Update README.md

README.md

@@ -28,27 +28,24 @@ This is an instruction fine-tuned llama-2 model, using synthetic data generated
 
 ### Prompt format
 
+The training code was updated to randomize newline vs space:
+https://github.com/jondurbin/qlora/blob/main/qlora.py#L559C1-L559C1
+
+
 ```
 A chat. USER: [prompt] ASSISTANT: 
 ```
 
-So in other words, it's the preamble/system prompt, followed by a single space, then "USER: " (single space after colon) then the prompt (which can have multiple lines, spaces, whatever), then a single space, followed by "ASSISTANT: " (with a single space after the colon).
-
-I updated the training code to randomize whitespace so it will hopefully be less important:
-https://github.com/jondurbin/qlora/blob/main/qlora.py#L559C1-L559C1
-
-It also uses the system prompt provided in the instructions, when available.
-
-In addition, I included a few thousand examples with alpaca style, but I would not recommend using this format.
+or
 
 ```
-Below is an instruction that describes a task. Write a response that appropriately completes the request.
-
-### Instruction:
-{prompt}
-### Response:
+A chat.
+USER: {prompt}
+ASSISTANT:
 ```
 
+So in other words, it's the preamble/system prompt, followed by a single space or newline, then "USER: " (single space after colon) then the prompt (which can have multiple lines, spaces, whatever), then a single space or newline, followed by "ASSISTANT: " (with a single space after the colon).
+
 ### Helpful usage tips
 
 *The prompts shown here are are just the text that would be included after USER: and before ASSISTANT: in the full prompt format above, the system prompt and USER:/ASSISTANT: have been omited for readability.*