Added universal date and time for training

Added inmemory streaming

done some more changes

done some major changes

modified aws pools

made sure to make trainig fast

Made more efficient memory management

added single file fetching for visualization and evaluation

Added multi satellite based fetching for training

starting 2 months data training

Resolved major errors

Added some changes

modified for 1 month training

Added more multi dynamic data preprocessing

Added dynamic data pipeline

Added some dynamic dataset for image interpolation

completed main logics

completed main logics

feat: implement CompositeLoss combining Charbonnier and Laplacian

../src/model/laplacian.py

feat: port RIFE architectures and modify for 1-channel TIR

feat: port RIFE architectures and modify for 1-channel TIR

Resolved error and other missing data

feat: add dataset and transforms for TIR triplets

Added main.py

feat: setup centralized config and logging