Update README.md

README.md

@@ -39,3 +39,56 @@ SillyTavern Settings:
 Llama-3-Instruct-With-Names -->>>> Remove the Brackets in `[{{name}}]` or `[{{char}}]` or `[{{user}}]` within the instruction template to match the format used for training.
 ```
 
+---
+
+Example Dataset Entry:
+
+```
+    {
+        "token_length": x,
+        "Unique_chars": 3,
+        "conversations": [
+            {
+                "from": "system",
+                "value": "text"
+            },
+            {
+                "from": "human-chat",
+                "name": "User-1",
+                "value": "text"
+            },
+            {
+                "from": "gpt-chat",
+                "name": "User-2",
+                "value": "text"
+            },
+            {
+                "from": "human-chat",
+                "name": "User-3",
+                "value": "text"
+            },
+            {
+                "from": "gpt-chat",
+                "name": "User-2",
+                "value": "text"
+            },
+            {
+                "from": "human-chat",
+                "name": "User-1",
+                "value": "text"
+            },
+            {
+                "from": "gpt-chat",
+                "name": "User-3",
+                "value": "text"
+            },
+            {
+                "from": "gpt-chat",
+                "name": "User-2",
+                "value": "text"
+            }
+        ]
+    },
+```
+
+