cc_net/jsonql.py (2 lines):
	- line 1014: # TODO: should we use another format ?
	- line 1121: # TODO: open the remote file in streaming mode.


cc_net/mine.py (2 lines):
	- line 300: # TODO: try to reduce this / make it a function of "hash_in_mem" / num_langs
	- line 415: # TODO: better default


cc_net/tokenizer.py (1 line):
	- line 56: # TODO find a tokenizer for those languages


cc_net/tools/expand_corpus.py (1 line):
	- line 48: # TODO use classic files directory.


cc_net/perplexity.py (1 line):
	- line 155: # TODO: try copying models file, try READ or PARALLEL_READ


cc_net/process_wet_file.py (1 line):
	- line 198: # TODO: start downloading the next segment in the background