ehartford commited on
Commit
223de7e
·
verified ·
1 Parent(s): 8312098

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -6
README.md CHANGED
@@ -6,7 +6,6 @@ tags:
6
  - persona
7
  - character
8
  - dialogue
9
- - chatml
10
  - interactive-fiction
11
  - NPC
12
  - erotic-roleplay
@@ -33,8 +32,7 @@ Fine-tuned by **BaiAI** and **Eric Hartford (QuixiAI)** using QLoRA + DPO on lar
33
 
34
  # **Supported Prompt Format**
35
 
36
- Ina **does not** use ChatML.
37
- It uses the **Llama-3.1 special token header format**, exactly like this:
38
 
39
  ```text
40
  <|begin_of_text|><|start_header_id|>system<|end_header_id>
@@ -65,11 +63,9 @@ Ah, welcome, weary traveler! Pull up a stool — our lamb stew is rich with rose
65
  <|eot_id|>
66
  ```
67
 
68
- This is **the officially correct format** for how Ina was trained.
69
-
70
  ---
71
 
72
- # **What Ina Is Used For**
73
 
74
  * Unrestricted explicit ERP with long-term consistency
75
  * OnlyFans/Fansly DM + caption generation
 
6
  - persona
7
  - character
8
  - dialogue
 
9
  - interactive-fiction
10
  - NPC
11
  - erotic-roleplay
 
32
 
33
  # **Supported Prompt Format**
34
 
35
+ Ina uses the **Llama-3.1 special token header format**, exactly like this:
 
36
 
37
  ```text
38
  <|begin_of_text|><|start_header_id|>system<|end_header_id>
 
63
  <|eot_id|>
64
  ```
65
 
 
 
66
  ---
67
 
68
+ # **Use Cases**
69
 
70
  * Unrestricted explicit ERP with long-term consistency
71
  * OnlyFans/Fansly DM + caption generation