Improved user experience when working with the gallery and inpaint, adding support for Forge/Reforge, Hires. Fix, ADetailer, and Fal AI. #614
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
… and Logger export
- Add tests for FalAiEndpointRepositoryImpl, FalAiGenerationRepositoryImpl, ForgeModulesRepositoryImpl - Add tests for TestFalAiApiKeyUseCaseImpl, GetForgeModulesUseCaseImpl, GetGalleryPagedIdsUseCaseImpl - Add tests for FalAiGenerationUseCaseImpl, ConnectToFalAiUseCaseImpl - Fix existing tests: add mediaFileManager, falAiApiKey, mediaPath/inputMediaPath mocks - Add mock files for FalAi and ForgeModule entities
- GetStableDiffusionModelsUseCaseImpl: only fetch when A1111 is active - FetchAndGetSwarmUiModelsUseCaseImpl: only fetch when SwarmUI is active - FetchAndGetHuggingFaceModelsUseCaseImpl: only fetch when HuggingFace is active - FetchAndGetStabilityAiEnginesUseCaseImpl: only fetch when StabilityAI is active This prevents unnecessary network requests to servers that are not currently configured as the active generation source.
Add model name field to generation results and display it in the gallery detail view. This allows users to see which model was used for each generated image. Changes: - Add modelName field to AiGenerationResult entity - Add MODEL_NAME column to GenerationResultEntity (DB migration v9) - Display model name in GalleryDetailScreen when available - Update test mocks
Add new LOCAL_QUALCOMM_QNN server source that enables on-device Stable Diffusion generation using Qualcomm QNN SDK with NPU (HTP) acceleration. Supports Snapdragon 8 Gen 1 and newer chipsets. Features: - NPU-accelerated inference using Qualcomm HTP backend - GPU fallback with OpenCL support (MNN backend) - Built-in models from HuggingFace (xororz/sd-qnn) - Custom model support with folder scanning - Multiple resolution presets (512x512, 768x768, etc.) - Configurable scheduler (DPM, Euler) - Real-time diffusion process visualization - Img2Img support with denoising Technical changes: - Add feature/qnn module with LocalDream-based implementation - Add DeviceChipsetDetector for Snapdragon detection - Add QNN preferences (model, runtime, scheduler, etc.) - Add QnnForm for server setup configuration - Add QnnRuntimeSelectionComponent for runtime switching - Add model scanning for custom QNN models - Update generation usecases to support QNN backend - Add prepare_qnn_libs.sh script for library setup Note: QNN libraries must be extracted separately using the prepare_qnn_libs.sh script (not included in repository).
- Add fal.ai as supported backend with FLUX models - Document local diffusion options (ONNX, MediaPipe, QNN) - Add QNN setup instructions for Snapdragon devices - Update feature list with ADetailer, Hires.Fix, Forge modules - Document gallery improvements and file-based storage
Update test mocks and fixtures to support QNN repository and use cases: - Add QnnGenerationRepository mocks to generation use case tests - Add GetForgeModulesUseCase to generation viewmodel tests - Add QNN preferences to configuration tests - Fix mapper tests for model name field - Add missing parameters to presentation viewmodel tests
crim50n
force-pushed
the
master
branch
2 times, most recently
from
358e155 to
a2e5954
Compare
- Rename log tag [SDAI] -> [PDAI] - Rename log file sdaiv1.log -> pdaiv1.log - Update custom model path SDAI -> PDAI - Rename SdaiWorkerFactory -> PdaiWorkerFactory - Update test stubs naming - Add CHANGES.md with 0.7.0 release notes
- Add gradle-versions-plugin for dependency updates checking - Add version-catalog-update-plugin for automated updates - Update stable dependencies to latest versions: - compileSdk: 35 -> 36 (required by new AndroidX libs) - Compose BOM: 2025.03.01 -> 2025.12.01 - AndroidX core-ktx: 1.15.0 -> 1.17.0 - AndroidX lifecycle: 2.8.7 -> 2.10.0 - AndroidX navigation: 2.8.9 -> 2.9.6 - AndroidX work: 2.10.0 -> 2.11.0 - Koin: 4.0.4 -> 4.1.1 - Retrofit: 2.11.0 -> 3.0.0 - ONNX: 1.21.0 -> 1.23.2 - Material: 1.12.0 -> 1.13.0 - And other library updates - Pin critical versions (paging, agp, kotlin, ksp) to prevent breaking changes - Keep minSdk=24 for Android 7.0+ compatibility
crim50n
force-pushed
the
master
branch
3 times, most recently
from
ea4dd38 to
77b538b
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Key Changes:
ZoomableImagecomponent for smoother zooming and gestures.Screenshots / Video (Optional)
VN20251230_181416.mp4
VN20251230_181241.mp4
Checklist
Tested On