hotpotqa / hotpot

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

MemoryError: during preprocessing

mominali12 opened this issue · comments

commented

python main.py --mode prepro --data_file hotpot_train_v1.1.json --para_limit 2250 --data_split train

Getting memory error after 11048 tasks are done.
I have 8GB of RAM and I believe that it runs out of memory while doing pre processing. Is there any way so that I could perhaps perform pre processing in chunks or some other way around ?

Try reducing n_jobs in Parallel calls. More parallel jobs speed up processing, but also consume more memory.