zbrunner commited on
Commit
3e63a73
·
verified ·
1 Parent(s): 7b6c4e1

Upload ./exp/asr_train_raw_en_bpe10000_sp/RESULTS.md with huggingface_hub

Browse files
exp/asr_train_raw_en_bpe10000_sp/RESULTS.md ADDED
@@ -0,0 +1,152 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ <!-- Generated by scripts/utils/show_asr_result.sh -->
2
+ # RESULTS
3
+ ## Environments
4
+ - date: `Tue Aug 20 21:46:00 BST 2024`
5
+ - python version: `3.10.8 (main, Nov 24 2022, 14:13:03) [GCC 11.2.0]`
6
+ - espnet version: `espnet 202402`
7
+ - pytorch version: `pytorch 1.13.1+cu116`
8
+ - Git hash: `06693f5abd8cc8c8a34d92cf655b81be8ca27db0`
9
+ - Commit date: `Thu Jun 27 19:07:30 2024 +0100`
10
+
11
+ ## exp/asr_train_raw_en_bpe10000_sp
12
+ ### WER
13
+
14
+ |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
15
+ |---|---|---|---|---|---|---|---|---|
16
+ |decode_asr_model_valid.acc.ave/test|1155|27500|92.1|4.8|3.1|1.1|9.1|70.9|
17
+ |decode_asr_model_valid.acc.ave/test_hb|1695|28708|91.2|4.8|4.0|2.6|11.3|74.6|
18
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_10snr_1s_crowd|1695|28708|89.7|5.5|4.8|2.7|13.0|79.9|
19
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_10snr_1s_white|1695|28708|89.5|5.7|4.8|2.9|13.5|80.5|
20
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_10snr_2s_crowd|1695|28708|88.6|6.3|5.1|2.8|14.2|82.1|
21
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_10snr_2s_white|1695|28708|88.4|6.6|5.0|3.0|14.6|82.0|
22
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_5snr_1s_crowd|1695|28708|87.6|6.3|6.1|2.7|15.1|83.7|
23
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_5snr_1s_white|1695|28708|87.4|6.7|5.9|3.0|15.5|84.1|
24
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_5snr_2s_crowd|1695|28708|83.8|9.6|6.6|3.2|19.5|89.3|
25
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_5snr_2s_white|1695|28708|84.4|9.1|6.5|3.1|18.7|87.3|
26
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_1s_10snr|855|28699|87.6|8.1|4.2|5.9|18.3|99.8|
27
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr|855|28708|87.6|7.8|4.6|6.1|18.5|99.8|
28
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr_reverb|855|28708|62.4|25.7|11.9|7.4|45.0|100.0|
29
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr_reverb_varsnr|0|0|0.0|0.0|0.0|0.0|0.0|0.0|
30
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr_silnoise|855|28699|89.7|6.0|4.3|5.0|15.3|99.4|
31
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr_varsnr|855|28708|87.0|8.2|4.8|7.6|20.6|99.6|
32
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr_varsnr_reverb|855|28708|61.0|26.3|12.8|8.6|47.7|100.0|
33
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_2snr_silnoise|855|28699|89.7|6.2|4.1|7.3|17.6|99.3|
34
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr|855|28699|81.5|12.8|5.6|9.5|28.0|99.9|
35
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_reverb|855|28699|44.9|31.3|23.9|5.6|60.7|100.0|
36
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_reverb_varsnr|0|0|0.0|0.0|0.0|0.0|0.0|0.0|
37
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_silnoise|855|28699|89.6|6.1|4.2|5.7|16.0|99.5|
38
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_silnoise_reverb|855|28699|71.9|20.0|8.1|15.2|43.4|100.0|
39
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_varsnr|855|28708|81.1|13.4|5.6|11.8|30.7|99.9|
40
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_varsnr_reverb|855|28699|44.1|31.0|24.9|5.9|61.8|100.0|
41
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_white|855|28699|79.4|14.2|6.4|6.1|26.8|99.9|
42
