Text Generation
Transformers
Safetensors
English
llama
conversational
Inference Endpoints
text-generation-inference
Sao10K commited on
Commit
ff0d22e
1 Parent(s): 381491e

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -4
README.md CHANGED
@@ -48,19 +48,18 @@ Testing Notes - Compared to v3.1
48
  **Recommended Samplers:**
49
 
50
  ```
51
- Temperature - 1.12 to 1.32
52
  Min-P - 0.075
53
- Top-K - 40
54
  Repetition Penalty - 1.1
55
  ```
56
 
57
  **Stopping Strings:**
58
 
59
  ```
60
- \n{{User}} # Or Equivalent, depending on Frontend
61
  <|eot_id|>
62
  <|end_of_text|>
63
- \n< # If there is leakage of XML tags in response. May happen Occasionally, Regenerate Answer as Needed. Happens rarely.
64
  ```
65
 
66
  **Prompting Template - Llama-3-Instruct**
 
48
  **Recommended Samplers:**
49
 
50
  ```
51
+ Temperature - 1.12-1.22
52
  Min-P - 0.075
53
+ Top-K - 50
54
  Repetition Penalty - 1.1
55
  ```
56
 
57
  **Stopping Strings:**
58
 
59
  ```
60
+ \n\n{{User}} # Or Equivalent, depending on Frontend
61
  <|eot_id|>
62
  <|end_of_text|>
 
63
  ```
64
 
65
  **Prompting Template - Llama-3-Instruct**