Hugging Face's logo

foxxy-hm
/
wav2vec2-base-finetune-vi-v2

Automatic Speech Recognition
Transformers
PyTorch
TensorBoard
wav2vec2
Generated from Trainer
Model card Files Metrics Community
4
foxxy-hm commited on
Commit
9d325c9
·
1 Parent(s): eba769f

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +38 -42
README.md CHANGED
@@ -7,16 +7,6 @@ metrics:
7
  model-index:
8
  - name: wav2vec2-base-finetune-vi-v2
9
  results: []
10
- widget:
11
- - example_title: SOICT 2023 - SLU public test 1
12
- src: >-
13
- https://huggingface.co/foxxy-hm/wav2vec2-base-finetune-vi/raw/main/audio-test/055R7BruAa333g9teFfamQH.wav
14
- - example_title: SOICT 2023 - SLU public test 2
15
- src: >-
16
- https://huggingface.co/foxxy-hm/wav2vec2-base-finetune-vi/raw/main/audio-test/0BLHhoJexE8THB8BrsZxWbh.wav
17
- - example_title: SOICT 2023 - SLU public test 3
18
- src: >-
19
- https://huggingface.co/foxxy-hm/wav2vec2-base-finetune-vi/raw/main/audio-test/1ArUTGWJQ9YALH2xaNhU6GV.wav
20
  ---
21
 
22
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -26,8 +16,8 @@ should probably proofread and complete it, then remove this comment. -->
26
 
27
  This model is a fine-tuned version of [nguyenvulebinh/wav2vec2-base-vietnamese-250h](https://huggingface.co/nguyenvulebinh/wav2vec2-base-vietnamese-250h) on the None dataset.
28
  It achieves the following results on the evaluation set:
29
- - Loss: 0.2294
30
- - Wer: 0.1457
31
 
32
  ## Model description
33
 
@@ -53,41 +43,47 @@ The following hyperparameters were used during training:
53
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
54
  - lr_scheduler_type: linear
55
  - lr_scheduler_warmup_steps: 1000
56
- - num_epochs: 20
57
 
58
  ### Training results
59
 
60
  | Training Loss | Epoch | Step | Validation Loss | Wer |
61
  |:-------------:|:-----:|:-----:|:---------------:|:------:|
62
- | 13.1354 | 0.67 | 500 | 3.0881 | 1.0186 |
63
- | 2.2088 | 1.34 | 1000 | 0.9805 | 0.4257 |
64
- | 1.122 | 2.0 | 1500 | 0.4928 | 0.2850 |
65
- | 0.7567 | 2.67 | 2000 | 0.4217 | 0.2466 |
66
- | 0.627 | 3.34 | 2500 | 0.3889 | 0.2212 |
67
- | 0.5369 | 4.01 | 3000 | 0.3496 | 0.2131 |
68
- | 0.4485 | 4.67 | 3500 | 0.3239 | 0.1994 |
69
- | 0.4478 | 5.34 | 4000 | 0.3143 | 0.1944 |
70
- | 0.4013 | 6.01 | 4500 | 0.2989 | 0.1871 |
71
- | 0.4542 | 6.68 | 5000 | 0.2996 | 0.1871 |
72
- | 0.351 | 7.34 | 5500 | 0.2719 | 0.1736 |
73
- | 0.3236 | 8.01 | 6000 | 0.2865 | 0.1702 |
74
- | 0.2954 | 8.68 | 6500 | 0.2708 | 0.1636 |
75
- | 0.3533 | 9.35 | 7000 | 0.2712 | 0.1639 |
76
- | 0.2996 | 10.01 | 7500 | 0.2609 | 0.1621 |
77
- | 0.2595 | 10.68 | 8000 | 0.2450 | 0.1627 |
78
- | 0.2914 | 11.35 | 8500 | 0.2748 | 0.1596 |
79
- | 0.253 | 12.02 | 9000 | 0.2496 | 0.1552 |
80
- | 0.2314 | 12.68 | 9500 | 0.2496 | 0.1549 |
81
- | 0.2232 | 13.35 | 10000 | 0.2594 | 0.1557 |
82
- | 0.2206 | 14.02 | 10500 | 0.2485 | 0.1529 |
83
- | 0.2026 | 14.69 | 11000 | 0.2365 | 0.1522 |
84
- | 0.2009 | 15.35 | 11500 | 0.2396 | 0.1513 |
85
- | 0.205 | 16.02 | 12000 | 0.2433 | 0.1499 |
86
- | 0.207 | 16.69 | 12500 | 0.2363 | 0.1496 |
87
- | 0.1895 | 17.36 | 13000 | 0.2280 | 0.1481 |
88
- | 0.1991 | 18.02 | 13500 | 0.2352 | 0.1481 |
89
- | 0.2109 | 18.69 | 14000 | 0.2353 | 0.1477 |
90
- | 0.1959 | 19.36 | 14500 | 0.2294 | 0.1457 |
 