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_10snr|855|28699|83.9|8.2|7.9|7.0|23.1|99.4|
43
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_10snr_silnoise|855|28699|87.5|6.5|6.0|5.5|18.0|99.8|
44
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_10snr_varsnr|855|28699|82.9|8.5|8.6|8.4|25.5|99.8|
45
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_10snr_varsnr_reverb|855|28699|52.4|23.8|23.7|6.0|53.6|100.0|
46
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_5snr_silnoise|855|28699|87.2|6.8|6.1|6.5|19.3|99.6|
47
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_5snr_varsnr|855|28699|75.7|12.4|11.9|10.3|34.5|100.0|
48
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_nonoise|855|28699|88.8|6.4|4.8|5.7|16.9|99.6|
49
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_nonoise|855|28699|90.5|5.6|4.0|5.1|14.7|99.6|
50
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_nonoise_reverb|855|28699|76.6|17.3|6.1|6.8|30.2|100.0|
51
+ |decode_asr_model_valid.acc.ave/test_hb_noisy_10snr_crowd1|1695|28699|85.5|9.2|5.3|3.4|17.9|84.8|
52
+ |decode_asr_model_valid.acc.ave/test_hb_noisy_10snr_white|1695|28699|85.6|9.2|5.2|3.0|17.4|84.4|
53
+ |decode_asr_model_valid.acc.ave/test_hb_noisyspeech_10snr_crowd1|1695|28708|85.5|9.2|5.2|3.4|17.8|84.8|
54
+ |decode_asr_model_valid.acc.ave/test_hb_noisyspeech_10snr_white|1695|28708|85.6|9.2|5.2|3.0|17.4|84.3|
55
+ |decode_asr_model_valid.acc.ave/test_hb_noisyspeech_5snr_crowd1|1695|28708|72.7|20.1|7.2|6.3|33.6|93.3|
56
+ |decode_asr_model_valid.acc.ave/test_hb_noisyspeech_5snr_white|1695|28708|76.2|16.9|6.8|4.0|27.8|90.4|
57
+ |decode_asr_model_valid.acc.ave/test_heldback|1564|26404|89.9|5.6|4.6|2.7|12.9|78.8|
58
+
59
+ ### CER
60
+
61
+ |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
62
+ |---|---|---|---|---|---|---|---|---|
63
+ |decode_asr_model_valid.acc.ave/test|1155|145066|95.9|1.2|2.9|1.0|5.1|70.9|
64
+ |decode_asr_model_valid.acc.ave/test_hb|1695|150646|95.3|1.2|3.6|2.2|6.9|74.6|
65
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_10snr_1s_crowd|1695|150646|94.3|1.4|4.2|2.4|8.0|79.9|
66
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_10snr_1s_white|1695|150646|94.2|1.5|4.3|2.5|8.3|80.5|
67
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_10snr_2s_crowd|1695|150646|93.7|1.8|4.5|2.5|8.8|82.1|
68
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_10snr_2s_white|1695|150646|93.5|1.9|4.6|2.5|9.1|82.0|
69
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_5snr_1s_crowd|1695|150646|92.8|1.8|5.4|2.4|9.5|83.7|
70
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_5snr_1s_white|1695|150646|92.7|2.0|5.3|2.5|9.8|84.1|
71
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_5snr_2s_crowd|1695|150646|90.6|3.2|6.2|3.1|12.4|89.3|
72
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_5snr_2s_white|1695|150646|90.9|2.8|6.3|2.7|11.9|87.3|
73
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_1s_10snr|855|154052|92.6|2.5|4.9|6.2|13.6|99.8|
74
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr|855|151486|93.3|2.3|4.4|5.4|12.1|99.8|
75
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr_reverb|855|151486|77.0|9.5|13.5|6.8|29.8|100.0|
76
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr_reverb_varsnr|0|0|0.0|0.0|0.0|0.0|0.0|0.0|
77
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr_silnoise|855|154052|93.2|1.8|4.9|6.4|13.1|99.4|
78
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr_varsnr|855|151486|93.0|2.5|4.5|7.1|14.1|99.6|
79
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr_varsnr_reverb|855|151486|75.8|9.8|14.4|7.9|32.1|100.0|
80
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_2snr_silnoise|855|154052|93.3|2.0|4.8|7.7|14.4|99.3|
