mirror of
https://github.com/MLSysBook/TinyTorch.git
synced 2026-05-06 06:48:26 -05:00
722679f165
- Fixed axis=1 error when CrossEntropyLoss receives 1D prediction arrays - Added robust handling for both 1D and 2D prediction inputs - Reshapes 1D arrays to 2D for consistent processing - All integration tests now pass (17/17) - All inline tests pass (6/6) - tito CLI integration working correctly Technical improvements: - Handles single sample predictions correctly - Maintains backward compatibility with batch inputs - Prevents numpy axis errors in edge cases - Ensures consistent shape handling across all loss functions