update model card README.md
Browse files
README.md
ADDED
@@ -0,0 +1,140 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
tags:
|
4 |
+
- generated_from_trainer
|
5 |
+
datasets:
|
6 |
+
- torgo
|
7 |
+
metrics:
|
8 |
+
- wer
|
9 |
+
model-index:
|
10 |
+
- name: torgo_xlsr_finetune_M01
|
11 |
+
results:
|
12 |
+
- task:
|
13 |
+
name: Automatic Speech Recognition
|
14 |
+
type: automatic-speech-recognition
|
15 |
+
dataset:
|
16 |
+
name: torgo
|
17 |
+
type: torgo
|
18 |
+
config: null
|
19 |
+
split: None
|
20 |
+
metrics:
|
21 |
+
- name: Wer
|
22 |
+
type: wer
|
23 |
+
value: 0.596551724137931
|
24 |
+
---
|
25 |
+
|
26 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
27 |
+
should probably proofread and complete it, then remove this comment. -->
|
28 |
+
|
29 |
+
# torgo_xlsr_finetune_M01
|
30 |
+
|
31 |
+
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the torgo dataset.
|
32 |
+
It achieves the following results on the evaluation set:
|
33 |
+
- Loss: 1.7808
|
34 |
+
- Wer: 0.5966
|
35 |
+
|
36 |
+
## Model description
|
37 |
+
|
38 |
+
More information needed
|
39 |
+
|
40 |
+
## Intended uses & limitations
|
41 |
+
|
42 |
+
More information needed
|
43 |
+
|
44 |
+
## Training and evaluation data
|
45 |
+
|
46 |
+
More information needed
|
47 |
+
|
48 |
+
## Training procedure
|
49 |
+
|
50 |
+
### Training hyperparameters
|
51 |
+
|
52 |
+
The following hyperparameters were used during training:
|
53 |
+
- learning_rate: 0.0001
|
54 |
+
- train_batch_size: 8
|
55 |
+
- eval_batch_size: 8
|
56 |
+
- seed: 42
|
57 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
58 |
+
- lr_scheduler_type: linear
|
59 |
+
- lr_scheduler_warmup_steps: 1000
|
60 |
+
- num_epochs: 30
|
61 |
+
|
62 |
+
### Training results
|
63 |
+
|
64 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
65 |
+
|:-------------:|:-----:|:-----:|:---------------:|:------:|
|
66 |
+
| 3.3568 | 0.45 | 500 | 3.2373 | 1.0 |
|
67 |
+
| 2.6662 | 0.89 | 1000 | 1.9097 | 0.9707 |
|
68 |
+
| 1.4433 | 1.34 | 1500 | 1.8284 | 0.8819 |
|
69 |
+
| 1.0086 | 1.79 | 2000 | 1.5207 | 0.8233 |
|
70 |
+
| 0.8253 | 2.24 | 2500 | 1.5046 | 0.7888 |
|
71 |
+
| 0.7438 | 2.68 | 3000 | 1.4814 | 0.7457 |
|
72 |
+
| 0.6364 | 3.13 | 3500 | 1.5690 | 0.7448 |
|
73 |
+
| 0.593 | 3.58 | 4000 | 1.7702 | 0.7293 |
|
74 |
+
| 0.555 | 4.03 | 4500 | 1.5775 | 0.7078 |
|
75 |
+
| 0.479 | 4.47 | 5000 | 1.4384 | 0.7026 |
|
76 |
+
| 0.4735 | 4.92 | 5500 | 1.6368 | 0.6940 |
|
77 |
+
| 0.4284 | 5.37 | 6000 | 1.9717 | 0.6879 |
|
78 |
+
| 0.429 | 5.81 | 6500 | 1.4882 | 0.6578 |
|
79 |
+
| 0.4128 | 6.26 | 7000 | 1.4697 | 0.6664 |
|
80 |
+
| 0.3869 | 6.71 | 7500 | 1.5555 | 0.6647 |
|
81 |
+
| 0.3725 | 7.16 | 8000 | 1.7755 | 0.6664 |
|
82 |
+
| 0.3104 | 7.6 | 8500 | 1.4753 | 0.65 |
|
83 |
+
| 0.3355 | 8.05 | 9000 | 1.5212 | 0.6526 |
|
84 |
+
