[BUGFIX] gc before fork in loader

[szha]

Description

Same workaround from https://github.com/apache/incubator-mxnet/pull/19925 for dataset loader in gluonnlp.

Checklist

Essentials

• [x] PR's title starts with a category (e.g. [BUGFIX], [MODEL], [TUTORIAL], [FEATURE], [DOC], etc)
• [x] Changes are complete (i.e. I finished coding on this PR)
• [x] All changes have test coverage
• [x] Code is well-documented

Changes

• [x] avoid dangling data before fork by waiting and gc

cc @dmlc/gluon-nlp-team

[github-actions[bot]]

The documentation website for preview: http://gluon-nlp-staging.s3-accelerate.dualstack.amazonaws.com/PR1525/175272ccbe14c826e7a1e96e40ad572ca650587a/index.html

[codecov[bot]]

Codecov Report

Merging #1525 (175272c) into master (0211c65) will increase coverage by 0.26%. The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master    #1525      +/-   ##
==========================================
+ Coverage   86.27%   86.54%   +0.26%     
==========================================
  Files          55       55              
  Lines        7513     7522       +9     
==========================================
+ Hits         6482     6510      +28     
+ Misses       1031     1012      -19     

Impacted Files	Coverage Δ
src/gluonnlp/data/loading.py	81.75% <100.00%> (+5.90%)	:arrow_up:
src/gluonnlp/data/filtering.py	82.60% <0.00%> (+4.34%)	:arrow_up:

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 0211c65...175272c. Read the comment docs.