- Add explicit pytest requirement to testing-patterns.mdc
- Update user-preferences.mdc to include pytest preference
- Modify development-workflow.mdc to reference pytest usage
- Specify that all tests must use pytest framework
- Document pytest features to use (classes, fixtures, assertions)
- Add comprehensive examples of proper pytest structure
- Include DO NOT USE section for prohibited testing approaches
- Ensure CLI integration expects pytest-compatible test files
- Make it clear that manual testing and other frameworks are not allowed
- Restore complete setup module with system information and developer profiles
- Add NBDev educational patterns with #| export and #| hide directives
- Implement SystemInfo class for platform detection and compatibility checking
- Add DeveloperProfile class with ASCII art customization and personalization
- Include comprehensive educational content with step-by-step learning progression
- Expand test suite to 20 comprehensive tests covering all functionality:
- Function execution and output validation
- Arithmetic operations (basic, negative, floating-point)
- System information collection and compatibility
- Developer profile creation and customization
- ASCII art file loading and fallback behavior
- Error recovery and integration testing
- Update README with complete feature documentation and usage examples
- Convert to executed notebook showing rich output and system information
- Maintain file-based ASCII art system with tinytorch_flame.txt
- All tests passing with professional pytest structure
- Replace manual testing with pytest test classes and assertions
- Add comprehensive test coverage:
- Function execution without errors
- Correct output content validation
- ASCII art file existence and content checks
- Error handling and fallback behavior
- Missing file graceful handling
- Update README with pytest testing instructions
- Make testing consistent with other TinyTorch modules
- All 5 tests passing with proper pytest integration
- Replace complex setup module with simple hello_tinytorch() function
- Keep only the ASCII art file (tinytorch_flame.txt) for visual appeal
- Simplify tests to just verify function runs and file exists
- Update README to reflect simplified purpose and usage
- Remove complex developer profiles and system info classes
- Focus on minimal introduction to TinyTorch workflow
🔧 Notebook Compatibility:
- Fix __file__ NameError in notebook environment
- Add try/except for file path detection
- Handle both script and notebook execution contexts
- Use os.getcwd() fallback when __file__ not available
📓 Notebook Execution:
- Execute setup_dev.ipynb with all outputs generated
- Beautiful ASCII art displays properly in notebook
- All test cells show expected results
- Student-friendly interactive experience
🎨 ASCII Art Display:
- Stunning flame ASCII art renders perfectly in notebook
- Full profile display with Tiny🔥Torch branding
- Interactive testing cells with clear outputs
- Students can see the ASCII art immediately
🧪 Testing:
- All 17 tests still pass after notebook conversion
- File loading works in both environments
- Comprehensive coverage maintained
- Ready for student use
Students can now run the setup_dev.ipynb notebook and immediately see the beautiful ASCII art output, making their first TinyTorch experience memorable and engaging
🎨 File-Based ASCII Art System:
- Add beautiful tinytorch_flame.txt with sophisticated flame design
- Dynamic file loading with fallback to simple flame
- Easy customization - students can edit the file directly
- Proper error handling for missing files
🔥 Enhanced Visual Experience:
- Stunning detailed flame ASCII art from file
- 'Tiny🔥Torch' branding with 'Build ML Systems from Scratch!' tagline
- Much more sophisticated design than hardcoded version
- Students can easily customize by editing tinytorch_flame.txt
📚 Educational Benefits:
- Two customization options: parameter override or file editing
- Teaches file I/O and error handling concepts
- Encourages creative expression and ownership
- Professional development practices
🧪 Testing:
- Updated all tests to work with file-based system
- All 17 tests pass successfully
- Comprehensive coverage of file loading and fallback
- Verified both custom and default ASCII art functionality
Students can now easily create their own ASCII art by editing the tinytorch_flame.txt file or providing custom art via parameters
🎨 ASCII Art Features:
- Add beautiful TinyTorch flame ASCII art as default
- Support custom ASCII art for student personalization
- New get_ascii_art() and get_full_profile() methods
- Enhanced visual appeal and student engagement
👨💻 Profile Updates:
- Update instructor info: vj@eecs.harvard.edu, @profvjreddi
- Add comprehensive ASCII art examples and ideas
- Encourage creative student customization
- Professional signature generation with visual flair
🧪 Testing:
- Add 2 new ASCII art tests (17 total tests pass)
- Test default flame art and custom art functionality
- Verify full profile display with proper formatting
- Comprehensive coverage of all new features
Students can now create truly personalized TinyTorch experiences with custom ASCII art, making their learning journey unique and memorable
