TODOS.md

TODOS

• Update dataloader to perform grid split of data
• Add learning rate scheduler (apopt from AlphaPeptDeep)
• Update/Test different loss functions
• Optimize inference and data types with torchao
• Add eval metrics to WandB logging
• Move sampling.py into codebase
• Add a raw mzML/tdf parser
• Obtain another dataset for testing
• Benchmark if time

In Progress

Completed

• - Dataloaders for MS1+MS2 data
• Implement base diffusion model from PyTorch
• Adapter layer for MS data to input dimensions of above
• Maybe custom training loop
• Eval code