Safetensors
openlm
text
mathewhe commited on
Commit
16d463a
·
1 Parent(s): d91e224

Improve the chat prompt explanation and example

Browse files
Files changed (1) hide show
  1. README.md +11 -0
README.md CHANGED
@@ -61,11 +61,22 @@ Each message should be a dict with two keys:
61
 
62
  For example:
63
  ```python
 
 
 
 
64
  messages = [
65
  {"role": "user", "content": "Solve for x: 3x=4"},
66
  {"role": "assistant", "content": "3x=4\n(3x)/3=(4)/3\nx=4/3"},
67
  {"role": "user", "content": "Please explain your work."},
68
  ]
 
 
 
 
 
 
 
69
  ```
70
 
71
  See the example code in the included `chat_class.py` module for more details.
 
61
 
62
  For example:
63
  ```python
64
+ from transformers import AutoTokenizer
65
+
66
+ tokenizer = AutoTokenizer.from_pretrained("mathewhe/DCLM-7B-Chat")
67
+
68
  messages = [
69
  {"role": "user", "content": "Solve for x: 3x=4"},
70
  {"role": "assistant", "content": "3x=4\n(3x)/3=(4)/3\nx=4/3"},
71
  {"role": "user", "content": "Please explain your work."},
72
  ]
73
+ print(tokenizer.apply_chat_template(messages, tokenize=False)
74
+ ```
75
+ outputs
76
+ ```
77
+ <|endoftext|>[INST] Solve for x: 3x=4 [/INST][ASST] 3x=4
78
+ (3x)/3=(4)/3
79
+ x=4/3 [/ASST]<|endoftext|><|endoftext|>[INST] Please explain your work [/INST]
80
  ```
81
 
82
  See the example code in the included `chat_class.py` module for more details.