GPU CI Improvements

[ethanwharris]

What does this PR do?

Makes the following changes:

• Fixes default weights to freeze
• Fixes Bleu metric usage
• Splits GPU CI into multiple jobs (one per data type) to run on each PR

Before submitting

• Was this discussed/approved via a Github issue? (no need for typos and docs improvements)
• Did you read the contributor guideline, Pull Request section?
• Did you make sure your PR does only one thing, instead of bundling different changes together?
• Did you make sure to update the documentation with your changes?
• Did you write any new necessary tests? [not needed for typos/docs]
• Did you verify new and existing tests pass locally with your changes?
• If you made a notable change (that affects users), did you update the CHANGELOG?

PR review

• Is this pull request ready for review? (if not, please submit in draft mode)

Anyone in the community is free to review the PR once the tests have passed.
If we didn't discuss your PR in Github issues there's a high chance it will not be merged.

Did you have fun?

Make sure you had fun coding 🙃

[codecov]

Codecov Report

Merging #992 (096f7fd) into master (1489f47) will decrease coverage by 0.03%.
The diff coverage is 15.38%.

@@            Coverage Diff             @@
##           master     #992      +/-   ##
==========================================
- Coverage   88.45%   88.41%   -0.04%     
==========================================
  Files         250      250              
  Lines       13161    13168       +7     
==========================================
+ Hits        11642    11643       +1     
- Misses       1519     1525       +6     

Flag	Coverage Δ
unittests	88.41% <15.38%> (-0.04%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
flash/core/finetuning.py	92.40% <ø> (ø)
flash/core/trainer.py	85.93% <0.00%> (-2.07%)	⬇️
flash/image/classification/model.py	75.00% <0.00%> (ø)
flash/text/classification/data.py	96.89% <ø> (ø)
flash/text/classification/model.py	91.83% <ø> (ø)
flash/text/seq2seq/summarization/model.py	80.00% <0.00%> (ø)
flash/text/seq2seq/translation/model.py	68.18% <0.00%> (-6.82%)	⬇️
flash/core/model.py	87.95% <100.00%> (+0.04%)	⬆️
flash/core/hooks.py	80.00% <0.00%> (-20.00%)	⬇️

---

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 1489f47...096f7fd. Read the comment docs.

[tchaton]

LGTM !

[tchaton]

neat !