies
it each run, checks for duplicates, and is silently skipped on macOS/WSL.
Default behaviour (no args): docker compose up -d --build.
All docker compose subcommands can be passed as args (logs, down, etc.).
README Quick Start and Updates sections now reference ./start-gpu.sh.
https://claude.ai/code/session_01WVDg7amsy1TTtxvpku7bcM
iptables fix is simpler (one command, no 13 GB download) and does not
affect container isolation — adds clarifying note so users understand
it only restores Docker's default outbound DNS behaviour.
Docker helper container kept as fallback option.
https://claude.ai/code/session_01WVDg7amsy1TTtxvpku7bcM
Replaces 'pip install huggingface-hub' (breaks on PEP 668 / Debian 12+)
with a docker run --rm python:3.11-slim one-liner that downloads directly
into ./data/hf_cache without touching host Python packages.
https://claude.ai/code/session_01WVDg7amsy1TTtxvpku7bcM
Build / pip layer fixes:
- Add BUILDID ARG to Dockerfile.gpu; pass from docker-compose.gpu.yml build args
so pip layers can be force-busted without --no-cache:
BUILDID=$(date +%s) docker compose -f docker-compose.gpu.yml up --build
Model download (DNS-blocked environments):
- Change HF model cache from named volume to ./data/hf_cache bind mount
so models can be pre-downloaded on the host (no rebuild needed)
- Remove now-unused hf_model_cache named volume
- README: add iptables fix + huggingface-cli offline download instructions
Error handling improvements:
- ai_edit_region: catch ConnectError/Errno-3 → return 503 with exact fix commands
- _require_remote: give actionable message when local_gpu provider fails to load
- _build_provider: catch AttributeError (torch.xpu from wrong diffusers) not just ImportError
- local_diffusion.py: fix docstring to reflect <0.29.0 pin
https://claude.ai/code/session_01WVDg7amsy1TTtxvpku7bcM
- Pin diffusers to >=0.28.0,<0.29.0 to avoid AttributeError on torch.xpu
(diffusers 0.29+ requires PyTorch 2.4 but base image ships 2.1.2)
- Pin transformers to <4.40.0 to match
- README: add SAM offline download troubleshooting for DNS-blocked containers
- SelectionActions panel: 'Selection ready' title + subtitle makes clear
nothing has fired yet; AI actions show inline description (not just tooltip);
section labels 'AI Actions' / 'Classic Tools'; scale hint updates live;
Enter key submits AI edit prompt; _actionCard hover border for clickability
https://claude.ai/code/session_01WVDg7amsy1TTtxvpku7bcM
Old README described the original React/Fabric.js UI and listed
"No local GPU inference" as a non-goal. Updated to reflect:
- miniPaint-based editor with SAM brush selection
- GPU quick-start (nvidia-container-toolkit prereqs, docker-compose.gpu.yml)
- Cloud API quick-start
- GPU tier auto-selection table (FLUX/SDXL/SD by VRAM)
- Full feature list (selection actions, print tools, progress bars)
- Correct clone URL and update commands
- Troubleshooting section
https://claude.ai/code/session_01WVDg7amsy1TTtxvpku7bcM
This commit implements a full-stack AI photo editing application that
allows users to regenerate only selected areas of images using AI.
Features implemented:
- Frontend (React + Fabric.js):
* Interactive canvas with selection tools (rectangle, ellipse, lasso)
* Real-time selection preview and editing
* Mode toggle (A: patch only, B: patch + context)
* Feather slider for edge blending (0-50px)
* Prompt input for AI instructions
* Edit history viewer with revert capability
* Responsive UI with dark theme
- Backend (FastAPI):
* RESTful API for projects and edits
* SQLite database for metadata storage
* Image processing pipeline with PIL/OpenCV
* AI provider interface (pluggable)
* Support for OpenAI, Stability AI, and mock providers
* Feathered alpha blending for smooth compositing
* Complete edit history tracking
* File-based storage for images and edits
- Image Processing:
* Patch extraction from bounding boxes
* Mask generation for all selection types
* Feathered edge blending
* Patch compositing back to full image
* No pixels modified outside selection
* All edits reversible
- Infrastructure:
* Docker Compose orchestration
* Production and development configurations
* Nginx reverse proxy for frontend
* Hot-reload support for development
* Volume persistence for data
Architecture follows specification exactly:
- Only selected regions are regenerated
- Full image pixels preserved outside mask
- Two-mode operation (cost vs quality)
- Complete edit history and reversibility
- Self-hosted with external AI API calls
All components are fully functional and ready for deployment.
Claude