### Setup the benchmarking environment Source: https://github.com/akitaonrails/franksherlock/blob/master/_face_ab_test/README.md Commands to navigate to the directory, create a virtual environment, and install necessary dependencies. ```bash cd _face_ab_test python -m venv .venv source .venv/bin/activate pip install -r requirements.txt ``` -------------------------------- ### Build and run the application Source: https://github.com/akitaonrails/franksherlock/blob/master/README.md Commands to download native dependencies, install frontend packages, and start the development environment. ```bash # 1. Download native libraries for your platform cd sherlock/desktop/src-tauri bash scripts/download-pdfium.sh bash scripts/download-onnxruntime.sh # 2. Install frontend dependencies cd sherlock/desktop npm install # 3. Start Ollama (in a separate terminal) ollama serve # 4. Run in dev mode npm run tauri:dev ``` -------------------------------- ### Run Development Environment Source: https://github.com/akitaonrails/franksherlock/blob/master/sherlock/README.md Commands to install dependencies and launch the Tauri development environment. ```bash npm install npm run tauri:dev ``` -------------------------------- ### Build and Test Commands Source: https://github.com/akitaonrails/franksherlock/blob/master/CLAUDE.md Standard commands for installing dependencies, building the Rust backend, and running tests for the frontend and backend. ```bash npm install # frontend deps cargo build # rust backend (from src-tauri/) cargo test # 322 unit tests npm run tauri:dev # launch dev mode npm run test # 299 frontend tests npm run tauri:build # produce AppImage/DMG/MSI ``` -------------------------------- ### Define ground truth JSON format Source: https://github.com/akitaonrails/franksherlock/blob/master/_face_ab_test/README.md Example structure for the ground truth JSON file used to calculate precision, recall, and clustering metrics. ```json { "images": { "photo1.jpg": { "faces": [ { "bbox_approx": [x, y, width, height], "person": "Alice" } ] } }, "expected_clusters": { "Alice": ["photo1.jpg:0", "photo3.jpg:0"], "Bob": ["photo1.jpg:1", "photo2.jpg:0"] } } ``` -------------------------------- ### Produce a release binary Source: https://github.com/akitaonrails/franksherlock/blob/master/README.md Command to build the production-ready application binary. ```bash cd sherlock/desktop npm run tauri:build ``` -------------------------------- ### Verify the environment Source: https://github.com/akitaonrails/franksherlock/blob/master/_face_ab_test/README.md Execute the verification script to check Python version, GPU availability, and model downloads. ```bash python scripts/phase0_verify_env.py ``` -------------------------------- ### Run Qwen Model Comparison Script Source: https://github.com/akitaonrails/franksherlock/blob/master/_research_ab_test/docs/RESULTS.md Executes a comparative benchmark between different Qwen model versions using the specified script and parameters. ```bash uv run python -u scripts/phase2e_qwen_size_compare.py --models qwen2.5vl:7b,qwen3-vl:8b,qwen3-vl:30b-a3b --repeats 3 ``` -------------------------------- ### CI/CD Build and Test Commands Source: https://github.com/akitaonrails/franksherlock/blob/master/README.md Lists the commands executed by GitHub Actions for code quality, testing, and building. ```bash cargo test cargo clippy cargo fmt --check npm run build npm run test cargo audit ``` -------------------------------- ### Build AppImage Source: https://github.com/akitaonrails/franksherlock/blob/master/sherlock/README.md Command to generate an AppImage bundle for the desktop application. ```bash npm run tauri:build ``` -------------------------------- ### Run project tests Source: https://github.com/akitaonrails/franksherlock/blob/master/README.md Commands to execute Rust backend tests and frontend unit tests. ```bash # Rust (322 tests) cd sherlock/desktop/src-tauri cargo test # Frontend (299 tests) cd sherlock/desktop npm run test ``` -------------------------------- ### Run Development with Wayland/NVIDIA Workaround Source: https://github.com/akitaonrails/franksherlock/blob/master/sherlock/README.md Environment variables to bypass WebKit rendering issues on specific Linux configurations. ```bash WEBKIT_DISABLE_DMABUF_RENDERER=1 GDK_BACKEND=wayland,x11 npm run tauri:dev ``` -------------------------------- ### Run detection benchmarks Source: https://github.com/akitaonrails/franksherlock/blob/master/_face_ab_test/README.md Commands to execute the detection phase, with options to specify detectors or limit the number of images processed. ```bash python scripts/phase1_detection.py python scripts/phase1_detection.py --detectors scrfd mtcnn # specific detectors python scripts/phase1_detection.py --limit 10 # first 10 images ``` -------------------------------- ### Project Directory Structure Source: https://github.com/akitaonrails/franksherlock/blob/master/README.md Visual representation of the main