expanded_checklist/checklist/eval_core.py (4 lines):
	- line 58: TODO: subsampling doesn't work for now.
	- line 334: # TODO: this works only if the data wasn't sampled (sampling is not
	- line 405: # TODO: fix the run_idx to match the new filtered data
	- line 556: # TODO: fix. Only running from a file is supported atm.


expanded_checklist/checklist/core_record.py (2 lines):
	- line 179: # TODO: this is NER specific - the code should be restructured
	- line 631: # TODO: there can be situations where the tokenization is context-


src/models/readers/simple_reader.py (1 line):
	- line 53: # NOTE and a TODO: for now this reader assumes there are no


expanded_checklist/checklist/lexicons.py (1 line):
	- line 23: # TODO: this assumes the sets of possible values are different