Hugging Face's logo

Laurie
/
gpt2-imdb-pos-v2

Model card Files
xet
Community
Laurie commited on
Commit
b389afd
·
1 Parent(s): 8eedc5e

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +10 -0
README.md ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ license: apache-2.0
3
+
4
+ datasets: mdb
5
+
6
+ language: English
7
+
8
+ 为了实现InstructGPT的SFT-RW-PPO
9
+
10
+ 此repo采用GPT2作为SFT模型，经过GPT2生成的文字再经过DistilBERT加以评估取生成positive的分数，再经过PPO优化