- Add DeveloperProfile class with customizable developer information
- Default to course instructor (Vijay Janapa Reddi) but allow student customization
- Include professional signature generation for code attribution
- Add comprehensive test coverage (6 new tests, 15 total tests pass)
- Encourage student ownership and personalization of their TinyTorch journey
- Maintain educational structure with TODO sections and hidden solutions
- Fix syntax errors in setup_dev.py test cells (proper indentation)
- Add comprehensive test suite for setup module (test_setup.py)
- Test coverage: functions, SystemInfo class, integration tests
- Custom test runner (no pytest dependency required)
- All 9 tests pass successfully
- Update .gitignore to allow modules/ directory for educational structure
- Fix CLI tool to look for tests in modules/{module}/tests/ instead of tests/
- Update test imports to use parent directory module imports
- Update Cursor rules to reflect new test structure:
* Project structure shows tests/ subdirectory
* Testing patterns show correct paths and import patterns
* Development workflow shows updated test locations
- Test imports now work: from tensor_dev import Tensor
- CLI commands now find tests in correct locations
Tests are now properly organized and discoverable
- Move test_tensor.py to modules/tensor/tests/test_tensor.py
- Create tests/ directories in setup and tensor modules
- Clean up setup module: remove setup_dev.html, setup_dev_files/, __init__.py
- Establish clean module structure pattern:
* {module}_dev.py - Main development file
* {module}_dev.ipynb - Generated notebook
* README.md - Module documentation
* tests/ - Test directory with test_{module}.py
This makes it crystal clear to students where tests are located.
- Fix frontmatter for all rules to include proper descriptions
- Update rule types according to Cursor guidelines:
* alwaysApply: project-structure, user-preferences
* globs: nbdev-educational-pattern, testing-patterns
* description only: development-workflow, ml-systems-course-context
- All rules under 500 lines, focused and actionable
- Follow MDC format with proper metadata
- Ready for proper Cursor rule functionality
- Project structure guide (always applied)
- NBDev educational pattern for module development files
- Development workflow and CLI usage
- Testing patterns and conventions
- ML Systems course context and learning approach
- User preferences and conventions
These rules will help with codebase navigation, understanding the educational approach,
and maintaining consistency across the TinyTorch framework.
🎯 Student Version (Visible):
- All functions raise NotImplementedError('Student implementation required')
- Clear TODO instructions with specific guidance
- Test cells handle NotImplementedError gracefully
- Students must implement everything from scratch
🔧 Instructor Version (Hidden with #|hide):
- Complete working implementations
- Same function signatures as student version
- Production-ready code for package export
📚 Added MODULE_GENERATION_GUIDE.md:
- Clear instructions for creating student/instructor modules
- NBDev workflow documentation
- Module structure patterns and templates
- Generation commands and checklist
🔄 Generation Flow:
1. Write .py with student stubs + hidden instructor solutions
2. Convert to notebook: jupytext --to notebook module_dev.py
3. Export package: python bin/tito.py sync --module name
4. Package gets instructor solutions, students see exercises
Perfect foundation for all TinyTorch modules
- Removed modules/setup/setup_nbdev_educational.* files
- Keep only the clean setup_dev.* files
- Setup module now focused and clean
- NBDev features work behind scenes without clutter
- Focused on original purpose: just setting up development environment
- Students don't see NBDev educational features explanations
- Simple workflow: hello world, basic class, export, test, progress
- NBDev #|hide directive works behind scenes for instructor solutions
- Clean and simple, just like the original but with hidden solutions
Back to basics: setup is about setup, not teaching NBDev features
SINGLE SOURCE approach: One notebook serves both instructors and students
🎯 Key Features:
- #|hide directive hides complete solutions from students
- #|code-fold creates collapsible sections for details
- #|export with educational metadata
- Progressive learning: simple → ML-relevant complexity
- Vector operations (add, dot product) as ML foundations
- ML-aware SystemInfo class with library checking
- Comprehensive testing for both student/instructor versions
🔄 Workflow:
- Students see exercises with TODOs and hints
- Instructors see complete solutions (hidden by default)
- Package exports get instructor (complete) implementations
- Single notebook maintains both audiences
This establishes the pattern for all TinyTorch modules
- Create tensor_nbdev_educational.py with NBDev directives
- Demonstrate #|hide, #|code-fold, #|filter_stream features
- Convert using standard Jupytext (# %% markers)
- Add comprehensive guides and comparisons
- Show superior approach vs custom generator
Key insight: NBDev already has mature educational capabilities!
