license: apache-2.0
tags:
- 训练生成
metrics:
- 准确率
model-index:
- name: 结果模型
results: []
结果模型
本模型是基于facebook/wav2vec2-large-xlsr-53在指定数据集上微调得到的版本。在评估集上取得了如下表现：
模型描述
该模型可处理乌尔都语音频数据，并将其分类为以下情感类别：
训练与评估数据
数据集可通过以下链接获取：
https://www.kaggle.com/datasets/kingabzpro/urdu-emotion-dataset
训练流程
训练代码详见：
https://www.kaggle.com/code/chtalhaanwar/urdu-emotions-hf
训练超参数
训练过程中采用以下超参数配置：
- 学习率: 5e-05
- 训练批大小: 32
- 评估批大小: 32
- 随机种子: 42
- 优化器: 带betas=(0.9,0.999)和epsilon=1e-08的Adam优化器
- 学习率调度器类型: 线性
- 训练轮次: 50
- 混合精度训练: 原生AMP
训练结果
| 训练损失 |
轮次 |
步数 |
验证损失 |
准确率 |
| 1.3838 |
1.0 |
10 |
1.3907 |
0.225 |
| 1.3732 |
2.0 |
20 |
1.3872 |
0.2125 |
| 1.3354 |
3.0 |
30 |
1.3116 |
0.6625 |
| 1.2689 |
4.0 |
40 |
1.1820 |
0.6375 |
| 1.1179 |
5.0 |
50 |
1.0075 |
0.7 |
| 0.9962 |
6.0 |
60 |
0.8707 |
0.7125 |
| 0.8842 |
7.0 |
70 |
0.7485 |
0.7625 |
| 0.786 |
8.0 |
80 |
0.6326 |
0.8 |
| 0.6757 |
9.0 |
90 |
0.5995 |
0.8 |
| 0.6104 |
10.0 |
100 |
0.4835 |
0.825 |
| 0.5821 |
11.0 |
110 |
0.3886 |
0.9 |
| 0.4721 |
12.0 |
120 |
0.3935 |
0.8625 |
| 0.3976 |
13.0 |
130 |
0.3020 |
0.925 |
| 0.4483 |
14.0 |
140 |
0.3171 |
0.9 |
| 0.2665 |
15.0 |
150 |
0.3016 |
0.9125 |
| 0.2119 |
16.0 |
160 |
0.2722 |
0.925 |
| 0.3376 |
17.0 |
170 |
0.3163 |
0.8875 |
| 0.1518 |
18.0 |
180 |
0.2681 |
0.9125 |
| 0.1559 |
19.0 |
190 |
0.3074 |
0.925 |
| 0.1031 |
20.0 |
200 |
0.3526 |
0.8875 |
| 0.1557 |
21.0 |
210 |
0.2254 |
0.9375 |
| 0.0846 |
22.0 |
220 |
0.2410 |
0.9375 |
| 0.0733 |
23.0 |
230 |
0.2369 |
0.925 |
| 0.0964 |
24.0 |
240 |
0.2273 |
0.9375 |
| 0.0574 |
25.0 |
250 |
0.2066 |
0.95 |
| 0.1113 |
26.0 |
260 |
0.2941 |
0.9125 |
| 0.1313 |
27.0 |
270 |
0.2715 |
0.925 |
| 0.0851 |
28.0 |
280 |
0.1725 |
0.9625 |
| 0.0402 |
29.0 |
290 |
0.2221 |
0.95 |
| 0.1075 |
30.0 |
300 |
0.2199 |
0.9625 |
| 0.0418 |
31.0 |
310 |
0.1699 |
0.95 |
| 0.1869 |
32.0 |
320 |
0.2287 |
0.9625 |
| 0.0637 |
33.0 |
330 |
0.3230 |
0.9125 |
| 0.0483 |
34.0 |
340 |
0.1602 |
0.975 |
| 0.0891 |
35.0 |
350 |
0.1615 |
0.975 |
| 0.0359 |
36.0 |
360 |
0.1571 |
0.975 |
| 0.1006 |
37.0 |
370 |
0.1809 |
0.9625 |
| 0.0417 |
38.0 |
380 |
0.1923 |
0.9625 |
| 0.0346 |