application directory layout. ```text sherlock/ <- Main application desktop/ src-tauri/src/ <- Rust backend classify.rs <- Ollama vision + Surya OCR pipeline thumbnail.rs <- Thumbnail generation + dHash computation scan.rs <- Incremental scanner (4-phase) with cancellation + resume db.rs <- SQLite + FTS5 + duplicate queries similarity.rs <- dHash + description similarity + Union-Find grouping face.rs <- Native ONNX face detection (SCRFD) + recognition (ArcFace) pdf.rs <- PDFium text extraction + page rendering video.rs <- ffmpeg metadata, keyframe extraction, subtitle parsing video_server.rs <- Localhost HTTP Range streaming for video preview config.rs <- App paths lib.rs <- Tauri commands, auto-cleanup query_parser.rs <- NL query parsing runtime.rs <- Ollama/GPU status platform/ <- OS abstraction (clipboard, GPU, Python paths) scripts/ surya_ocr.py <- Isolated OCR script (bundled as Tauri resource) src/ <- React frontend utils.ts <- Shared utilities (basename, errorMessage) __tests__/fixtures.ts <- Shared test mock objects _classification/ <- Python PoC of the classification pipeline ``` -------------------------------- ### Run quality filter benchmarks Source: https://github.com/akitaonrails/franksherlock/blob/master/_face_ab_test/README.md Command to test confidence, face size, and blur thresholds for noise reduction. ```bash python scripts/phase4_quality_filter.py ``` -------------------------------- ### Run Classification Script Source: https://github.com/akitaonrails/franksherlock/blob/master/_classification/README.md Executes the classification process using the default configuration. ```bash uv run python classification/run_classification.py ``` -------------------------------- ### Application Data Directory Structure Source: https://github.com/akitaonrails/franksherlock/blob/master/README.md Displays the file system layout for application data, including database, cache, and virtual environment directories. ```text db/index.sqlite <- SQLite database with FTS5 cache/thumbnails/ <- Generated thumbnails (mirrored path structure) cache/classifications/ <- Classification cache cache/tmp/ <- Temporary files (GIF frames, etc.) surya_venv/ <- Isolated Python venv for Surya OCR ``` -------------------------------- ### Run embedding benchmarks Source: https://github.com/akitaonrails/franksherlock/blob/master/_face_ab_test/README.md Commands to execute the embedding phase, comparing different models while using a fixed detector. ```bash python scripts/phase2_embedding.py python scripts/phase2_embedding.py --models arcface_insightface facenet ``` -------------------------------- ### Run Rust Unit Tests Source: https://github.com/akitaonrails/franksherlock/blob/master/sherlock/README.md Executes the test suite for core logic including path resolution, query parsing, and SQLite operations. ```bash npm run test:rust ``` -------------------------------- ### Execute video classification script Source: https://github.com/akitaonrails/franksherlock/blob/master/_research_ab_test/docs/RESULTS.md Runs the video classification pipeline with specified models and repeat trials. ```bash uv run python -u scripts/phase4b_video_classify.py --models qwen2.5vl:7b,qwen3-vl:8b,qwen3-vl:30b-a3b --repeats 3 ``` -------------------------------- ### Run clustering benchmarks Source: https://github.com/akitaonrails/franksherlock/blob/master/_face_ab_test/README.md Command to execute the clustering phase on ArcFace embeddings. ```bash python scripts/phase3_clustering.py ``` -------------------------------- ### Verify PyTorch CUDA availability Source: https://github.com/akitaonrails/franksherlock/blob/master/_research_ab_test/docs/RESULTS.md Checks if CUDA is available for hardware acceleration during Whisper execution. ```bash uv run python -c "import torch; print(torch.cuda.is_available())" ``` -------------------------------- ### Run Classification Smoke Test Source: https://github.com/akitaonrails/franksherlock/blob/master/_classification/README.md Executes the classification script with a limited number of images for testing purposes. ```bash uv run python classification/run_classification.py --max-images 5 ``` -------------------------------- ### Research Directory Structure Source: https://github.com/akitaonrails/franksherlock/blob/master/README.md Layout of the historical research and A/B testing directories. ```text _research_ab_test/ scripts/ <- A/B benchmark scripts docs/ <- Research notes (IDEA.md, RESULTS.md, etc.) lib/ <- Shared Python helpers results/ <- Generated benchmark outputs (gitignored) test_files/ <- Test corpus (gitignored, see note below) ``` -------------------------------- ### Override Ollama Vision Model Source: https://github.com/akitaonrails/franksherlock/blob/master/README.md Specify a custom vision model by editing the settings.toml file. The application must be restarted for changes to take effect. ```toml model_override = "qwen2.5vl:3b" ``` === COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.