add fleurs conformer+sc-ctc results

[wanchichen]

Add Conformer + self-conditioned ctc results for FLEURS
Trained model: https://huggingface.co/espnet/wanchichen_fleurs_asr_conformer_scctc
Note: the trained checkpoint uses the older version of s3prl w/ the fairseq dependency

The models that condition on LID use the newer version of s3prl. Will add in a separate PR with the code changes necessary to run them.

[codecov]

Codecov Report

Merging #4746 (3719a13) into master (af70425) will decrease coverage by 0.11%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #4746      +/-   ##
==========================================
- Coverage   80.44%   80.32%   -0.12%     
==========================================
  Files         527      527              
  Lines       46230    46311      +81     
==========================================
+ Hits        37188    37200      +12     
- Misses       9042     9111      +69     

Flag	Coverage Δ
test_integration_espnet1	66.37% <ø> (ø)
test_integration_espnet2	48.96% <ø> (-0.08%)	⬇️
test_python	68.56% <ø> (-0.10%)	⬇️
test_utils	23.30% <ø> (ø)

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

Impacted Files	Coverage Δ
espnet2/fileio/xml_scp.py	24.10% <0.00%> (-6.25%)	⬇️
espnet2/text/phoneme_tokenizer.py	75.93% <0.00%> (-4.80%)	⬇️
espnet2/tasks/svs.py	0.00% <0.00%> (ø)
espnet2/svs/espnet_model.py	0.00% <0.00%> (ø)
espnet2/train/preprocessor.py	29.27% <0.00%> (+0.57%)	⬆️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[sw005320]

You can replace it as a default config.

[sw005320]

Thanks a lot!
Let's use this pre-trained models as our base!!!