Skip to content
This repository was archived by the owner on Sep 28, 2025. It is now read-only.

v0.5.0 - Enhanced Memory Profiling & Visualization Improvements

Choose a tag to compare

@Goldziher Goldziher released this 05 Jul 09:12
· 153 commits to main since this release
2f9961e

πŸš€ Enhanced Memory Profiling & Visualization Improvements

πŸ”§ Memory Profiling Fixes

  • Fixed critical memory profiling issues - All frameworks now properly report memory and CPU metrics
  • Enhanced sync/async profilers - Improved sample collection and data sharing mechanisms
  • Better resource monitoring - Added fallback mechanisms for edge cases and fast operations
  • Accurate performance metrics - No more zero values for memory and CPU usage

πŸ“Š Visualization Enhancements

  • Separated memory and throughput charts - Clear distinction between memory usage and data processing speed
  • Installation size analysis - New 3-panel comparison showing size, dependencies, and storage efficiency
  • Improved chart organization - Better labeling and categorization on GitHub Pages
  • Consistent color scheme - Framework colors maintained across all visualizations

✨ CLI & Workflow Improvements

  • Integrated installation size checking - Automatically collected during benchmark runs
  • Enhanced error reporting - Better stack traces and debugging information
  • Fixed import issues - Resolved Path-related CLI problems
  • Streamlined benchmark workflow - Installation sizes included in standard benchmark output

🎯 Key Features

  • Memory usage analysis with CPU metrics visualization
  • Data throughput comparison by framework and category
  • Installation size comparison with dependency analysis
  • Proper profiling for both sync and async extractors
  • GitHub Pages with correctly labeled visualizations

πŸ› Bug Fixes

  • Fixed memory profiling returning 0.0 values for all frameworks
  • Corrected chart labeling confusion between memory and throughput
  • Resolved CLI import errors in benchmark command
  • Fixed visualization file naming and organization

This release significantly improves the accuracy and clarity of benchmark results, providing proper memory profiling and better organized visualizations for comprehensive framework analysis.

πŸ€– Generated with Claude Code