update model card README.md
Browse files
README.md
CHANGED
@@ -13,8 +13,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
13 |
|
14 |
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on an unknown dataset.
|
15 |
It achieves the following results on the evaluation set:
|
16 |
-
- Loss:
|
17 |
-
- Wer: 1.
|
18 |
|
19 |
## Model description
|
20 |
|
@@ -34,77 +34,77 @@ More information needed
|
|
34 |
|
35 |
The following hyperparameters were used during training:
|
36 |
- learning_rate: 0.0001
|
37 |
-
- train_batch_size:
|
38 |
- eval_batch_size: 8
|
39 |
- seed: 42
|
40 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
41 |
- lr_scheduler_type: linear
|
42 |
- lr_scheduler_warmup_steps: 1000
|
43 |
-
- num_epochs:
|
44 |
|
45 |
### Training results
|
46 |
|
47 |
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
48 |
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
49 |
-
|
|
50 |
-
|
|
51 |
-
|
|
52 |
-
|
|
53 |
-
|
|
54 |
-
|
|
55 |
-
|
|
56 |
-
|
|
57 |
-
|
|
58 |
-
|
|
59 |
-
|
|
60 |
-
|
|
61 |
-
|
|
62 |
-
|
|
63 |
-
| 1.
|
64 |
-
| 1.
|
65 |
-
| 1.
|
66 |
-
|
|
67 |
-
|
|
68 |
-
|
|
69 |
-
|
|
70 |
-
|
|
71 |
-
|
|
72 |
-
|
|
73 |
-
|
|
74 |
-
|
|
75 |
-
|
|
76 |
-
|
|
77 |
-
|
|
78 |
-
|
|
79 |
-
|
|
80 |
-
|
|
81 |
-
|
|
82 |
-
|
|
83 |
-
|
|
84 |
-
|
|
85 |
-
|
|
86 |
-
|
|
87 |
-
|
|
88 |
-
|
|
89 |
-
|
|
90 |
-
|
|
91 |
-
|
|
92 |
-
|
|
93 |
-
|
|
94 |
-
|
|
95 |
-
|
|
96 |
-
|
|
97 |
-
|
|
98 |
-
|
|
99 |
-
|
|
100 |
-
|
|
101 |
-
|
|
102 |
-
|
|
103 |
-
|
|
104 |
-
|
|
105 |
-
|
|
106 |
-
|
|
107 |
-
|
|
108 |
|
109 |
|
110 |
### Framework versions
|
|
|
13 |
|
14 |
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on an unknown dataset.
|
15 |
It achieves the following results on the evaluation set:
|
16 |
+
- Loss: 1.7114
|
17 |
+
- Wer: 1.0056
|
18 |
|
19 |
## Model description
|
20 |
|
|
|
34 |
|
35 |
The following hyperparameters were used during training:
|
36 |
- learning_rate: 0.0001
|
37 |
+
- train_batch_size: 16
|
38 |
- eval_batch_size: 8
|
39 |
- seed: 42
|
40 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
41 |
- lr_scheduler_type: linear
|
42 |
- lr_scheduler_warmup_steps: 1000
|
43 |
+
- num_epochs: 60
|
44 |
|
45 |
### Training results
|
46 |
|
47 |
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
48 |
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
49 |
+
| 92.1483 | 1.01 | 100 | 42.2435 | 1.0 |
|
50 |
+
| 63.5438 | 2.02 | 200 | 30.3017 | 1.0 |
|
51 |
+
| 40.1961 | 3.03 | 300 | 16.5376 | 1.0 |
|
52 |
+
| 17.0478 | 4.04 | 400 | 6.9455 | 1.0 |
|
53 |
+
| 6.3103 | 5.05 | 500 | 4.9786 | 1.0 |
|
54 |
+
| 4.8201 | 6.06 | 600 | 4.8737 | 1.0 |
|
55 |
+
| 4.6806 | 7.07 | 700 | 4.8619 | 1.0 |
|
56 |
+
