Hugging Face's logo

dbdmg
/
wav2vec2-xls-r-300m-italian

Automatic Speech Recognition
Transformers
PyTorch
Safetensors
Italian
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
g8a9 commited on
Commit
d72f759
·
1 Parent(s): 85f14aa

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +177 -18
README.md CHANGED
@@ -1,18 +1,177 @@
1
- ---
2
-
3
- language:
4
- - it
5
- thumbnail: ""
6
- tags:
7
- - it
8
- - robust-speech-event
9
- - automatic-speech-recognition
10
- - mozilla-foundation/common_voice_7_0
11
- license: mit
12
- datasets:
13
- - mozilla-foundation/common_voice_7_0
14
- metrics:
15
- - CER
16
- - WER
17
-
18
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - it
4
+ license: apache-2.0
5
+ tags:
6
+ - automatic-speech-recognition
7
+ - mozilla-foundation/common_voice_7_0
8
+ - generated_from_trainer
9
+ datasets:
10
+ - common_voice
11
+ model-index:
12
+ - name: wav2vec2-xls-r-300m-italian
13
+ results: []
14
+ ---
15
+
16
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
17
+ should probably proofread and complete it, then remove this comment. -->
18
+
19
+ # wav2vec2-xls-r-300m-italian
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 - IT dataset.
22
+ It achieves the following results on the evaluation set:
23
+ - Loss: inf
24
+ - Wer: 0.1710
25
+
26
+ ## Model description
27
+
28
+ More information needed
29
+
30
+ ## Intended uses & limitations
31
+
32
+ More information needed
33
+
34
+ ## Training and evaluation data
35
+
36
+ More information needed
37
+
38
+ ## Training procedure
39
+
40
+ ### Training hyperparameters
41
+
42
+ The following hyperparameters were used during training:
43
+ - learning_rate: 0.0003
44
+ - train_batch_size: 64
45
+ - eval_batch_size: 8
46
+ - seed: 42
47
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
48
+ - lr_scheduler_type: linear
49
+ - lr_scheduler_warmup_steps: 500
50
+ - num_epochs: 5.0
51
+ - mixed_precision_training: Native AMP
52
+
53
+ ### Training results
54
+
55
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
56
+ |:-------------:|:-----:|:-----:|:---------------:|:------:|
57
+ | No log | 0.04 | 100 | inf | 1.0 |
58
+ | No log | 0.09 | 200 | inf | 0.9983 |
59
+ | No log | 0.13 | 300 | inf | 0.7672 |
60
+ | No log | 0.18 | 400 | inf | 0.6919 |
61
+ | 2.9929 | 0.22 | 500 | inf | 0.6266 |
62
+ | 2.9929 | 0.26 | 600 | inf | 0.5513 |
63
+ | 2.9929 | 0.31 | 700 | inf | 0.5081 |
64
+ | 2.9929 | 0.35 | 800 | inf | 0.4945 |
65
