From b504ac314fbc4bc17fc18a92c128cd797a71d9cd Mon Sep 17 00:00:00 2001 From: jinzr <60612200+JinZr@users.noreply.github.com> Date: Tue, 24 Oct 2023 11:35:53 +0800 Subject: [PATCH] minor fixes --- .../run-aishell-zipformer-2023-10-24.sh | 45 ++++++------------- egs/aishell/ASR/RESULTS.md | 2 +- 2 files changed, 15 insertions(+), 32 deletions(-) diff --git a/.github/scripts/run-aishell-zipformer-2023-10-24.sh b/.github/scripts/run-aishell-zipformer-2023-10-24.sh index 66b7965f7..865e29799 100755 --- a/.github/scripts/run-aishell-zipformer-2023-10-24.sh +++ b/.github/scripts/run-aishell-zipformer-2023-10-24.sh @@ -29,23 +29,18 @@ log "Display test files" tree $repo/ ls -lh $repo/test_wavs/*.wav -./zipformer/pretrained.py \ - --method greedy_search \ - --max-sym-per-frame $sym \ - --checkpoint $repo/exp/pretrained.pt \ - --tokens $repo/data/lang_char/tokens.txt \ - $repo/test_wavs/BAC009S0764W0121.wav \ - $repo/test_wavs/BAC009S0764W0122.wav \ - $repo/test_wavs/BAC009S0764W0123.wav - -for method in modified_beam_search beam_search fast_beam_search; do +for method in modified_beam_search greedy_search fast_beam_search; do log "$method" ./zipformer/pretrained.py \ --method $method \ - --beam-size 4 \ + --context-size 1 \ --checkpoint $repo/exp/pretrained.pt \ --tokens $repo/data/lang_char/tokens.txt \ + --num-encoder-layers 2,2,4,5,4,2 \ + --feedforward-dim 512,768,1536,2048,1536,768 \ + --encoder-dim 192,256,512,768,512,256 \ + --encoder-unmasked-dim 192,192,256,320,256,192 \ $repo/test_wavs/BAC009S0764W0121.wav \ $repo/test_wavs/BAC009S0764W0122.wav \ $repo/test_wavs/BAC009S0764W0123.wav @@ -62,21 +57,13 @@ log "Display test files" tree $repo/ ls -lh $repo/test_wavs/*.wav -./zipformer/pretrained.py \ - --method greedy_search \ - --max-sym-per-frame $sym \ - --checkpoint $repo/exp/pretrained.pt \ - --tokens $repo/data/lang_char/tokens.txt \ - $repo/test_wavs/BAC009S0764W0121.wav \ - $repo/test_wavs/BAC009S0764W0122.wav \ - $repo/test_wavs/BAC009S0764W0123.wav -for method in modified_beam_search beam_search fast_beam_search; do +for method in modified_beam_search greedy_search fast_beam_search; do log "$method" ./zipformer/pretrained.py \ --method $method \ - --beam-size 4 \ + --context-size 1 \ --checkpoint $repo/exp/pretrained.pt \ --tokens $repo/data/lang_char/tokens.txt \ $repo/test_wavs/BAC009S0764W0121.wav \ @@ -96,23 +83,19 @@ log "Display test files" tree $repo/ ls -lh $repo/test_wavs/*.wav -./zipformer/pretrained.py \ - --method greedy_search \ - --max-sym-per-frame $sym \ - --checkpoint $repo/exp/pretrained.pt \ - --tokens $repo/data/lang_char/tokens.txt \ - $repo/test_wavs/BAC009S0764W0121.wav \ - $repo/test_wavs/BAC009S0764W0122.wav \ - $repo/test_wavs/BAC009S0764W0123.wav -for method in modified_beam_search beam_search fast_beam_search; do +for method in modified_beam_search greedy_search fast_beam_search; do log "$method" ./zipformer/pretrained.py \ --method $method \ - --beam-size 4 \ + --context-size 1 \ --checkpoint $repo/exp/pretrained.pt \ --tokens $repo/data/lang_char/tokens.txt \ + --num-encoder-layers 2,2,2,2,2,2 \ + --feedforward-dim 512,768,768,768,768,768 \ + --encoder-dim 192,256,256,256,256,256 \ + --encoder-unmasked-dim 192,192,192,192,192,192 \ $repo/test_wavs/BAC009S0764W0121.wav \ $repo/test_wavs/BAC009S0764W0122.wav \ $repo/test_wavs/BAC009S0764W0123.wav diff --git a/egs/aishell/ASR/RESULTS.md b/egs/aishell/ASR/RESULTS.md index 002db08fe..0b22f41a1 100644 --- a/egs/aishell/ASR/RESULTS.md +++ b/egs/aishell/ASR/RESULTS.md @@ -143,7 +143,7 @@ for m in greedy_search modified_beam_search fast_beam_search ; do ./zipformer/decode.py \ --epoch 56 \ --avg 23 \ - --exp-dir ./zipformer/exp-small \ + --exp-dir ./zipformer/exp-large \ --lang-dir data/lang_char \ --context-size 1 \ --decoding-method $m \