Llama 3 and 3.3 Prompt Headers: Keep the Newlines

Meta's official Llama 3 documentation highlights that newlines (0x0A) are part of the prompt format, and the prompt must always end with the last user message followed by the assistant header <|start_header_id|>assistant<|end_header_id|>\n\n to prompt the model to start generating.

<|begin_of_text|><|start_header_id|>system<|end_header_id|>

You are a helpful AI assistant for travel tips and recommendations<|eot_id|><|start_header_id|>user<|end_header_id|>

What can I see in Prague?<|eot_id|><|start_header_id|>assistant<|end_header_id|>