+ | 2.9929 | 0.39 | 900 | inf | 0.4720 |
66
+ | 0.5311 | 0.44 | 1000 | inf | 0.4387 |
67
+ | 0.5311 | 0.48 | 1100 | inf | 0.4411 |
68
+ | 0.5311 | 0.53 | 1200 | inf | 0.4429 |
69
+ | 0.5311 | 0.57 | 1300 | inf | 0.4322 |
70
+ | 0.5311 | 0.61 | 1400 | inf | 0.4532 |
71
+ | 0.4654 | 0.66 | 1500 | inf | 0.4492 |
72
+ | 0.4654 | 0.7 | 1600 | inf | 0.3879 |
73
+ | 0.4654 | 0.75 | 1700 | inf | 0.3836 |
74
+ | 0.4654 | 0.79 | 1800 | inf | 0.3743 |
75
+ | 0.4654 | 0.83 | 1900 | inf | 0.3687 |
76
+ | 0.4254 | 0.88 | 2000 | inf | 0.3793 |
77
+ | 0.4254 | 0.92 | 2100 | inf | 0.3766 |
78
+ | 0.4254 | 0.97 | 2200 | inf | 0.3705 |
79
+ | 0.4254 | 1.01 | 2300 | inf | 0.3272 |
80
+ | 0.4254 | 1.05 | 2400 | inf | 0.3185 |
81
+ | 0.3997 | 1.1 | 2500 | inf | 0.3244 |
82
+ | 0.3997 | 1.14 | 2600 | inf | 0.3082 |
83
+ | 0.3997 | 1.18 | 2700 | inf | 0.3040 |
84
+ | 0.3997 | 1.23 | 2800 | inf | 0.3028 |
85
+ | 0.3997 | 1.27 | 2900 | inf | 0.3112 |
86
+ | 0.3668 | 1.32 | 3000 | inf | 0.3110 |
87
+ | 0.3668 | 1.36 | 3100 | inf | 0.3067 |
88
+ | 0.3668 | 1.4 | 3200 | inf | 0.2961 |
89
+ | 0.3668 | 1.45 | 3300 | inf | 0.3081 |
90
+ | 0.3668 | 1.49 | 3400 | inf | 0.2936 |
91
+ | 0.3645 | 1.54 | 3500 | inf | 0.3037 |
92
+ | 0.3645 | 1.58 | 3600 | inf | 0.2974 |
93
+ | 0.3645 | 1.62 | 3700 | inf | 0.3010 |
94
+ | 0.3645 | 1.67 | 3800 | inf | 0.2985 |
95
+ | 0.3645 | 1.71 | 3900 | inf | 0.2976 |
96
+ | 0.3624 | 1.76 | 4000 | inf | 0.2928 |
97
+ | 0.3624 | 1.8 | 4100 | inf | 0.2860 |
98
+ | 0.3624 | 1.84 | 4200 | inf | 0.2922 |
99
+ | 0.3624 | 1.89 | 4300 | inf | 0.2866 |
100
+ | 0.3624 | 1.93 | 4400 | inf | 0.2776 |
101
+ | 0.3527 | 1.97 | 4500 | inf | 0.2792 |
102
+ | 0.3527 | 2.02 | 4600 | inf | 0.2858 |
103
+ | 0.3527 | 2.06 | 4700 | inf | 0.2767 |
104
+ | 0.3527 | 2.11 | 4800 | inf | 0.2824 |
105
+ | 0.3527 | 2.15 | 4900 | inf | 0.2799 |
106
+ | 0.3162 | 2.19 | 5000 | inf | 0.2673 |
107
+ | 0.3162 | 2.24 | 5100 | inf | 0.2962 |
108
+ | 0.3162 | 2.28 | 5200 | inf | 0.2736 |
109
+ | 0.3162 | 2.33 | 5300 | inf | 0.2652 |
110
+ | 0.3162 | 2.37 | 5400 | inf | 0.2551 |
111
+ | 0.3063 | 2.41 | 5500 | inf | 0.2680 |
112
+ | 0.3063 | 2.46 | 5600 | inf | 0.2558 |
113
+ | 0.3063 | 2.5 | 5700 | inf | 0.2598 |
114
+ | 0.3063 | 2.54 | 5800 | inf | 0.2518 |
115
+ | 0.3063 | 2.59 | 5900 | inf | 0.2541 |
116
+ | 0.2913 | 2.63 | 6000 | inf | 0.2507 |
117
+ | 0.2913 | 2.68 | 6100 | inf | 0.2500 |
118
+ | 0.2913 | 2.72 | 6200 | inf | 0.2435 |
119
+ | 0.2913 | 2.76 | 6300 | inf | 0.2376 |
