Vijay Janapa Reddi
c7dbf68dcf
Fix training pipeline: Parameter class, Variable.sum(), gradient handling
...
Major fixes for complete training pipeline functionality:
Core Components Fixed:
- Parameter class: Now wraps Variables with requires_grad=True for proper gradient tracking
- Variable.sum(): Essential for scalar loss computation from multi-element tensors
- Gradient handling: Fixed memoryview issues in autograd and activations
- Tensor indexing: Added __getitem__ support for weight inspection
Training Results:
- XOR learning: 100% accuracy (4/4) - network successfully learns XOR function
- Linear regression: Weight=1.991 (target=2.0), Bias=0.980 (target=1.0)
- Integration tests: 21/22 passing (95.5% success rate)
- Module tests: All individual modules passing
- General functionality: 4/5 tests passing with core training working
Technical Details:
- Fixed gradient data access patterns throughout activations.py
- Added safe memoryview handling in Variable.backward()
- Implemented proper Parameter-Variable delegation
- Added Tensor subscripting for debugging access(https://claude.ai/code )
2025-09-28 19:14:11 -04:00
Vijay Janapa Reddi
c3fa592a5e
Prepare for v0.1 release
...
Documentation:
• Add comprehensive student quickstart guide
• Create instructor guide with grading workflow
• Update README with v0.1 features and capabilities
• Document interactive ML Systems questions
• Add tito grade command documentation
Cleanup:
• Remove __pycache__ directories (1073 removed)
• Clean .ipynb_checkpoints
• Remove experimental Python files
• Clean up temporary files (.pyc, .DS_Store)
Features in v0.1:
• 17 educational modules from tensors to transformers
• Interactive ML Systems thinking questions (NBGrader)
• TinyGPT demonstrating 70% framework reuse
• 16-checkpoint capability progression system
• Simplified tito CLI wrapping all functionality
• Complete instructor grading workflow
Ready for v0.1 release tag.
2025-09-17 19:29:16 -04:00
Vijay Janapa Reddi