Add comprehensive GPU support for embedding generation while maintaining optimized CPU processing for distance calculations:
- Add device option to TransformerEmbeddingOptions (auto, cpu, webgpu, cuda, gpu)
- Implement smart auto-detection of best available GPU (WebGPU for browsers, CUDA for Node.js)
- Add automatic CPU fallback if GPU initialization fails
- Fix misleading GPU acceleration claims in distance functions and HNSW search
- Update documentation to accurately reflect GPU usage (embeddings only)
- Add comprehensive example demonstrating GPU acceleration usage
- Maintain full backward compatibility with existing code
Performance improvements: 3-5x faster embedding generation when GPU is available, while keeping faster CPU processing for 384-dim vector distance calculations.
BREAKING CHANGE: Complete migration from TensorFlow.js to Transformers.js for embedding generation
This is a major architectural change that replaces TensorFlow.js (USE model) with Transformers.js (all-MiniLM-L6-v2) for significantly improved performance and reduced complexity.
Key Changes:
- Replace TensorFlow.js Universal Sentence Encoder with Transformers.js all-MiniLM-L6-v2
- Reduce model size from 525MB to 87MB (83% reduction)
- Reduce embedding dimensions from 512 to 384 (faster distance calculations)
- Remove TensorFlow.js Float32Array patching (caused ONNX conflicts)
- Implement smart bundled model detection for offline operation
- Add explicit model download script for Docker deployments
- Remove complex environment variables in favor of simple configuration
- Update all distance functions to use optimized pure JavaScript
- Remove TensorFlow-specific utilities and type definitions
Performance Improvements:
- Model loading: 5x faster (87MB vs 525MB)
- Memory usage: 75% reduction (~200-400MB vs ~1.5GB)
- Distance calculations: Faster pure JS vs GPU overhead for small vectors
- Cold start performance: Significantly improved
Files Changed:
- Updated package.json: New dependencies, simplified scripts
- Rewrote src/utils/embedding.ts: Complete Transformers.js implementation
- Updated src/utils/distance.ts: Optimized JavaScript distance functions
- Simplified src/setup.ts: Removed TensorFlow-specific patching
- Simplified src/utils/textEncoding.ts: Only Node.js TextEncoder/Decoder patches
- Deleted src/utils/robustModelLoader.ts: TensorFlow-specific loader
- Deleted src/types/tensorflowTypes.ts: TensorFlow type definitions
- Added scripts/download-models.cjs: Docker-compatible model downloader
- Added comprehensive documentation: README.md, OFFLINE_MODELS.md, analysis docs
Testing:
- All 19 tests passing
- Removed test mocking in favor of real implementation testing
- Updated test environment for Transformers.js compatibility
- Performance tests validate improved efficiency
This migration resolves production issues with Docker egress limitations and provides a more robust, performant foundation for vector operations.
- Add npm overrides to force form-data version 4.0.4 or higher
- Fixes GitHub security advisory GHSA-fjxv-7rqg-78g4
- Vulnerability was in transitive dependency via TensorFlow.js
- No functionality changes, all tests passing
- **Documentation Additions**:
- Created `brainy_architecture_diagram.md` to detail Brainy's architecture using diagrams and structured descriptions:
- Added overviews of the system, core architecture, and augmentation pipeline.
- Defined data models, graph structures, storage architecture, and performance optimizations.
- Explained vector search engine design, HNSW index structure, and usage flow examples.
- Developed `brainy_architecture_visual.md` to complement the architecture with visual aids in Mermaid.js:
- Provided detailed flowcharts, mind maps, and sequence diagrams for system components and data flow.
- **Purpose**:
- Provide in-depth technical insights into Brainy's architecture for developers and stakeholders.
- Enhance understanding of the system's core design principles with easy-to-follow diagrams and examples.
This change implements standard-version for versioning and changelog management:
- Adds standard-version as a dev dependency
- Updates package.json scripts to use standard-version
- Creates .versionrc.json configuration file
- Updates CHANGELOG.md with instructions for conventional commits
- Removes obsolete manual changelog update script
Remove `@modelcontextprotocol/server-sequential-thinking` and its nested dependencies from `devDependencies` in `package-lock.json` as they are no longer required for the project. This reduces the overall package size and improves maintenance.
- **Removed Files**:
- Deleted outdated statistics documentation files (`statistics.md`, `statistics-flush-solution.md`, `statistics-summary.md`) to clean up the repository and avoid confusion.
- **Added Standards**:
- Introduced `DOCUMENTATION_STANDARDS.md` to outline naming conventions and troubleshooting practices for more consistent and maintainable project documentation.
- **Tests**:
- Added a new test file `edge-cases.test.ts` to verify handling of edge cases, ensuring robust behavior against boundary values and invalid inputs.
**Purpose**: Cleans up deprecated documentation while introducing concrete standards for maintaining and updating documentation. Enhances test coverage for unusual or boundary inputs, improving overall system resilience.
- **Integration Tests**:
- Introduced `api-integration.test.ts` to validate API functionality:
- Verifies text insertion, vector embedding generation, and search operations.
- Confirms HNSW index correctness for vector similarity search.
- Ensures no dimensional mismatches in embeddings.
- **Test Server**:
- Added test server utilizing Express for endpoint simulation (`/insert` and `/search/text`).
- **Dependencies**:
- Introduced `express` and `node-fetch` as new dependencies for testing purposes.
- **Vitest Fix**:
- Updated `vitest.config.ts` to resolve the `process.memoryUsage` error by setting `logHeapUsage: false`.
- **Package Updates**:
- Modified `package-lock.json` to include newly added dependencies and updates.
**Purpose**: Guarantees the stability of core API endpoints and vector-related functionality, ensuring reliable behavior for end-to-end scenarios.
