Skip to content

GitLab

Switch branch/tag
Find file Normal viewHistoryPermalink
contents.rst 2.5 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 


------------------------------------
ERNIE-M模型汇总
------------------------------------



下表汇总介绍了目前PaddleNLP支持的ERNIE-M模型对应预训练权重。

+----------------------------------------------------------------------------------+--------------+----------------------------------------------------------------------------------+
| Pretrained Weight                                                                | Language     | Details of the model                                                             |
+==================================================================================+==============+==================================================================================+
|``ernie-m-base``                                                                  | Multilingual | 12-layer, 768-hidden,                                                            |
|                                                                                  |              | 12-heads, _M parameters.                                                         |
|                                                                                  |              | Trained on  pseudo-parallel sentence                                             |
|                                                                                  |              | pairs on a monolingual corpus.                                                   |
+----------------------------------------------------------------------------------+--------------+----------------------------------------------------------------------------------+
|``ernie-m-large``                                                                 | Multilingual | 24-layer, 1024-hidden,                                                           |
|                                                                                  |              | 16-heads, _M parameters.                                                         |
|                                                                                  |              | Trained on  pseudo-parallel sentence                                             |
|                                                                                  |              | pairs on a monolingual corpus.                                                   |
+----------------------------------------------------------------------------------+--------------+----------------------------------------------------------------------------------+