120
+ | 0.2913 | 2.81 | 6400 | inf | 0.2348 |
121
+ | 0.2797 | 2.85 | 6500 | inf | 0.2512 |
122
+ | 0.2797 | 2.9 | 6600 | inf | 0.2382 |
123
+ | 0.2797 | 2.94 | 6700 | inf | 0.2523 |
124
+ | 0.2797 | 2.98 | 6800 | inf | 0.2522 |
125
+ | 0.2797 | 3.03 | 6900 | inf | 0.2409 |
126
+ | 0.2766 | 3.07 | 7000 | inf | 0.2453 |
127
+ | 0.2766 | 3.12 | 7100 | inf | 0.2326 |
128
+ | 0.2766 | 3.16 | 7200 | inf | 0.2286 |
129
+ | 0.2766 | 3.2 | 7300 | inf | 0.2342 |
130
+ | 0.2766 | 3.25 | 7400 | inf | 0.2305 |
131
+ | 0.2468 | 3.29 | 7500 | inf | 0.2238 |
132
+ | 0.2468 | 3.33 | 7600 | inf | 0.2321 |
133
+ | 0.2468 | 3.38 | 7700 | inf | 0.2305 |
134
+ | 0.2468 | 3.42 | 7800 | inf | 0.2174 |
135
+ | 0.2468 | 3.47 | 7900 | inf | 0.2201 |
136
+ | 0.2439 | 3.51 | 8000 | inf | 0.2133 |
137
+ | 0.2439 | 3.55 | 8100 | inf | 0.2217 |
138
+ | 0.2439 | 3.6 | 8200 | inf | 0.2189 |
139
+ | 0.2439 | 3.64 | 8300 | inf | 0.2105 |
140
+ | 0.2439 | 3.69 | 8400 | inf | 0.2118 |
141
+ | 0.2357 | 3.73 | 8500 | inf | 0.2093 |
142
+ | 0.2357 | 3.77 | 8600 | inf | 0.2103 |
143
+ | 0.2357 | 3.82 | 8700 | inf | 0.2035 |
144
+ | 0.2357 | 3.86 | 8800 | inf | 0.2019 |
145
+ | 0.2357 | 3.91 | 8900 | inf | 0.2032 |
146
+ | 0.2217 | 3.95 | 9000 | inf | 0.2056 |
147
+ | 0.2217 | 3.99 | 9100 | inf | 0.2022 |
148
+ | 0.2217 | 4.04 | 9200 | inf | 0.1932 |
149
+ | 0.2217 | 4.08 | 9300 | inf | 0.1935 |
150
+ | 0.2217 | 4.12 | 9400 | inf | 0.1906 |
151
+ | 0.2025 | 4.17 | 9500 | inf | 0.1879 |
152
+ | 0.2025 | 4.21 | 9600 | inf | 0.1882 |
153
+ | 0.2025 | 4.26 | 9700 | inf | 0.1854 |
154
+ | 0.2025 | 4.3 | 9800 | inf | 0.1865 |
155
+ | 0.2025 | 4.34 | 9900 | inf | 0.1844 |
156
+ | 0.1869 | 4.39 | 10000 | inf | 0.1822 |
157
+ | 0.1869 | 4.43 | 10100 | inf | 0.1815 |
158
+ | 0.1869 | 4.48 | 10200 | inf | 0.1812 |
159
+ | 0.1869 | 4.52 | 10300 | inf | 0.1792 |
160
+ | 0.1869 | 4.56 | 10400 | inf | 0.1797 |
161
+ | 0.1863 | 4.61 | 10500 | inf | 0.1774 |
162
+ | 0.1863 | 4.65 | 10600 | inf | 0.1767 |
163
+ | 0.1863 | 4.7 | 10700 | inf | 0.1765 |
164
+ | 0.1863 | 4.74 | 10800 | inf | 0.1753 |
165
+ | 0.1863 | 4.78 | 10900 | inf | 0.1731 |
166
+ | 0.178 | 4.83 | 11000 | inf | 0.1727 |
167
+ | 0.178 | 4.87 | 11100 | inf | 0.1724 |
168
+ | 0.178 | 4.91 | 11200 | inf | 0.1722 |
169
+ | 0.178 | 4.96 | 11300 | inf | 0.1712 |
170
+
171
+
172
+ ### Framework versions
173
+
174
+ - Transformers 4.16.0.dev0
175
+ - Pytorch 1.10.1+cu102
176
+ - Datasets 1.17.1.dev0
177
+ - Tokenizers 0.11.0