[Model] Add ABCNet recog decoder and fix docstring in CRNN

[gaotongxiao]

Modification

1. Renamed mmocr.models.textrecog.layers.lstm_layer to mmocr.models.textrecog.layers.rnn_layers
2. Added AttentionGRUCell to mmocr.models.textrecog.layers.rnn_layers
3. Added ABCNet's recognition decoder
4. Fixed docstring in CRNN

BC-breaking

Users who have imported anything from mmocr.models.textrecog.layers.lstm_layer should change the path to mmocr.models.textrecog.layers.rnn_layers or mmocr.models.textrecog.layers (recommended). But since this layer is rarely used, the potential negative impact is neglectible.

[codecov[bot]]

Codecov Report

Merging #747 (3266faa) into main (8248802) will increase coverage by 0.08%. The diff coverage is 95.23%.

@@            Coverage Diff             @@
##             main     #747      +/-   ##
==========================================
+ Coverage   84.59%   84.68%   +0.08%     
==========================================
  Files         164      166       +2     
  Lines       10642    10745     +103     
  Branches     1618     1630      +12     
==========================================
+ Hits         9003     9099      +96     
- Misses       1295     1299       +4     
- Partials      344      347       +3     

Flag	Coverage Δ
unittests	84.68% <95.23%> (+0.08%)	:arrow_up:

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

Impacted Files	Coverage Δ
mmocr/models/textrecog/decoders/crnn_decoder.py	75.00% <ø> (ø)
...ocr/models/textrecog/decoders/abc_recog_decoder.py	94.44% <94.44%> (ø)
mmocr/models/textrecog/layers/rnn_layers.py	95.83% <95.83%> (ø)
mmocr/utils/lmdb_util.py	74.28% <0.00%> (ø)
mmocr/core/deployment/deploy_utils.py	20.54% <0.00%> (ø)
mmocr/datasets/openset_kie_dataset.py	90.96% <0.00%> (ø)
mmocr/models/textrecog/recognizer/satrn.py	100.00% <0.00%> (ø)
mmocr/models/common/detectors/single_stage.py	83.33% <0.00%> (ø)
mmocr/models/textrecog/encoders/satrn_encoder.py	100.00% <0.00%> (ø)
mmocr/models/common/modules/transformer_module.py	98.71% <0.00%> (ø)
... and 5 more

---

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 8248802...3266faa. Read the comment docs.