81
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr|855|154052|88.8|4.5|6.7|8.6|19.8|99.9|
82
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_reverb|855|154052|60.8|11.5|27.7|5.7|44.9|100.0|
83
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_reverb_varsnr|0|0|0.0|0.0|0.0|0.0|0.0|0.0|
84
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_silnoise|855|154052|93.2|1.9|4.9|6.7|13.5|99.5|
85
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_silnoise_reverb|855|154052|82.7|7.5|9.8|13.0|30.2|100.0|
86
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_varsnr|855|151486|89.6|4.6|5.8|10.2|20.6|99.9|
87
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_varsnr_reverb|855|154052|59.9|11.4|28.7|6.0|46.1|100.0|
88
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_white|855|154052|87.6|4.8|7.7|6.5|19.0|99.9|
89
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_10snr|855|154052|88.5|2.6|8.9|6.5|18.0|99.4|
90
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_10snr_silnoise|855|154052|91.0|2.0|7.0|6.6|15.6|99.8|
91
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_10snr_varsnr|855|154052|87.6|2.8|9.6|7.6|19.9|99.8|
92
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_10snr_varsnr_reverb|855|154052|64.9|8.6|26.4|5.6|40.7|100.0|
93
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_5snr_silnoise|855|154052|90.8|2.2|7.1|7.3|16.6|99.6|
94
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_5snr_varsnr|855|154052|82.5|4.3|13.2|8.8|26.3|100.0|
95
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_nonoise|855|154052|92.5|1.7|5.8|5.8|13.3|99.6|
96
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_nonoise|855|154052|93.9|1.4|4.6|5.3|11.3|99.6|
97
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_nonoise_reverb|855|154052|86.4|6.1|7.5|7.2|20.8|100.0|
98
+ |decode_asr_model_valid.acc.ave/test_hb_noisy_10snr_crowd1|1695|153212|91.4|2.9|5.7|3.6|12.2|84.8|
99
+ |decode_asr_model_valid.acc.ave/test_hb_noisy_10snr_white|1695|153212|91.3|2.8|5.9|3.3|11.9|84.4|
100
+ |decode_asr_model_valid.acc.ave/test_hb_noisyspeech_10snr_crowd1|1695|150646|92.2|2.9|4.9|3.2|11.0|84.8|
101
+ |decode_asr_model_valid.acc.ave/test_hb_noisyspeech_10snr_white|1695|150646|92.2|2.8|5.0|2.9|10.6|84.3|
102
+ |decode_asr_model_valid.acc.ave/test_hb_noisyspeech_5snr_crowd1|1695|150646|84.8|7.6|7.6|6.4|21.6|93.3|
103
+ |decode_asr_model_valid.acc.ave/test_hb_noisyspeech_5snr_white|1695|150646|86.7|6.0|7.3|4.1|17.4|90.4|
104
+ |decode_asr_model_valid.acc.ave/test_heldback|1564|140769|93.7|1.5|4.9|2.9|9.2|78.8|
105
+
106
+ ### TER
107
+
108
+ |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
109
+ |---|---|---|---|---|---|---|---|---|
110
+ |decode_asr_model_valid.acc.ave/test|1155|31518|90.4|4.7|5.0|1.1|10.7|70.9|
111
+ |decode_asr_model_valid.acc.ave/test_hb|1695|113174|94.7|1.6|3.6|2.3|7.5|74.6|
112
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_10snr_1s_crowd|1695|113174|93.7|1.9|4.4|2.4|8.7|79.9|
113
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_10snr_1s_white|1695|113174|93.6|2.0|4.5|2.6|9.0|80.5|
114
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_10snr_2s_crowd|1695|113174|93.1|2.3|4.7|2.6|9.5|82.1|
115
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_10snr_2s_white|1695|113174|92.8|2.4|4.7|2.6|9.8|82.0|
116
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_5snr_1s_crowd|1695|113174|92.1|2.3|5.6|2.4|10.3|83.7|
117
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_5snr_1s_white|1695|113174|92.0|2.5|5.5|2.5|10.6|84.1|
118
