Add dataset slicing and fix remove annotations

[potipot]

add dataset slicing

This simple PR adds functionality of slicing dataset to return a new sub dataset. This can be used in inference for example like so:

infer_dl = model_type.infer_dl(valid_ds[:4], batch_size=4, shuffle=False)
preds = model_type.predict_from_dl(model, infer_dl, keep_images=True)

remove annotation

The remove annotation for detection component was broken, it only deleted the label_id from record but not the string encoded label itself.

[ai-fast-track]

Great idea Pawel! Thanks!

[potipot]

it seems like the error is coming from pytorch-lightning 1.7.0
the version we install

Requirement already satisfied: pytorch-lightning>=1.4.5 in /opt/hostedtoolcache/Python/3.7.13/x64/lib/python3.7/site-packages (from icevision==0.12.0) (1.7.1)

argument deprecated since 1.7.0:
https://pytorch-lightning.readthedocs.io/en/stable/generated/CHANGELOG.html#removed

[codecov]

Codecov Report

❌ Patch coverage is 87.50000% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 85.83%. Comparing base (356d29b) to head (500ddf2).
⚠️ Report is 10 commits behind head on master.

Files with missing lines	Patch %	Lines
icevision/data/dataset.py	85.71%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1131      +/-   ##
==========================================
+ Coverage   85.48%   85.83%   +0.35%     
==========================================
  Files         305      305              
  Lines        6675     6713      +38     
==========================================
+ Hits         5706     5762      +56     
+ Misses        969      951      -18     

Flag	Coverage Δ
unittests	85.83% <87.50%> (+0.35%)	⬆️

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

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.