Hugging Face's logo

AndrewMcDowell
/
wav2vec2-xls-r-300m-german-de

Automatic Speech Recognition
Transformers
PyTorch
German
wav2vec2
Generated from Trainer
hf-asr-leaderboard
mozilla-foundation/common_voice_7_0
robust-speech-event
Eval Results
Inference Endpoints
Model card Files Community
1
AndrewMcDowell commited on
Commit
31af956
•
1 Parent(s): 20de31b

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +89 -0
README.md CHANGED
@@ -19,6 +19,9 @@ should probably proofread and complete it, then remove this comment. -->
19
  #
20
 
21
  This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the MOZILLA-FOUNDATION/COMMON_VOICE_7_0 - DE dataset.
 
 
 
22
 
23
  ## Model description
24
 
@@ -49,6 +52,92 @@ The following hyperparameters were used during training:
49
  - num_epochs: 0.0
50
  - mixed_precision_training: Native AMP
51
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
52
  ### Framework versions
53
 
54
  - Transformers 4.16.0.dev0
 
19
  #
20
 
21
  This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the MOZILLA-FOUNDATION/COMMON_VOICE_7_0 - DE dataset.
22
+ It achieves the following results on the evaluation set:
23
+ - Loss: 0.1768
24
+ - Wer: 0.2016
25
 
