Skip to content

GitLab

Switch branch/tag
Find file Normal viewHistoryPermalink
contents.rst 2.91 KB
Newer Older
yuguo-Jack's avatar
llama_paddle  
yuguo-Jack committed
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26 


------------------------------------
UNIMO模型汇总
------------------------------------



下表汇总介绍了目前PaddleNLP支持的UNIMO模型对应预训练权重。
关于模型的具体细节可以参考对应链接。

+----------------------------------------------------------------------------------+--------------+----------------------------------------------------------------------------------+
| Pretrained Weight                                                                | Language     | Details of the model                                                             |
+==================================================================================+==============+==================================================================================+
|``unimo-text-1.0``                                                                | Chinese      | 12-layer, 768-hidden,                                                            |
|                                                                                  |              | 12-heads, 99M parameters.                                                        |
|                                                                                  |              | UNIMO-text-1.0 model.                                                            |
+----------------------------------------------------------------------------------+--------------+----------------------------------------------------------------------------------+
|``unimo-text-1.0-lcsts-new``                                                      | Chinese      | 12-layer, 768-hidden,                                                            |
|                                                                                  |              | 12-heads, 99M parameters.                                                        |
|                                                                                  |              | Finetuned on lcsts_new dataset.                                                  |
+----------------------------------------------------------------------------------+--------------+----------------------------------------------------------------------------------+
|``unimo-text-1.0-large``                                                          | Chinese      | 24-layer, 768-hidden,                                                            |
|                                                                                  |              | 16-heads, 316M parameters.                                                       |
|                                                                                  |              | UNIMO-text-1.0 large model.                                                      |
+----------------------------------------------------------------------------------+--------------+----------------------------------------------------------------------------------+