| 0.3119 | 8.5 | 9500 | 1.6810 | 0.6345 |
|
85 |
+
| 0.3186 | 8.94 | 10000 | 2.2611 | 0.6534 |
|
86 |
+
| 0.3065 | 9.39 | 10500 | 1.2431 | 0.6397 |
|
87 |
+
| 0.2862 | 9.84 | 11000 | 1.5408 | 0.6371 |
|
88 |
+
| 0.2628 | 10.29 | 11500 | 2.0784 | 0.6474 |
|
89 |
+
| 0.2792 | 10.73 | 12000 | 1.6698 | 0.6293 |
|
90 |
+
| 0.2944 | 11.18 | 12500 | 1.7610 | 0.6336 |
|
91 |
+
| 0.237 | 11.63 | 13000 | 1.5215 | 0.6371 |
|
92 |
+
| 0.2471 | 12.08 | 13500 | 1.6935 | 0.65 |
|
93 |
+
| 0.2154 | 12.52 | 14000 | 1.7569 | 0.6319 |
|
94 |
+
| 0.2367 | 12.97 | 14500 | 1.6718 | 0.6284 |
|
95 |
+
| 0.2499 | 13.42 | 15000 | 1.4102 | 0.6302 |
|
96 |
+
| 0.2306 | 13.86 | 15500 | 1.5523 | 0.6147 |
|
97 |
+
| 0.2337 | 14.31 | 16000 | 1.8419 | 0.6129 |
|
98 |
+
| 0.206 | 14.76 | 16500 | 1.5285 | 0.6069 |
|
99 |
+
| 0.2098 | 15.21 | 17000 | 1.7365 | 0.6198 |
|
100 |
+
| 0.1994 | 15.65 | 17500 | 1.6756 | 0.6052 |
|
101 |
+
| 0.1943 | 16.1 | 18000 | 1.9950 | 0.6241 |
|
102 |
+
| 0.2077 | 16.55 | 18500 | 1.6966 | 0.6121 |
|
103 |
+
| 0.2135 | 16.99 | 19000 | 1.7379 | 0.6233 |
|
104 |
+
| 0.1721 | 17.44 | 19500 | 2.1036 | 0.6147 |
|
105 |
+
| 0.1876 | 17.89 | 20000 | 1.7061 | 0.6129 |
|
106 |
+
| 0.187 | 18.34 | 20500 | 1.7549 | 0.5897 |
|
107 |
+
| 0.1699 | 18.78 | 21000 | 1.7215 | 0.6009 |
|
108 |
+
| 0.1833 | 19.23 | 21500 | 1.5892 | 0.6069 |
|
109 |
+
| 0.1545 | 19.68 | 22000 | 1.7657 | 0.6155 |
|
110 |
+
| 0.1624 | 20.13 | 22500 | 1.5143 | 0.6103 |
|
111 |
+
| 0.1631 | 20.57 | 23000 | 1.4752 | 0.5974 |
|
112 |
+
| 0.1476 | 21.02 | 23500 | 1.5408 | 0.6 |
|
113 |
+
| 0.141 | 21.47 | 24000 | 1.7880 | 0.6112 |
|
114 |
+
| 0.1617 | 21.91 | 24500 | 1.7662 | 0.6190 |
|
115 |
+
| 0.1435 | 22.36 | 25000 | 1.8012 | 0.6034 |
|
116 |
+
| 0.1456 | 22.81 | 25500 | 1.7721 | 0.6052 |
|
117 |
+
| 0.147 | 23.26 | 26000 | 1.6913 | 0.6 |
|
118 |
+
| 0.137 | 23.7 | 26500 | 1.8138 | 0.5983 |
|
119 |
+
| 0.1269 | 24.15 | 27000 | 2.0274 | 0.5974 |
|
120 |
+
| 0.1203 | 24.6 | 27500 | 1.8193 | 0.5879 |
|
121 |
+
| 0.1321 | 25.04 | 28000 | 1.7929 | 0.5853 |
|
122 |
+
| 0.1333 | 25.49 | 28500 | 1.9791 | 0.6017 |
|
123 |
+
| 0.1061 | 25.94 | 29000 | 1.7340 | 0.5957 |
|
124 |
+
| 0.1099 | 26.39 | 29500 | 1.7547 | 0.6017 |
|
125 |
+
| 0.111 | 26.83 | 30000 | 1.7777 | 0.5991 |
|
126 |
+
| 0.1239 | 27.28 | 30500 | 1.7383 | 0.5991 |
|
127 |
+
| 0.1179 | 27.73 | 31000 | 1.8627 | 0.6009 |
|
128 |
+
| 0.0995 | 28.18 | 31500 | 1.7775 | 0.5966 |
|
129 |
+
| 0.0924 | 28.62 | 32000 | 1.8651 | 0.5966 |
|
130 |
+
| 0.1144 | 29.07 | 32500 | 1.7918 | 0.5974 |
|
131 |
+
| 0.1082 | 29.52 | 33000 | 1.7697 | 0.5957 |
|
132 |
+
| 0.1008 | 29.96 | 33500 | 1.7808 | 0.5966 |
|
133 |
+
|
134 |
+
|
135 |
+
### Framework versions
|
136 |
+
|
137 |
+
- Transformers 4.26.1
|
138 |
+
- Pytorch 2.1.0+cu118
|
139 |
+
- Datasets 2.15.0
|
140 |
+
- Tokenizers 0.13.3
|