Merge pull request #11 from zb-ss/feature/gpu-reexec-selection

feat(gpu): add GPU selection with device passthrough for AMD multi-GPU

Merge pull request #10 from zb-ss/feature/ltx2-docker-improvements

feat: add LTX-2 model support and improve Docker interactive mode

chore: bump version to 0.6.0

Includes multi-GPU support and video output fix since v0.5.0.

Merge pull request #6 from zb-ss/feature/add-flux2-klein

feat: add FLUX.2 Klein 9B image-to-image model

Merge pull request #4 from zb-ss/fix/interactive-prompt-navigation

fix: enable proper text editing in interactive prompt inputs

Merge pull request #3 from zb-ss/feature/add-qwen-image-2512

feat: add Qwen-Image-2512 text-to-image model

Merge pull request #2 from zb-ss/bugfix/interactive-mode-fixes

fix: Interactive mode bug fixes and improvements

Merge pull request #1 from zb-ss/feature/ux-improvements

feat: UX Improvements v0.4.0 (Phases 1-4 + Interactive Wizard)

v0.3.1 - Update author email

v0.3.0 - Qwen Image Edit support

Features:
- Add Qwen Image Edit (Qwen/Qwen-Image-Edit-2511) as default i2i model
- Multi-image input support for combining images
- Multi-GPU support with HFTOOL_MULTI_GPU env var
- Auto-upgrade dependencies when downloading models
- Suppress noisy warnings by default (HFTOOL_DEBUG=1 to show)
- File logging with HFTOOL_LOG_FILE

Fixes:
- Auto-open now works correctly for i2i tasks with JSON input
- Skip dependency install if already satisfied
- Better output filename generation from JSON input