search

hirofumi0810 / neural_sp

End-to-end ASR/LM implementation with PyTorch
Apache License 2.0
593 stars 140 forks source link

Refactor LM loading during evaluation #336

Closed hirofumi0810 closed 3 years ago

hirofumi0810 commented 3 years ago

Modularized LM loading before ASR evaluation

codecov-commenter commented 3 years ago

Codecov Report

Merging #336 (4c0fac3) into master (03baea2) will increase coverage by 0.21%. The diff coverage is 88.88%.

Impacted file tree graph

@@            Coverage Diff             @@
##           master     #336      +/-   ##
==========================================
+ Coverage   68.00%   68.22%   +0.21%     
==========================================
  Files          96       96              
  Lines       10803    10786      -17     
==========================================
+ Hits         7347     7359      +12     
+ Misses       3456     3427      -29
Impacted Files Coverage Δ
neural_sp/bin/asr/plot_attention.py 77.22% <50.00%> (+3.15%) :arrow_up:
neural_sp/bin/eval_utils.py 89.28% <93.33%> (+2.92%) :arrow_up:
neural_sp/bin/asr/ctc_forced_align.py 92.85% <100.00%> (ø)
neural_sp/bin/asr/eval.py 65.21% <100.00%> (+5.21%) :arrow_up:
neural_sp/bin/asr/plot_ctc.py 92.45% <100.00%> (ø)
neural_sp/models/seq2seq/frontends/streaming.py 57.62% <0.00%> (-3.39%) :arrow_down:
neural_sp/models/seq2seq/speech2text.py 71.39% <0.00%> (-0.67%) :arrow_down:
neural_sp/models/seq2seq/decoders/las.py 54.34% <0.00%> (+0.22%) :arrow_up:
neural_sp/models/seq2seq/decoders/beam_search.py 80.35% <0.00%> (+11.60%) :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 03baea2...4c0fac3. Read the comment docs.