Pull from dev to Main #1
Merged
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
- Remove old Docker README and benchmark files - Complete README.md rewrite with comprehensive documentation - Add full API documentation, performance benchmarks, and examples - Implement new FastAPI server with async inference endpoints - Add production-ready features: monitoring, health checks, batch processing - Include optimization guides for TensorRT, ONNX, and quantization - Add complete installation guide with uv package manager - Document 2-10x performance improvements and memory optimizations - Provide Docker deployment configuration and development setup - Add comprehensive troubleshooting and configuration sections This represents a complete transformation from a basic inference service to a production-ready framework with extensive documentation and examples.
…, and configuration management - Update README to structured docs approach with focused overview - Remove consolidated DOCUMENTATION_CONSOLIDATION.md file - Add comprehensive development dependencies (pytest plugins, security tools) - Implement enhanced configuration management in main.py - Add monitoring improvements with request tracking and performance stats - Update pyproject.toml with extensive pytest configuration and coverage settings - Add security tools (bandit, safety) and testing infrastructure (tox, pytest-xdist) - Enhanced requirements with development, testing, and security packages - Update uv lockfile with new comprehensive dependency tree - Configure device type enum with string conversion support
- Framework: Fix model_manager function reference bug and add performance tracking - Config: Add validation, property accessors, and safer torch_compile defaults - Error handling: Add graceful degradation for compilation and dependency failures - Optimizers: Add standardized optimize() interfaces and better error recovery - Processors: Add CustomPreprocessor/Postprocessor for unknown input/output types - Testing: Improve test stability, add proper mocking, and enhance error capture - Dependencies: Add optional import handling for enterprise features - Batch processing: Enhanced batch prediction with fallback mechanisms - Memory: Add gradient checkpointing and CUDA memory optimizations - Monitoring: Expand metrics collection with flexible input handling - Compatibility: Add backward compatibility layers and dict conversion methods This comprehensive update improves framework robustness, adds missing features, and enhances error handling throughout the codebase while maintaining backward compatibility.
…ty issues - Fix model_manager initialization to store instance instead of function reference - Add backward compatibility property for model_manager access - Separate async and sync cleanup methods for better lifecycle management - Fix batch tensor stacking for inputs with existing batch dimensions - Enhance device configuration with better validation and error handling - Add fallback support for missing optimizer dependencies (TensorRT, ONNX) - Improve timeout handling and logging in inference engine - Strengthen config manager with flexible device type parsing - Update test suite with better mocking and error handling - Remove outdated batch scripts and empty test files - Update requirements.txt with additional development dependencies Breaking changes: - model_manager is now an instance property instead of function call - cleanup() method now defaults to synchronous, use cleanup_async() for async - DeviceType.from_string() now raises ValueError for invalid device types
- Created more robust CI/CD pipeline with proper job dependencies - Added validation job to ensure workflow runs correctly - Fixed docs job with MkDocs configuration file - Added basic workflow as fallback - Improved error handling and continue-on-error flags - Added manual workflow dispatch trigger - Fixed security scans and package building steps
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
No description provided.