 
 
 
 
 
91
 
92
 
93
  ### Framework versions
 
7
  model-index:
8
  - name: wav2vec2-base-finetune-vi-v2
9
  results: []
 
 
 
 
 
 
 
 
 
 
10
  ---
11
 
12
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
16
 
17
  This model is a fine-tuned version of [nguyenvulebinh/wav2vec2-base-vietnamese-250h](https://huggingface.co/nguyenvulebinh/wav2vec2-base-vietnamese-250h) on the None dataset.
18
  It achieves the following results on the evaluation set:
19
+ - Loss: 0.2188
20
+ - Wer: 0.1391
21
 
22
  ## Model description
23
 
 
43
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
44
  - lr_scheduler_type: linear
45
  - lr_scheduler_warmup_steps: 1000
46
+ - num_epochs: 24
47
 
48
  ### Training results
49
 
50
  | Training Loss | Epoch | Step | Validation Loss | Wer |
51
  |:-------------:|:-----:|:-----:|:---------------:|:------:|
52
+ | 4.3873 | 0.67 | 500 | 2.4321 | 0.9719 |
53
+ | 1.4812 | 1.34 | 1000 | 0.5449 | 0.3062 |
54
+ | 0.7731 | 2.0 | 1500 | 0.3793 | 0.2263 |
55
+ | 0.542 | 2.67 | 2000 | 0.3021 | 0.2002 |
56
+ | 0.4461 | 3.34 | 2500 | 0.2905 | 0.1862 |
57
+ | 0.4175 | 4.01 | 3000 | 0.2687 | 0.1771 |
58
+ | 0.3878 | 4.67 | 3500 | 0.2958 | 0.1751 |
59
+ | 0.3373 | 5.34 | 4000 | 0.2713 | 0.1721 |
60
+ | 0.3046 | 6.01 | 4500 | 0.2505 | 0.1616 |
61
+ | 0.2933 | 6.68 | 5000 | 0.2561 | 0.1611 |
62
+ | 0.285 | 7.34 | 5500 | 0.2405 | 0.1617 |
63
+ | 0.2998 | 8.01 | 6000 | 0.2363 | 0.1578 |
64
+ | 0.2486 | 8.68 | 6500 | 0.2254 | 0.1570 |
65
+ | 0.2682 | 9.35 | 7000 | 0.2306 | 0.1547 |
66
+ | 0.2327 | 10.01 | 7500 | 0.2289 | 0.1537 |
67
+ | 0.2141 | 10.68 | 8000 | 0.2383 | 0.1499 |
68
+ | 0.2124 | 11.35 | 8500 | 0.2261 | 0.15 |
69
+ | 0.2156 | 12.02 | 9000 | 0.2142 | 0.1511 |
70
+ | 0.2082 | 12.68 | 9500 | 0.2386 | 0.1467 |
71
+ | 0.1814 | 13.35 | 10000 | 0.2301 | 0.1448 |
72
+ | 0.1836 | 14.02 | 10500 | 0.2302 | 0.1446 |
73
+ | 0.18 | 14.69 | 11000 | 0.2244 | 0.1445 |
74
+ | 0.1756 | 15.35 | 11500 | 0.2280 | 0.1439 |
75
+ | 0.1693 | 16.02 | 12000 | 0.2307 | 0.1426 |
76
+ | 0.1588 | 16.69 | 12500 | 0.2164 | 0.1422 |
77
+ | 0.1587 | 17.36 | 13000 | 0.2198 | 0.1417 |
78
+ | 0.1738 | 18.02 | 13500 | 0.2282 | 0.1411 |
79
+ | 0.1524 | 18.69 | 14000 | 0.2274 | 0.1394 |
80
+ | 0.1569 | 19.36 | 14500 | 0.2178 | 0.1396 |
81
+ | 0.1433 | 20.03 | 15000 | 0.2200 | 0.1413 |
82
+ | 0.1512 | 20.69 | 15500 | 0.2193 | 0.1382 |
83
+ | 0.1375 | 21.36 | 16000 | 0.2174 | 0.1393 |
84
+ | 0.1302 | 22.03 | 16500 | 0.2246 | 0.1391 |
85
+ | 0.146 | 22.7 | 17000 | 0.2222 | 0.1392 |
86
+ | 0.1265 | 23.36 | 17500 | 0.2188 | 0.1391 |
87
 
88
 
89
  ### Framework versions