demonstrating-data-to-knowledge-pipelines

- .gitignore: Added new entry for 'Trajectory_Data' to exclude additional trajectory data files.
- Coscine files: Renamed 'coscine_app_profile.ttl' to 'coscine_app_profile-Trajectory-data.ttl'.
- Dockerfiles:
  - Added locale and timezone settings to ensure consistent environment configuration (set timezone to Europe/Berlin).
  - Improved localization setup across foundation and cron job Dockerfiles.
- Requirements: Updated 'requirements.txt' to require 'coscine' version '>=0.11.5', addressing download limitations in older versions.
- Data Retrieval: Introduced a method to limit the number of stored trajectories per robot to optimize disk usage and ensure efficient file management.
- Training: Enhanced model training script with updated configuration and sweep setups. Added the ability to handle dynamic notes and model configurations based on different benchmarks.
- New Script: Added 'train_instance.py' to manage instance training using ITA data, including a new method to streamline data processing, training, and model evaluation.
- Code Refactoring:
  - Improved existing functions for better readability and maintainability.
  - Updated logging to provide better traceability during operations.
  - Refined sweep configurations to dynamically create and handle hyperparameters more efficiently.