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_5snr_2s_crowd|1695|113174|89.8|3.9|6.4|3.1|13.4|89.3|
119
+ |decode_asr_model_valid.acc.ave/test_hb_initnoise_5snr_2s_white|1695|113174|90.1|3.5|6.4|2.8|12.7|87.3|
120
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_1s_10snr|855|33855|84.1|7.7|8.2|7.4|23.4|99.8|
121
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr|855|113174|92.5|3.0|4.5|5.5|13.0|99.8|
122
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr_reverb|855|113174|75.2|11.3|13.5|6.9|31.7|100.0|
123
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr_reverb_varsnr|0|0|0.0|0.0|0.0|0.0|0.0|0.0|
124
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr_silnoise|855|33855|85.7|5.7|8.6|5.3|19.6|99.4|
125
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr_varsnr|855|113174|92.2|3.2|4.6|7.3|15.1|99.6|
126
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_10snr_varsnr_reverb|855|113174|74.0|11.6|14.4|7.9|34.0|100.0|
127
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_2snr_silnoise|855|33855|85.8|6.2|8.0|7.1|21.3|99.3|
128
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr|855|33855|77.9|12.5|9.6|9.5|31.6|99.9|
129
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_reverb|855|33855|42.4|29.2|28.4|5.6|63.2|100.0|
130
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_reverb_varsnr|0|0|0.0|0.0|0.0|0.0|0.0|0.0|
131
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_silnoise|855|33855|85.7|5.9|8.4|5.8|20.1|99.5|
132
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_silnoise_reverb|855|33855|67.9|19.6|12.5|13.5|45.6|100.0|
133
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_varsnr|855|113174|88.5|5.6|6.0|10.2|21.8|99.9|
134
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_varsnr_reverb|855|33855|41.7|28.9|29.4|6.0|64.2|100.0|
135
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_2s_5snr_white|855|33855|75.7|13.2|11.1|7.8|32.1|99.9|
136
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_10snr|855|33855|80.3|8.1|11.6|7.8|27.6|99.4|
137
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_10snr_silnoise|855|33855|83.8|6.2|10.0|6.1|22.3|99.8|
138
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_10snr_varsnr|855|33855|79.2|8.5|12.3|8.4|29.2|99.8|
139
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_10snr_varsnr_reverb|855|33855|49.3|22.2|28.5|5.8|56.5|100.0|
140
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_5snr_silnoise|855|33855|83.4|6.6|10.0|6.7|23.3|99.6|
141
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_5snr_varsnr|855|33855|72.0|12.0|15.9|9.8|37.7|100.0|
142
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_3s_nonoise|855|33855|85.5|6.1|8.4|7.9|22.4|99.6|
143
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_nonoise|855|33855|86.9|5.2|7.8|7.0|20.0|99.6|
144
+ |decode_asr_model_valid.acc.ave/test_hb_multi2_nonoise_reverb|855|33855|72.8|16.4|10.7|8.3|35.5|100.0|
145
+ |decode_asr_model_valid.acc.ave/test_hb_noisy_10snr_crowd1|1695|33855|82.1|8.5|9.4|3.9|21.9|84.8|
146
+ |decode_asr_model_valid.acc.ave/test_hb_noisy_10snr_white|1695|33855|81.9|8.3|9.8|3.4|21.5|84.4|
147
+ |decode_asr_model_valid.acc.ave/test_hb_noisyspeech_10snr_crowd1|1695|113174|91.3|3.6|5.1|3.3|11.9|84.8|
148
+ |decode_asr_model_valid.acc.ave/test_hb_noisyspeech_10snr_white|1695|113174|91.3|3.5|5.2|2.9|11.6|84.3|
149
+ |decode_asr_model_valid.acc.ave/test_hb_noisyspeech_5snr_crowd1|1695|113174|83.2|8.9|7.8|6.3|23.0|93.3|
150
+ |decode_asr_model_valid.acc.ave/test_hb_noisyspeech_5snr_white|1695|113174|85.4|7.0|7.6|4.1|18.7|90.4|
151
+ |decode_asr_model_valid.acc.ave/test_heldback|1564|31126|86.5|5.1|8.3|3.4|16.8|78.8|
152
+