Hugging Face's logo

WENGSYX
/
CirBERTa-Chinese-Base

Transformers
PyTorch
deberta-v2
Inference Endpoints
Model card Files Community
YixuanWeng commited on
Commit
6b37856
•
1 Parent(s): 8939cf3

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -3
README.md CHANGED
@@ -1,13 +1,13 @@
1
  # CirBERTa
2
- Apply the Circular to the Pretraining Model
 
3
 
4
 
5
- ### 使用通用语料（WuDao 200G) 进行无监督预训练
6
 
7
  | 预训练模型 | 学习率 | batchsize | 设备 | 语料库 | 时间 | 优化器 |
8
  | --------------------- | ------ | --------- | ------ | ------ | ---- | ------ |
9
  | CirBERTa-Chinese-Base | 1e-5 | 256 | 10张3090+3张A100 | 200G | 2月 | AdamW |
10
-
11
  在多项中文理解任务上，CirBERTa-Base模型超过MacBERT-Chinese-Large/RoBERTa-Chinese-Large
12
 
13
 
 
1
  # CirBERTa
2
+ ### Apply the Circular to the Pretraining Model
3
+
4
 
5
 
 
6
 
7
  | 预训练模型 | 学习率 | batchsize | 设备 | 语料库 | 时间 | 优化器 |
8
  | --------------------- | ------ | --------- | ------ | ------ | ---- | ------ |
9
  | CirBERTa-Chinese-Base | 1e-5 | 256 | 10张3090+3张A100 | 200G | 2月 | AdamW |
10
+ 使用通用语料（WuDao 200G) 进行无监督预训练
11
  在多项中文理解任务上，CirBERTa-Base模型超过MacBERT-Chinese-Large/RoBERTa-Chinese-Large
12
 
13