Commit History
-
Implement service layers for backup, model management, file handling, tokenization, and training
muhammadjalaludin- Added BackupService for managing cache backups and uploads to Hugging Face Hub. - Introduced ModelManager for loading and managing ML models and tokenizers. - Created SpaceFileService for handling file operations in the local model directory. - Developed TokenizationService for text tokenization and vocabulary analysis. - Implemented TrainingService for training ML models with data augmentation and reproducibility features. - Added comprehensive unit tests for all new services to ensure functionality and error handling.1424118
-
Refactor and enhance API services for predictions and training
thisisjackii- Improved predict.py by adding structured response formatting. - Refactored train.py to modularize file validation and data extraction. - Updated transaction_dataset.py constructor for better readability. - Enhanced main.py health check response. - Refactored model_service.py for clearer logging and error handling. - Improved prediction_service.py logging for better traceability. - Refactored training_service.py for enhanced logging and modularity. - Updated node_service.py to include timeout for API requests. - Refactored node_util.py to use async HTTP client and improved error handling.54c349d
-
Refactor model and prediction services; enhance logging and error handling; add log upload functionality
thisisjackii4494f2e
-
Pydantic for Predict API
thisisjackiid826e4a -
Refactor prediction logic to handle errors and return confidence scores; update file paths to use Path objects for better compatibility.
thisisjackii93a3d1c -
Adding LR as hybrid approach
thisisjackii93cd14c
-
Refactoring
Muhammmad Jalaludin Qurthubif8fd395 -
Refactoring
Muhammmad Jalaludin Qurthubid8afb70
-
Quickstart
Muhammmad Jalaludin Qurthubi1c7af52