Inconsistency in input template between paper and code

[ElementQi]

Really nice work!

I noticed that in the paper section 2.1:

we use the default official system prompt from Yang et al. (2024a) for Qwen2.5-Math, no system prompt for other models, and no user prompt for any models in our main experiments.

But in your training script for meta-llama/Llama-3.1-8B, line 76, you set the DeepSeek R1-Zero template as the input_template:

"A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. The reasoning process and answer are enclosed within and tags, respectively, i.e., reasoning process here answer here .\n\n<|user|>\n{}\n<|assistant|>\n"

There are some inconsistancy between the code and paper. Is there something typo or did I miss something important?