No need to build custom tools when industry standards exist.
Introduces a Tensor class that wraps numpy arrays, enabling
fundamental ML operations like addition, subtraction,
multiplication, and division.
Adds utility methods such as reshape, transpose, sum, mean, max,
min, item, and numpy to the Tensor class.
Updates tests to accommodate both scalar and Tensor results
when checking mean values.
Introduces a README with project overview, setup instructions,
and course structure.
Adds a VISION document outlining the project's goals, conventions,
and architecture.
Includes updates to the setup module's README to clarify module to
package mapping.
✅ Setup Module Implementation:
- Created comprehensive setup_dev.ipynb with TinyTorch workflow tutorial
- Added hello_tinytorch(), add_numbers(), and SystemInfo class
- Updated README with clear learning objectives and development workflow
- All 11 tests passing for complete workflow validation
🔧 CLI Enhancements:
- Added --module flag to 'tito sync' for module-specific exports
- Implemented 'tito reset' command with --force option
- Smart auto-generated file detection and cleanup
- Interactive confirmation with safety preservations
📚 Documentation Updates:
- Updated all references to use [module]_dev.ipynb naming convention
- Enhanced test coverage for new functionality
- Clear error handling and user guidance
This establishes the foundation workflow that students will use throughout TinyTorch development.
Extends the sync command to allow users to synchronize
specific modules instead of the entire project. This
improves efficiency by reducing the scope of the nbdev
export process. Adds argument parsing for module selection.
Adds a comprehensive VISION.md document outlining the philosophy, learning approach, module design principles, and target learning outcomes for TinyTorch.
This document serves as a canonical guide for module development, emphasizing a systems-first approach to understanding ML frameworks, and provides a framework for thinking about design trade-offs in accuracy, runtime, and memory footprint.
Introduces the VISION.md document, which outlines the core philosophy, learning approach, module design principles, target learning outcomes, module progression logic, and success metrics for TinyTorch.
This document serves as the "ground truth" for course design decisions, emphasizing a systems-first approach to understanding machine learning frameworks. It details the "Loops to Systems" learning approach, focusing on the systems engineering challenges that arise when implementing ML concepts at scale.
Updates references to the development notebook naming convention from `[module].ipynb` to `[module]_dev.ipynb` in documentation. This change ensures consistency across the project and aligns with the intended naming scheme for development notebooks.
Introduces a standardized module structure with README, notebooks, tutorials, tests, and solutions.
Refactors the project to emphasize a modular learning path, enhancing clarity and consistency across the TinyTorch course.
Changes the virtual environment path to ".venv".
Updates the project to use `.venv` as the standard virtual environment directory. This change:
- Updates `.gitignore` to ignore `.venv/`.
- Modifies the activation script to create and activate `.venv`.
- Adjusts the `tito.py` script to check for `.venv`'s existence and activation.
- Updates documentation and setup scripts to reflect the new virtual environment naming convention.
This change streamlines environment management and aligns with common Python practices.
Adds a comprehensive .gitignore file to exclude common build artifacts, logs, temporary files, and other project-specific files that should not be tracked by Git.
This ensures a cleaner repository and prevents accidental inclusion of sensitive or unnecessary files.
Introduces a command-line interface (CLI) named 'tito'
to streamline project setup, testing, and information retrieval.
Includes a setup script to automate virtual environment
creation and dependency installation.
Improves the user experience by providing clear instructions
and status indicators within the CLI.
Introduces the TinyTorch project, a pedagogical tool for building ML systems from scratch.
Provides an overview of the project's goals, system architecture, curriculum integration, and getting started guide.
Sets up the foundational project structure for the TinyTorch ML system, including the CLI entry point, project directories, and setup scripts.
This commit introduces the `tito` CLI for project management, testing, and information display.
It also includes setup scripts to automate environment creation and verification, along with initial documentation.
Introduces the foundational CLI structure and core components for the TinyTorch project.
This initial commit establishes the command-line interface (CLI) using `argparse` for training, evaluation, benchmarking, and system information. It also lays out the basic directory structure and essential modules, including tensor operations, autograd, neural network layers, optimizers, data loading, and MLOps components.
