From f79c5e15bcaa5e381bec2d927563e0f0d650b7b5 Mon Sep 17 00:00:00 2001 From: zr_jin Date: Tue, 20 Feb 2024 14:59:00 +0800 Subject: [PATCH] Update RESULTS.md --- egs/spgispeech/ASR/RESULTS.md | 71 +++++++++++++++++------------------ 1 file changed, 35 insertions(+), 36 deletions(-) diff --git a/egs/spgispeech/ASR/RESULTS.md b/egs/spgispeech/ASR/RESULTS.md index 30a354e30..4edaccb4d 100644 --- a/egs/spgispeech/ASR/RESULTS.md +++ b/egs/spgispeech/ASR/RESULTS.md @@ -14,7 +14,7 @@ The WERs are: |---------------------------|------------|------------|------------------------------------------| | greedy search | 2.08 | 2.14 | --epoch 30 --avg 10 | | modified beam search | 2.05 | 2.09 | --epoch 30 --avg 10 --beam-size 4 | -n| fast beam search | 2.07 | 2.17 | --epoch 30 --avg 10 --beam 20 --max-contexts 8 --max-states 64 | +| fast beam search | 2.07 | 2.17 | --epoch 30 --avg 10 --beam 20 --max-contexts 8 --max-states 64 | **NOTE:** SPGISpeech transcripts can be prepared in `ortho` or `norm` ways, which refer to whether the transcripts are orthographic or normalized. These WERs correspond to the normalized transcription @@ -39,31 +39,30 @@ The decoding command is: ``` # greedy search python ./zipformer/decode.py \ - --epoch $epoch \ - --avg $avg \ - --exp-dir ./zipformer/exp \ - --max-duration 1000 \ - --decoding-method modified_beam_search - --decoding-method greedy_search + --epoch $epoch \ + --avg $avg \ + --exp-dir ./zipformer/exp \ + --max-duration 1000 \ + --decoding-method greedy_search # modified beam search python ./zipformer/decode.py \ - --epoch $epoch \ - --avg $avg \ - --exp-dir ./zipformer/exp \ - --max-duration 1000 \ - --decoding-method modified_beam_search + --epoch $epoch \ + --avg $avg \ + --exp-dir ./zipformer/exp \ + --max-duration 1000 \ + --decoding-method modified_beam_search # fast beam search python ./zipformer/decode.py \ - --epoch $epoch \ - --avg $avg \ - --exp-dir ./zipformer/exp \ - --max-duration 1000 \ - --decoding-method fast_beam_search - --beam 4 \ - --max-contexts 4 \ - --max-states 8 + --epoch $epoch \ + --avg $avg \ + --exp-dir ./zipformer/exp \ + --max-duration 1000 \ + --decoding-method fast_beam_search + --beam 4 \ + --max-contexts 4 \ + --max-states 8 ``` ### SPGISpeech BPE training results (Pruned Transducer) @@ -109,28 +108,28 @@ The decoding command is: ``` # greedy search ./pruned_transducer_stateless2/decode.py \ - --iter 696000 --avg 10 \ - --exp-dir ./pruned_transducer_stateless2/exp \ - --max-duration 100 \ - --decoding-method greedy_search + --iter 696000 --avg 10 \ + --exp-dir ./pruned_transducer_stateless2/exp \ + --max-duration 100 \ + --decoding-method greedy_search # modified beam search ./pruned_transducer_stateless2/decode.py \ - --iter 696000 --avg 10 \ - --exp-dir ./pruned_transducer_stateless2/exp \ - --max-duration 100 \ - --decoding-method modified_beam_search \ - --beam-size 4 + --iter 696000 --avg 10 \ + --exp-dir ./pruned_transducer_stateless2/exp \ + --max-duration 100 \ + --decoding-method modified_beam_search \ + --beam-size 4 # fast beam search ./pruned_transducer_stateless2/decode.py \ - --iter 696000 --avg 10 \ - --exp-dir ./pruned_transducer_stateless2/exp \ - --max-duration 1500 \ - --decoding-method fast_beam_search \ - --beam 4 \ - --max-contexts 4 \ - --max-states 8 + --iter 696000 --avg 10 \ + --exp-dir ./pruned_transducer_stateless2/exp \ + --max-duration 1500 \ + --decoding-method fast_beam_search \ + --beam 4 \ + --max-contexts 4 \ + --max-states 8 ``` Pretrained model is available at