39.0 |
390 |
0.2035 |
0.9625 |
| 0.0417 |
40.0 |
400 |
0.1737 |
0.9625 |
| 0.0396 |
41.0 |
410 |
0.1833 |
0.9625 |
| 0.0202 |
42.0 |
420 |
0.1946 |
0.9625 |
| 0.0137 |
43.0 |
430 |
0.1785 |
0.9625 |
| 0.0214 |
44.0 |
440 |
0.1841 |
0.9625 |
| 0.0304 |
45.0 |
450 |
0.1690 |
0.9625 |
| 0.0199 |
46.0 |
460 |
0.1646 |
0.975 |
| 0.0122 |
47.0 |
470 |
0.1622 |
0.975 |
| 0.0324 |
48.0 |
480 |
0.1615 |
0.975 |
| 0.0269 |
49.0 |
490 |
0.1625 |
0.975 |
| 0.0245 |
50.0 |
500 |
0.1638 |
0.975 |
框架版本
- Transformers 4.18.0
- Pytorch 1.11.0
- Datasets 2.1.0
- Tokenizers 0.12.1
PyTorch
Transformers%20--%3e%3cdefs%3e%3cstyle%3e%20.st0%20{%20fill:%20none;%20}%20.st1%20{%20fill:%20%235ce6cf;%20}%20.st2%20{%20fill:%20%230080ff;%20}%20.st3%20{%20clip-path:%20url(%23clippath);%20}%20%3c/style%3e%3cclipPath%20id='clippath'%3e%3crect%20class='st0'%20y='0'%20width='120'%20height='32'/%3e%3c/clipPath%3e%3c/defs%3e%3cg%20class='st3'%3e%3cg%3e%3cpath%20class='st1'%20d='M22.4,32H3.2c-1,0-2-.5-2.6-1.4-.6-.9-.7-2-.4-2.9L9.8,2.1C10.3.8,11.5,0,12.8,0s2.5.8,3,2.1l9.6,25.6c.4,1,.2,2.1-.4,2.9-.6.9-1.6,1.4-2.6,1.4ZM7.8,25.6h10l-5-13.3-5,13.3Z'/%3e%3cpath%20class='st2'%20d='M33.6,32c-1.3,0-2.5-.8-3-2.1L21,4.3c-.6-1.7.2-3.5,1.9-4.1,1.7-.6,3.5.2,4.1,1.9l9.6,25.6c.6,1.7-.2,3.5-1.9,4.1-.4.1-.8.2-1.1.2h0Z'/%3e%3cpath%20d='M71.8,9.1h-6.9v4.6h6.9c2.5,0,4.6,2,4.6,4.6h-6.9c-3.8,0-6.9,3.1-6.9,6.9s3.1,6.9,6.9,6.9h2.3c1.7,0,3.2-.4,4.6-1.2v1.2h4.6v-13.7c0-5-4.1-9.1-9.1-9.1h0ZM71.8,27.4h-2.3c-1.3,0-2.3-1-2.3-2.3s1-2.3,2.3-2.3h6.9c0,2.5-2,4.6-4.6,4.6h0Z'/%3e%3cpath%20d='M52.3,9.1h-4.6V0h-4.6V32h9.1c5,0,9.1-4.1,9.1-9.1v-4.6c0-5.1-4.1-9.1-9.1-9.1h0ZM56.9,22.9c0,2.5-2,4.6-4.6,4.6h-4.6v-13.7h4.6c2.5,0,4.6,2,4.6,4.6v4.6h0Z'/%3e%3cpath%20d='M110.9,13.7h9.1v-4.6h-9.1c-5.1,0-9.1,4.1-9.1,9.1v4.6c0,5.1,4.1,9.1,9.1,9.1h9.1v-4.6h-9.1c-2.5,0-4.6-2-4.6-4.6h13.7v-4.6h-13.7c0-2.5,2-4.6,4.6-4.6h0Z'/%3e%3cpath%20d='M93.6,18.3h-4.6c-1.3,0-2.3-1-2.3-2.3s1-2.3,2.3-2.3h10.3v-4.6h-10.3c-3.8,0-6.9,3.1-6.9,6.9s3.1,6.9,6.9,6.9h4.6c1.3,0,2.3,1,2.3,2.3s-1,2.3-2.3,2.3h-10.3v4.6h10.3c3.8,0,6.9-3.1,6.9-6.9s-3.1-6.9-6.9-6.9Z'/%3e%3c/g%3e%3c/g%3e%3c/svg%3e)