NLP -> HF Datasets

[zphang]

Switch from nlp to datasets==1.1.2.

In code, try to refer to HF Datasets / hf_datasets rather than just "datasets" to minimize confusion.

Tested with the following command:

BASE_PATH=/path/to/download
STATUS_PATH=${BASE_PATH}/status
mkdir -p ${STATUS_PATH}
python jiant/scripts/download_data/runscript.py \
    download \
    --benchmark GLUE \
    --output_path ${BASE_PATH}/tasks/ && \
    touch ${STATUS_PATH}/glue_done
python jiant/scripts/download_data/runscript.py \
    download \
    --benchmark SUPERGLUE \
    --output_path ${BASE_PATH}/tasks/ && \
    touch ${STATUS_PATH}/superglue_done
# Doesn't include PAN-X because it requires an external file
python jiant/scripts/download_data/runscript.py \
    download \
    --tasks xnli pawsx udpos xquad mlqa tydiqa bucc2018 tatoeba \
    --output_path ${BASE_PATH}/tasks/ && \
    touch ${STATUS_PATH}/xtreme_done
# Datasets
python jiant/scripts/download_data/runscript.py \
    download \
    --tasks snli commonsenseqa hellaswag cosmosqa socialiqa scitail\
    --output_path ${BASE_PATH}/tasks/ && \
    touch ${STATUS_PATH}/hf_datasets_done
# Files
python jiant/scripts/download_data/runscript.py \
    download \
    --tasks squad_v1 squad_v2 abductive_nli swag qamr qasrl\
    --output_path ${BASE_PATH}/tasks/ && \
    touch ${STATUS_PATH}/dl_datasets_done

[pep8speaks]

Hello @zphang! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

• In the file setup.py:

Line 2:101: E501 line too long (118 > 100 characters)
Line 26:101: E501 line too long (101 > 100 characters)
Line 39:101: E501 line too long (106 > 100 characters)
Line 47:101: E501 line too long (126 > 100 characters)

You can repair most issues by installing black and running: black -l 100 ./*. If you contribute often, have a look at the 'Contributing' section of the README for instructions on doing this automatically.

Comment last updated at 2020-10-19 17:50:30 UTC

[zphang]

(Not a high priority to merge)

[jeswan]

Thanks!