Tanrei commited on
Commit
2cb2fb1
·
1 Parent(s): 12392f9

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +4 -16
README.md CHANGED
@@ -12,26 +12,14 @@ General-purpose Swich transformer based Japanese language model
12
 
13
  ```python
14
  >>> from transformers import AutoModel, AutoTokenizer
15
- >>> model = AutoModel.from_pretrained("Tanrei/GPTSAN-japanese")
16
- >>> tokenizer = AutoTokenizer.from_pretrained("Tanrei/GPTSAN-japanese")
17
- >>> x_tok = tokenizer.encode("武田信玄は、")
18
- >>> model = model.cuda()
19
- >>> res = model.generator.generate_lm(x_tok, tokenizer)
20
- >>> res[0]
21
- '勝頼の父であり、天正四年(1576)に死去するまで甲府14万石の大名として甲府を治めた戦国大名ですが...'
22
- ```
23
 
24
- ## Masked Language Model
25
-
26
- ```python
27
- >>> from transformers import AutoModel, AutoTokenizer
28
  >>> model = AutoModel.from_pretrained("Tanrei/GPTSAN-japanese")
29
  >>> tokenizer = AutoTokenizer.from_pretrained("Tanrei/GPTSAN-japanese")
30
- >>> x_tok = tokenizer.encode("武田信玄は、<|inputmask|>時代ファンならぜひ押さえ<|inputmask|>きたい名将の一人。")
31
  >>> model = model.cuda()
32
- >>> res = model.generator.predict_mlm(x_tok, tokenizer)
33
- >>> res[0]
34
- '武田信玄は、戦国時代ファンならぜひ押さえておきたい名将の一人。'
35
  ```
36
 
37
 
 
12
 
13
  ```python
14
  >>> from transformers import AutoModel, AutoTokenizer
 
 
 
 
 
 
 
 
15
 
 
 
 
 
16
  >>> model = AutoModel.from_pretrained("Tanrei/GPTSAN-japanese")
17
  >>> tokenizer = AutoTokenizer.from_pretrained("Tanrei/GPTSAN-japanese")
18
+ >>> x_tok = tokenizer.encode("武田信玄は、", return_tensors="pt")
19
  >>> model = model.cuda()
20
+ >>> c = model.generate(x_tok.cuda(), max_new_tokens=50, random_seed=63)
21
+ >>> tokenizer.decode(c[0])
22
+ '武田信玄は、戦国の頃より「智勇兼備」した英雄として織田信長に比されてきた戦国武将であり、...'
23
  ```
24
 
25