Rewrite confusing model training instructions

These instructions in USAGE.md confused me
> Create a single cell manifest (e.g. csv, parquet) for each dataset with a column corresponding to final processed paths, and create a split column corresponding to train/test/validation split.
Update the final single cell dataset path (SINGLE_CELL_DATASET_PATH) and the column in the manifest for appropriate input modality (SDF_COLUMN/SEG_COLUMN/POINTCLOUD_COLUMN/IMAGE_COLUMN) in each [datamodule file](https://github.com/AllenCell/benchmarking_representations/blob/main/configs/data). e.g. for PCNA data these yaml files are located here -

We should rewrite. (Discussed specifics in person)