26
  ## Model description
27
 
 
52
  - num_epochs: 0.0
53
  - mixed_precision_training: Native AMP
54
 
55
+ ### Training results
56
+
57
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
58
+ |:-------------:|:-----:|:-----:|:---------------:|:------:|
59
+ | 5.7531 | 0.04 | 500 | 5.4564 | 1.0 |
60
+ | 2.9882 | 0.08 | 1000 | 3.0041 | 1.0 |
61
+ | 2.1953 | 0.13 | 1500 | 1.1723 | 0.7121 |
62
+ | 1.2406 | 0.17 | 2000 | 0.3656 | 0.3623 |
63
+ | 1.1294 | 0.21 | 2500 | 0.2843 | 0.2926 |
64
+ | 1.0731 | 0.25 | 3000 | 0.2554 | 0.2664 |
65
+ | 1.051 | 0.3 | 3500 | 0.2387 | 0.2535 |
66
+ | 1.0479 | 0.34 | 4000 | 0.2345 | 0.2512 |
67
+ | 1.0026 | 0.38 | 4500 | 0.2270 | 0.2452 |
68
+ | 0.9921 | 0.42 | 5000 | 0.2212 | 0.2353 |
69
+ | 0.9839 | 0.47 | 5500 | 0.2141 | 0.2330 |
70
+ | 0.9907 | 0.51 | 6000 | 0.2122 | 0.2334 |
71
+ | 0.9788 | 0.55 | 6500 | 0.2114 | 0.2270 |
72
+ | 0.9687 | 0.59 | 7000 | 0.2066 | 0.2323 |
73
+ | 0.9777 | 0.64 | 7500 | 0.2033 | 0.2237 |
74
+ | 0.9476 | 0.68 | 8000 | 0.2020 | 0.2194 |
75
+ | 0.9625 | 0.72 | 8500 | 0.1977 | 0.2191 |
76
+ | 0.9497 | 0.76 | 9000 | 0.1976 | 0.2175 |
77
+ | 0.9781 | 0.81 | 9500 | 0.1956 | 0.2159 |
78
+ | 0.9552 | 0.85 | 10000 | 0.1958 | 0.2191 |
79
+ | 0.9345 | 0.89 | 10500 | 0.1964 | 0.2158 |
80
+ | 0.9528 | 0.93 | 11000 | 0.1926 | 0.2154 |
81
+ | 0.9502 | 0.98 | 11500 | 0.1953 | 0.2149 |
82
+ | 0.9358 | 1.02 | 12000 | 0.1927 | 0.2167 |
83
+ | 0.941 | 1.06 | 12500 | 0.1901 | 0.2115 |
84
+ | 0.9287 | 1.1 | 13000 | 0.1936 | 0.2090 |
85
+ | 0.9491 | 1.15 | 13500 | 0.1900 | 0.2104 |
86
+ | 0.9478 | 1.19 | 14000 | 0.1931 | 0.2120 |
87
+ | 0.946 | 1.23 | 14500 | 0.1914 | 0.2134 |
88
+ | 0.9499 | 1.27 | 15000 | 0.1931 | 0.2173 |
89
+ | 0.9346 | 1.32 | 15500 | 0.1913 | 0.2105 |
90
+ | 0.9509 | 1.36 | 16000 | 0.1902 | 0.2137 |
91
+ | 0.9294 | 1.4 | 16500 | 0.1895 | 0.2086 |
92
+ | 0.9418 | 1.44 | 17000 | 0.1913 | 0.2183 |
93
+ | 0.9302 | 1.49 | 17500 | 0.1884 | 0.2114 |
94
+ | 0.9418 | 1.53 | 18000 | 0.1894 | 0.2108 |
95
+ | 0.9363 | 1.57 | 18500 | 0.1886 | 0.2132 |
96
+ | 0.9338 | 1.61 | 19000 | 0.1856 | 0.2078 |
97
+ | 0.9185 | 1.66 | 19500 | 0.1852 | 0.2056 |
98
+ | 0.9216 | 1.7 | 20000 | 0.1874 | 0.2095 |
99
+ | 0.9176 | 1.74 | 20500 | 0.1873 | 0.2078 |
100
+ | 0.9288 | 1.78 | 21000 | 0.1865 | 0.2097 |
101
+ | 0.9278 | 1.83 | 21500 | 0.1869 | 0.2100 |
102
+ | 0.9295 | 1.87 | 22000 | 0.1878 | 0.2095 |
103
+ | 0.9221 | 1.91 | 22500 | 0.1852 | 0.2121 |
104
+ | 0.924 | 1.95 | 23000 | 0.1855 | 0.2042 |
105
+ | 0.9104 | 2.0 | 23500 | 0.1858 | 0.2105 |
106
+ | 0.9284 | 2.04 | 24000 | 0.1850 | 0.2080 |
107
+ | 0.9162 | 2.08 | 24500 | 0.1839 | 0.2045 |
108
+ | 0.9111 | 2.12 | 25000 | 0.1838 | 0.2080 |
109
+ | 0.91 | 2.17 | 25500 | 0.1889 | 0.2106 |
110
+ | 0.9152 | 2.21 | 26000 | 0.1856 | 0.2026 |
111
+ | 0.9209 | 2.25 | 26500 | 0.1891 | 0.2133 |
112
+ | 0.9094 | 2.29 | 27000 | 0.1857 | 0.2089 |
113
+ | 0.9065 | 2.34 | 27500 | 0.1840 | 0.2052 |
114
+ | 0.9156 | 2.38 | 28000 | 0.1833 | 0.2062 |
115
+ | 0.8986 | 2.42 | 28500 | 0.1789 | 0.2001 |
116
+ | 0.9045 | 2.46 | 29000 | 0.1769 | 0.2022 |
117
+ | 0.9039 | 2.51 | 29500 | 0.1819 | 0.2073 |
118
+ | 0.9145 | 2.55 | 30000 | 0.1828 | 0.2063 |
119
+ | 0.9081 | 2.59 | 30500 | 0.1811 | 0.2049 |
120
+ | 0.9252 | 2.63 | 31000 | 0.1833 | 0.2086 |
121
+ | 0.8957 | 2.68 | 31500 | 0.1795 | 0.2083 |
122
+ | 0.891 | 2.72 | 32000 | 0.1809 | 0.2058 |
123
+ | 0.9023 | 2.76 | 32500 | 0.1812 | 0.2061 |
124
+ | 0.8918 | 2.8 | 33000 | 0.1775 | 0.1997 |
125
+ | 0.8852 | 2.85 | 33500 | 0.1790 | 0.1997 |
126
+ | 0.8928 | 2.89 | 34000 | 0.1767 | 0.2013 |
127
+ | 0.9079 | 2.93 | 34500 | 0.1735 | 0.1986 |
128
+ | 0.9032 | 2.97 | 35000 | 0.1793 | 0.2024 |
129
+ | 0.9018 | 3.02 | 35500 | 0.1778 | 0.2027 |
130
+ | 0.8846 | 3.06 | 36000 | 0.1776 | 0.2046 |
131
+ | 0.8848 | 3.1 | 36500 | 0.1812 | 0.2064 |
132
+ | 0.9062 | 3.14 | 37000 | 0.1800 | 0.2018 |
133
+ | 0.9011 | 3.19 | 37500 | 0.1783 | 0.2049 |
134
+ | 0.8996 | 3.23 | 38000 | 0.1810 | 0.2036 |
135
+ | 0.893 | 3.27 | 38500 | 0.1805 | 0.2056 |
136
+ | 0.897 | 3.31 | 39000 | 0.1773 | 0.2035 |
137
+ | 0.8992 | 3.36 | 39500 | 0.1804 | 0.2054 |
138
+ | 0.8987 | 3.4 | 40000 | 0.1768 | 0.2016 |
139
+
140
+
141
  ### Framework versions
142
 
143
  - Transformers 4.16.0.dev0