| 4.6275 | 8.08 | 800 | 4.8368 | 1.0 |
|
57 |
+
| 4.6017 | 9.09 | 900 | 4.8350 | 1.0 |
|
58 |
+
| 4.5535 | 10.1 | 1000 | 4.8050 | 1.0 |
|
59 |
+
| 4.4901 | 11.11 | 1100 | 4.6832 | 1.0 |
|
60 |
+
| 3.9919 | 12.12 | 1200 | 3.8253 | 1.0 |
|
61 |
+
| 2.7548 | 13.13 | 1300 | 2.9664 | 1.1167 |
|
62 |
+
| 2.0868 | 14.14 | 1400 | 2.6503 | 1.1778 |
|
63 |
+
| 1.6825 | 15.15 | 1500 | 2.4266 | 1.0278 |
|
64 |
+
| 1.3986 | 16.16 | 1600 | 2.2006 | 1.0944 |
|
65 |
+
| 1.1716 | 17.17 | 1700 | 2.1571 | 1.0389 |
|
66 |
+
| 0.9922 | 18.18 | 1800 | 2.0690 | 1.0333 |
|
67 |
+
| 0.8654 | 19.19 | 1900 | 1.9874 | 1.0222 |
|
68 |
+
| 0.7447 | 20.2 | 2000 | 1.9131 | 1.0278 |
|
69 |
+
| 0.6263 | 21.21 | 2100 | 1.8423 | 1.0111 |
|
70 |
+
| 0.5429 | 22.22 | 2200 | 1.7591 | 1.0111 |
|
71 |
+
| 0.4715 | 23.23 | 2300 | 1.7380 | 1.0 |
|
72 |
+
| 0.4245 | 24.24 | 2400 | 1.7531 | 1.0111 |
|
73 |
+
| 0.3971 | 25.25 | 2500 | 1.8016 | 1.0 |
|
74 |
+
| 0.3495 | 26.26 | 2600 | 1.6868 | 0.9944 |
|
75 |
+
| 0.317 | 27.27 | 2700 | 1.8013 | 1.0 |
|
76 |
+
| 0.2886 | 28.28 | 2800 | 1.6985 | 1.0056 |
|
77 |
+
| 0.2597 | 29.29 | 2900 | 1.6931 | 1.0056 |
|
78 |
+
| 0.2269 | 30.3 | 3000 | 1.6208 | 1.0111 |
|
79 |
+
| 0.2175 | 31.31 | 3100 | 1.6451 | 1.0 |
|
80 |
+
| 0.2101 | 32.32 | 3200 | 1.7385 | 1.0056 |
|
81 |
+
| 0.2033 | 33.33 | 3300 | 1.6909 | 1.0111 |
|
82 |
+
| 0.184 | 34.34 | 3400 | 1.7458 | 1.0 |
|
83 |
+
| 0.1812 | 35.35 | 3500 | 1.7152 | 1.0 |
|
84 |
+
| 0.1582 | 36.36 | 3600 | 1.7101 | 1.0 |
|
85 |
+
| 0.1556 | 37.37 | 3700 | 1.6729 | 1.0056 |
|
86 |
+
| 0.1398 | 38.38 | 3800 | 1.6982 | 0.9944 |
|
87 |
+
| 0.1357 | 39.39 | 3900 | 1.6891 | 1.0167 |
|
88 |
+
| 0.1261 | 40.4 | 4000 | 1.6817 | 1.0 |
|
89 |
+
| 0.1226 | 41.41 | 4100 | 1.7411 | 1.0 |
|
90 |
+
| 0.1217 | 42.42 | 4200 | 1.7909 | 1.0 |
|
91 |
+
| 0.115 | 43.43 | 4300 | 1.6764 | 1.0 |
|
92 |
+
| 0.1127 | 44.44 | 4400 | 1.6728 | 0.9944 |
|
93 |
+
| 0.104 | 45.45 | 4500 | 1.7181 | 1.0056 |
|
94 |
+
| 0.105 | 46.46 | 4600 | 1.7491 | 1.0 |
|
95 |
+
| 0.0925 | 47.47 | 4700 | 1.7661 | 1.0056 |
|
96 |
+
| 0.0942 | 48.48 | 4800 | 1.7376 | 1.0 |
|
97 |
+
| 0.0943 | 49.49 | 4900 | 1.6908 | 1.0111 |
|
98 |
+
| 0.0859 | 50.51 | 5000 | 1.7193 | 1.0111 |
|
99 |
+
| 0.0845 | 51.52 | 5100 | 1.7051 | 1.0 |
|
100 |
+
| 0.0825 | 52.53 | 5200 | 1.6917 | 1.0 |
|
101 |
+
| 0.08 | 53.54 | 5300 | 1.7093 | 1.0111 |
|
102 |
+
| 0.0749 | 54.55 | 5400 | 1.7160 | 1.0111 |
|
103 |
+
| 0.0731 | 55.56 | 5500 | 1.7215 | 1.0111 |
|
104 |
+
| 0.0725 | 56.57 | 5600 | 1.7251 | 1.0056 |
|
105 |
+
| 0.0704 | 57.58 | 5700 | 1.7275 | 1.0056 |
|
106 |
+
| 0.0734 | 58.59 | 5800 | 1.7119 | 1.0056 |
|
107 |
+
| 0.0729 | 59.6 | 5900 | 1.7114 | 1.0056 |
|
108 |
|
109 |
|
110 |
### Framework versions
|